HOW QUALITATIVE DATA

ANALYSIS HAPPENS

How Qualitative Data Analysis Happens: Moving Beyond “Themes Emerged”

(Expanded Edition) offers an in-­depth look into how qualitative social sci-
ence researchers studying a wide range of human experiences and dynamics
approach their data analyses. This expanded edition includes 13 new chap-
ters from a broad range of disciplines that document the stories about how
qualitative data analysis occurred.
Chapters for this expanded edition represent a diversity of disciplines
(e.g., criminology, family science, education, health, nutrition, sociology,
and sport psychology) that focus on the human experience and describe a
diversity of methodological approaches. These chapters may be used to
introduce readers to newer or innovative ways of analyzing data. It moves
beyond the usual vague statement of “themes emerged from the data” to
show readers how researchers actively and consciously arrive at their themes
and conclusions, revealing the complexity and time involved in making sense
of thousands of pages of interview data, multiple data sources, and diverse
types of data. The various authors provide detailed narratives into how they
analyzed their data from previous publications. The methodologies range
from arts-­based research, autoethnography, community-­based participatory
research, ethnography, and grounded theory to narrative analysis. The vol-
ume allows readers to be seemingly “in the room” with these international
scholars (representing Canada, the United States, Austria, Germany, the
United Kingdom, and the Philippines) and getting their own hands vicari-
ously dirty with the data.
 This new edition also includes a conclusion chapter in which the authors
reflect on commonalities across the chapters. Supplemental figures, images,
and screenshots, which are referred to in the chapters, are included in an
accompanying eResource (that can be accessed at [Link].
com/9781032183213), as well as links to the previously published work on
which the chapters are based. This book is an invaluable resource for experi-
enced and novice qualitative researchers throughout the social sciences, as
well as undergraduate and postgraduate students in the field.

Áine M. Humble, PhD, is Professor, Department of Family Studies and

Gerontology at Mount Saint Vincent University, Halifax, Nova Scotia,
Canada. She carries out primarily qualitative research in family studies and
in gerontology. Her current research interests focus on LGBTQ (lesbian,
gay, bisexual, transgender, queer) families and aging issues, older women
and healthy aging, qualitative research methods, and computer-assisted
qualitative data analysis software programs (CAQDAS). She has published
her work in scholarly journals such as the Canadian Journal on Aging, Family
Relations, Journal of Family Theory & Review, Journal of GLBT Family
Studies, and the Journal of Woman & Aging. She currently serves on the
editorial board of the Journal of Women and Aging, and she is an Editorial
Assistant for Forum Qualitative Sozialforschung / Forum: Qualitative Social
Research. From 2009 to 2023, Dr. Humble provided introductory and
advanced training on MAXQDA qualitative data analysis software through
her consulting company, Fada Research Consulting, to academic and gov-
ernment audiences.

M. Elise Radina, PhD, CFLE, is Professor, Department of Family Science

and Social Work, Miami University, Oxford, Ohio. Elise is a qualitative
methodologist whose research focuses broadly on families and health. She
primarily studies mid- and later-life women in family contexts (e.g., breast
cancer survivorship, positive aging, breast cancer-related lymphedema). She
has published over 60 peer-reviewed articles and book chapters. Dr. Radina’s
work has appeared in the Journal of Family Theory & Review, Family
Relations, the Journal of Women & Aging, Cancer Nursing, and the Journal
of Family Nursing. She serves on the Editorial Board for the Journal of
Family Theory & Review and has served as a guest co-editor on two special
issues for the journal. The special issue whose title was Qualitative
Methodology, Research, and Theory in Family Studies was co-guest edited
with Anisa Zvonkovic and Elizabeth Sharp.
HOW QUALITATIVE DATA
ANALYSIS HAPPENS
Moving Beyond “Themes Emerged”
Expanded Edition

Edited by
Áine M. Humble
M. Elise Radina
Designed cover image: getty images @ Dusan Stankovic
Second edition published 2025
by Routledge
605 Third Avenue, New York, NY 10158
and by Routledge
4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2025 selection and editorial matter, Áine M. Humble and M. Elise Radina;
individual chapters, the contributors
The right of Áine M. Humble and M. Elise Radina to be identified as the
authors of the editorial material, and of the authors for their individual
chapters, has been asserted in accordance with sections 77 and 78 of the
Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or
utilised in any form or by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying and recording, or in any
information storage or retrieval system, without permission in writing from the
publishers.
Trademark notice: Product or corporate names may be trademarks or registered
trademarks, and are used only for identification and explanation without intent
to infringe.
First edition published by Routledge 2018
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Names: Humble, Áine ( Áine M.), editor. | Radina, M. (M. Elise), editor.
Title: How qualitative data analysis happens : moving beyond “themes
emerged” / edited by Áine M. Humble and M. Elise Radina.
Description: 2nd edition. | New York, NY : Routledge, 2025. | Includes
bibliographical references and index.
Identifiers: LCCN 2024010460 (print) | LCCN 2024010461 (ebook) | ISBN
9781032183213 (hardback) | ISBN 9781032183220 (paperback) | ISBN
9781003254041 (ebook)
Subjects: LCSH: Social sciences--Research--Methodology. | Qualitative
research--Methodology.
Classification: LCC H62 .H655 2025 (print) | LCC H62 (ebook) | DDC
300.72/1--dc23/eng/20240315
LC record available at [Link]
LC ebook record available at [Link]
ISBN: 978-1-032-18321-3 (hbk)
ISBN: 978-1-032-18322-0 (pbk)
ISBN: 978-1-003-25404-1 (ebk)
DOI: 10.4324/9781003254041
Typeset in Times New Roman
by SPi Technologies India Pvt Ltd (Straive)
Access the Support Material (eResource): [Link]/9781032183213
CONTENTS

List of Figures viii

List of Tables x
List of Documents xii
Contributors xiii
Acknowledgments xvii

Introduction: How This Second (and Expanded)

Volume Happened 1
M. Elise Radina and Áine M. Humble

1 How Autoethnography Begins and Never Ends: A

Tracing of the “Self ” in Relation to #MeToo and
Higher Education 9
Angela Underhill

2 Using Institutional Ethnography to Trace the Ruling

of Weight Surveillance Work 26
Alexa R. Ferdinands and Kim D. Raine

3 Writing a New Materialist Ethnography on

Polyamorous Parents 44
Cornelia Schadler
vi Contents

4 Reading between the Lines of After Death

Communication Stories: Using Narrative Analysis to
Make the Implicit Explicit 60
Sara E. Hackett and Kate de Medeiros

5 Data Analytic Strategies Used in a Remote

Photovoice Project of Filipino Single Mothers during
the COVID-19 Pandemic 75
Dennis S. Erasga, Jerome V. Cleofas, Mary Rose Jean
Andrada-Poa, and Ronaldo F. Jabal

6 Phenomenological Analysis and Racial Socialization

Interpretation of Interviews with African American
Parents of Toddler Sons 93
Sheresa Boone Blanchard, Stephanie Irby Coard, and
Mariana Mereoiu

7 Black Feminist Theory and Thematic Analysis:

Analyzing the Motherwork of Black Women Nursing
Professionals during the COVID-19 Pandemic 110
Adrienne L. Edwards-Bianchi

8 Visualizing Relationships to Explore Opportunities

for Family Engagement in Diabetes and Hypertension
Management 126
Meredith P. Fort, Cornelia J. Santos, Maria de los
Angeles Villaverde, and Kelly R. Moore

9 Voices from Inside Prison: Centering People through

Intentional Sampling, Coding, and Analysis within
Large Research Teams 142
Danielle S. Rudes, Shannon Magnuson,
and Sydney N. Ingel

10 Triangulating Partners’ Views Over Time: Analyzing

Multiple Perspective Qualitative Longitudinal
Interviews on Non-normative Work-Care
Arrangements in the Transition to Parenthood in Practice 160
Susanne Vogl, Eva-Maria Schmidt, and Ulrike Zartler
 Contents vii

11 Media Priming and Racialized Production Decisions

in College Football Broadcasts: Extrapolating
Strategies for Analyzing Video Data 179
Sara E. Grummert and Siduri J. Haslerig

12 Reflections on Conducting Team-Based Qualitatively

Oriented Mixed Methods Research about Students
with Disabilities in STEM Clubs 199
Peggy Shannon-Baker, Karin Fisher, and Kania Greer

13 Mixing Methods to Advance our Understanding of

Parental Stress and Coping in Youth Sport 218
Sam N. Thrower, Travis E. Dorsch, Camilla J. Knight,
and Chris G. Harwood

14 Final Reflections 233

Áine M. Humble and M. Elise Radina

Index 246
FIGURES

Note: All figures except Figure 8.4 are located at the eResource
(Companion Website) at [Link]/9781032183213

1.1 The Viral Hashtag

1.2 An Early Depiction of an Ongoing Analysis
1.3 Moving from Prompts to Analysis
2.1 Mapping the Social Organization of Young People’s
Weight Surveillance Work
3.1 Tagging
3.2 Network of Tags
3.3 Referencing: Example 1
3.4 Referencing: Example 2
4.1 Example of Coded Text in Dedoose
4.2 Example of Memo Note
4.3 Output from Dedoose’s Packed Code Cloud Tool
5.1 Screenshot of the Shared Google Drive Folder with
Analysis Files
5.2 Coding of Selected Photographs
5.3 Flowchart for “Following a Thread”
5.4 Sharing Analytical Musings on Instant Messenger
7.1 Final Coding Tree: Major Theme, Subthemes, and Focused Codes
8.1 Sample Genogram from Original Article (Fort et al., 2020)
8.2 Sample Eco-Map in Spanish (with Identifying
Information Removed or Covered)
 Figures ix

8.3 Sample Eco-Map in English (with Identifying Information

Removed or Covered)
8.4 Spectrum of Types of Family Support 134
8.5 Composite Eco-Map from Original Article (Fort et al., 2020)
8.6 Excerpts from Two Transcripts with Coded Segments in
[Link]
9.1 [Link] File Organization
9.2 Dropbox File Organization
9.3 [Link] Coded Document
9.4 [Link] Code Comment for Right: Shower
9.5 [Link] Queried Code Report
9.6 Model of Deep Analytical Processing for Qualitative Data
9.7 Diagram of Contextualizing Analysis within Law and Policy
10.1 Screenshot of MAXQDA File with Documents,
Transcripts, Coding, and a Code Memo
11.1 Example of Video Data with Multiple Coders
11.2 Example of “Code Development” Memo
11.3 All Excerpts Coded for “Injury”
11.4 Comparing and Subcoding an Excerpt Coded for “Injury”
12.1 Circular-Based Joint Display Draft
12.2 Quadrant-Based Joint Display Draft
12.3 Final Published Quadrant Joint Display (Fisher et al., 2022)
13.1 Example Stressors from the Guidebook given to Participants
13.2 Example of How Qualitative Data were Collected
13.3 Example of How Quantitative Data were Collected
TABLES

Note: Tables with an * are located at the eResource (Companion

Website) at [Link]/9781032183213

0.1 Summary of Chapters, Primary Study Publication

Sources, and First Authors’ Academic Disciplines 5
1.1 Prompts for Analysis*
1.2 Original, Condensed Summary of Forces*
2.1 Sample Index Headings*
2.2 Things I Found Interesting and Worth Exploring 35
3.1 Comparison of Text before and after the Revision 57
4.1 Output from Dedoose’s Code Presence Tool*
4.2 Output from Dedoose’s Code Application Tool*
4.3 Output from Dedoose’s Code Co-occurrence Tool*
5.1 Sample Coding of Textual Data under “More Intimate
Moments with Children” 81
6.1 First Author Bracketing Excerpt 97
6.2 Sources for Interview Questions*
6.3 Excerpt from Family 1 Textural Description*
6.4 Excerpt from Family 1 Structural Description*
6.5 Excerpt from Family 1 Descriptive Summary*
6.6 Strategies to Determine Rigor 103
6.7 Manuscript Publication Journey 104
7.1 Examples of First Cycle Codes 117
7.2 Data Analysis Notes for Subthemes 120
8.1 Excerpts from the Initial Codebook 131
 Tables xi

8.2 Analysis Table for Manuscript*

9.1 Concentration of Reported Instances of Inaccessible Rights 153
9.2 Expressive and Instrumental Rationales 155
10.1 Comparison Levels in MPQLI Data and Related Aims 164
10.2 Excel Sheet Comparing Individuals on a Couple Level*
10.3 Couple Profile along Thematic Codes 166
10.4 Dissonance in Triangulating Perspectives: Julia and Jim
Talk about Their Plans for Taking Parental Leave in Wave 1 168
10.5 Dissonance in Triangulating Perspectives: Chris’ and
Claudia’s Story about Chris’ Re-entry to Work 170
10.6 Complementary Accounts in Triangulation: Maria and
Max Talk about Re-traditionalization 171
10.7 Excerpt from a Note on Clusters of Couples: Example of
a Typical Couple, Their Difficulties, and Feelings 172
10.8 Maria and Max’ Parental Care Plans Over Time 173
10.9 Breadwinning as Care? Draft of a Typology Table*
11.1 Components of Video Data 185
11.2 Excerpt from Coding Decision Log 189
11.3 Final Codebook*
11.4 Code Co-occurrences Chart*
11.5 Injury Data Being Further Analyzed in Excel*
12.1 Excerpt of Data Compiled Under the Theme of
“Recruitment” for Study 1*
12.2 Excerpt from the Table-Based Joint Display Draft*
13.1 Raw Data Extracted from the Online Survey*
13.2 Initial Coding of Participant Stressors*
13.3 Example of How Qualitative Stressor Data Were Coded at
a Lower-Order, Higher-Order, and General Dimension Level*
13.4 Example of How Qualitative Coping Strategy Data Were
Coded at a Lower-Order, Higher-Order, and General
Dimension Level*
DOCUMENTS

Note: All documents are located at the eResource (Companion

Website) at [Link]/9781032183213

2.1 Example of Color-Coded Interview Transcript

2.2 Sample Indexing of an Electronic Interview Transcript
2.3 Jenny’s Informant Summary
7.1 First Draft of Coded Transcript—Phoenix
7.2 Analytic Memo: Phoenix, Ella, and Tracy
7.3 Analytic Memo: Thrive, Farrah, and Sasha
8.1 Summary Slide of Family Engagement Strategies
8.2 List of Quotes from an [Link] Code Query
10.1 Evolving Theory Memo Case 11
12.1 Excerpt of Handwritten Field Notes from Study 1 Interview
12.2 Excerpt of an Interview from Study 1 with Margin Notes
and Color-Coded Categories
12.3 Example of Initial Coding from a Member of Group 1 for Study 2
12.4 Excerpt from the Code and Category List for Study 2
12.5 Revised Excerpt from the Code and Category List for Study 2
CONTRIBUTORS

Mary Rose Jean Andrada-­Poa, MA, is a PhD candidate in the Department of

Sociology and Behavioral Sciences, Dela Salle University, Manila,
Philippines.

Sheresa Boone Blanchard, PhD, is a senior education researcher and techni-

cal assistance specialist, SRI Education. She is also Associate Professor in
the Department of Human Development and Family Science, East Carolina
University, Greenville, North Carolina.

Jerome V. Cleofas, PhD, RN, is Associate Professor in the Department of

Sociology and Behavioral Sciences, De La Salle University, Manila,
Philippines.

Maria de los Angeles Villaverde, MS, is retired. She previously worked at the
Centers for American Indian and Alaska Native Health, Colorado School
of Public Health, Aurora, Colorado.

Kate de Medeiros, PhD, is Professor, Department of Sociology and

Anthropology, Concordia University, Montreal, Quebec, Canada.

Travis E. Dorsch, PhD, is Associate Professor & Graduate Program Director,

Department of Human Development and Family Studies, Emma Eccles
Jones College of Education and Human Services, Utah State University,
Logan, Utah.
xiv Contributors

Adrienne L. Edwards-­Bianchi, PhD, CFLE, is Core Faculty, Infant and Early

Childhood Development, Fielding Graduate University, Santa Barbara,
California.

Dennis S. Erasga, PhD, is Professor, Department of Sociology and Behavioral

Sciences, De La Salle University, Manila, Philippines.

Alexa R. Ferdinands, PhD, RD, is Assistant Professor, Health and Nutrition,

Faculty of Science and Technology, Athabasca University, Athabasca,
Alberta, Canada.

Karin Fisher, PhD, is Associate Professor, College of Education, Georgia

Southern University, Statesboro, Georgia.

Meredith P. Fort, PhD, MPH is Research Assistant Professor, Department

of Health Systems, Management and Policy, Centers for American Indian
and Alaska Native Health, Colorado School of Public Health, Aurora,
Colorado.

Kania Greer, Ed.D, is Coordinator, Center for STEM Education, College of

Education, Georgia Southern University, Statesboro, Georgia.

Sara E. Grummert, PhD, is Academic Director, Institute for Mixed Methods

Research, Manhattan Beach, California.

Sara E. Hackett, PhD, is Assistant Professor of Instruction, School of Aging

Studies, University of South Florida, Tampa, Florida.

Chris G. Harwood, PhD, is Professor of Sport Psychology, Sport, Health

and Performance Enhancement (SHAPE) Research Centre, Department of
Sport Science, School of Science and Technology, Nottingham Trent
University, Nottingham, United Kingdom.

Siduri J. Haslerig, PhD, is Associate Dean and Associate Professor, Boise

State University, Boise, Idaho.

Sydney N. Ingel, MA, is a doctoral candidate, Department of Criminology,

Law and Society, George Mason University, Fairfax, Virginia.

Stephanie Irby Coard, PhD, is Professor, Department of Human Development

and Family Studies, University of North Carolina – Greensboro, Greensboro,
North Carolina.
 Contributors xv

Ronaldo F. Jabal, MA, is a PhD candidate, Department of Sociology and

Behavioral Sciences, Dela Salle University, Manila, Philippines.

Camilla J. Knight, PhD, is Professor of Sport Psychology and Youth Sport,

Department of Sport and Exercise Sciences, Swansea University, United
Kingdom. She is also associated with the Department of Sport Science and
Physical Education, University of Agder, Norway.

Shannon Magnuson, PhD, is Senior Associate, Justice System Partners,

South Easton, Massachusetts.

Mariana Mereoiu, PhD, is Associate Professor, School of Inclusive Teacher

Education, Bowling Green State University, Bowling Green, Ohio.

Kelly R. Moore, MD, FAAP, is Associate Professor and retired. She formerly
worked at the Centers for American Indian and Alaska Native Health,
Department of Community and Behavioral Health, Colorado School of
Public Health, Aurora, Colorado.

Kim D. Raine, PhD, FCAHS, is Distinguished Professor Emeritus, School of

Public Health, University of Alberta, Edmonton, Alberta, Canada.

Danielle S. Rudes, PhD, is Professor, Department of Criminal Justice &

Criminology, Sam Houston State University, and Deputy Director, Center
for Advancing Correctional Excellence (ACE!), Huntsville, Texas.

Cornelia J. Santos, PhD, is Assistant Professor, Colorado School of Public

Health, Aurora, Colorado. She was formerly with the Centers for American
Indian and Alaska Native Health.

Cornelia Schadler, PhD, is Senior Lecturer, Department of Education,

University of Vienna, Vienna, Austria.

Eva-­Maria Schmidt, PhD, Senior Scientist, Department of Sociology and

Austrian Institute for Family Studies, University of Vienna, Vienna, Austria.

Peggy Shannon-­Baker, PhD, is Associate Professor, Department of

Curriculum, Foundations, and Reading, Georgia Southern University,
Statesboro, Georgia.

Sam N. Thrower, PhD, is Senior Lecturer in Sport & Exercise Psychology,

Department of Sport, Health Sciences and Social Work, Faculty of Health
& Life Sciences. Oxford Brookes University. Oxford, United Kingdom.
xvi Contributors

Angela Underhill, MSc, is a doctoral candidate in the Department of Family

Relations and Applied Nutrition, University of Guelph, Guelph, Ontario,
Canada.

Susanne Vogl, PhD, is Full Professor, Division of Sociology and Research

Methods, Department of Social Sciences, University of Stuttgart, Stuttgart,
Germany.

Ulrike Zartler, PhD, is Full Professor, Department of Sociology, University

of Vienna, Vienna, Austria.
ACKNOWLEDGMENTS

We would like to thank the entire team at Taylor & Francis for their support
and assistance throughout this process. We are also grateful for the fine work
of Michael Goldstein, who provided us with valuable indexing services.
Finally, we both know that work such as this takes time and energy. We are
therefore grateful to our families, colleagues, and friends for their unwaver-
ing support to us.
INTRODUCTION
How This Second (and Expanded) Volume
Happened

M. Elise Radina and Áine M. Humble

In the Introduction (Radina & Humble, 2019) to the first edition of this book
How Qualitative Data Analysis Happens: Moving Beyond “Themes Emerged”,
we focused on uncovering the messy—but energizing—complexity of qualita-
tive methodologies within the fields of family science and gerontology. The
result was a collection of chapters that provided a behind-­the-­scenes look at
how a diversity of researchers approached their analyses.
For this expanded volume, we have curated 13 new chapters from a broad
range of disciplines that document the data analysis details for recently pub-
lished work (i.e., in the last five years or so). Like others who have given
advice about qualitative research (e.g., Goldberg & Allen, 2015), the intent
of these chapters is not to provide a “step-­by-­step formula” (p. 3) for carry-
ing out data analysis. Instead, the purpose is to provide readers with a vari-
ety of stories about how data analysis actually occurred. To be mindful that
there is not just one correct way to analyze qualitative data, we have inten-
tionally identified chapter authors for this book whose work represents var-
ied ways of how data can be analyzed. The book may also introduce readers
to newer or innovative ways of analyzing data.
In the interim and as an outcome of the earlier edition, we shared a 2014
reading by Saldaña and engaged in our own dialogue. Saldaña discusses how
researchers label themselves (or not) as a certain type of qualitative
researcher. Elise commented about how much she has identified with
Saldaña’s point about not embracing labels, stating that she is “more inter-
ested in the question and getting to answers in whatever way makes sense
than holding tightly to one school of thought” (Radina, personal communi-
cation, June 1, 2023). Saldaña discusses how qualitative research is like

DOI: 10.4324/9781003254041-1
2 M. Elise Radina and Áine M. Humble

perfecting a craft as a craftsman who chooses the right tool(s) for the right
project. In curating this edition of the book, we focused routinely on trying
to draw out the authors with regard to their positionality in a desire to
understand the human tool(s) that were used in the craft and what they are
about. We regularly asked the authors during chapter revisions to review
their humanness because understanding where the individuals are in the
work is essential to feeling connected to the story being told.
As we noted in the introduction to the first edition (Radina & Humble,
2019), numerous researchers have identified that far too often, little detail is
provided on how data analysis actually occurs. For example, Richards and
Morse (2013) note that “how abstracting is done is far too often presented
as a mystery. A constant theme in the literature is that the theory ‘emerges’
from the data, but we have never had the privilege of seeing such an aspira-
tion” (p. 120). Similarly, Dickie (2003) notes that

just as the account of the findings needs to tell a story, so does the account
of the process itself. This account needs to go far beyond the use of terms
such as “open coding”, “triangulation of data sources”, and “line by line
reading”. Writing a separate data analysis section in a qualitative study
can too easily result in a cursory text composed more of labels than
descriptions.
(p. 50)

The reality for both researchers and their readership is that they are left with
limited and vague representations of qualitative analysis processes that offer
little explanation about the realities of what such processes look like. This
limited glimpse can lead to misunderstandings and misconceptions about the
nature of the methodology—for example, its (often inevitable) complex messi-
ness, decision-­making of the researcher and/or research team, how techniques
are executed perhaps differently by various researchers, and attention to rigor.
This edited volume offers an expanded look into how qualitative social
science researchers approach their data analysis. As with the first volume, we
intentionally included a variety of methodologies (e.g., photovoice, institu-
tional ethnography, and hierarchical mapping) and theoretical perspectives,
some of which were new to us, such as New Materialism. This volume also
builds on the earlier edited volume in three distinct ways. First, we broad-
ened our inclusion of social science disciplines beyond family science and
gerontology to include a variety of social sciences (e.g., public health, psy-
chology, education, and criminal justice) that put the lived experiences of
humans at the center of their inquiries. This has resulted in 13 new chapters.
Second, our first edition was mainly limited to researchers in North America.
This expanded addition includes a more global representation of researchers
 Introduction 3

from outside North America including Austria, Germany, and the

Philippines. Third, based on feedback from the first issue, we included mixed
methods research (Chapters 12 and 13) and added a concluding chapter.
Our primary aim is still to shed light on the true nature of these processes
that are all too often, and by necessity, stricken from work published in
scholarly journals and other outlets. The main reason for this missing infor-
mation is that researchers typically never have room to describe how they
carried out their analysis in scholarly journal articles, which are the usual
outlet for reporting findings. Peer-­reviewed journal articles typically have
word count or page number limits. Goldberg and Allen (2015) suggested
that the Methods section in a journal article might be about 7 pages long in
a 35-­page article, but this section consists of many topics in addition to data
analysis (e.g., data collection methods, interview questions, the researcher’s
positionality, and ethical issues). The proliferation of digital over print pub-
lication may provide an opportunity for change, although some argue that
consumers of digital information prefer shorter articles instead of longer
ones (Joaquin & Tan, 2021). Space limitations placed on researchers by aca-
demic publishers should not be mistaken for a lack of complexity in analysis
from which much can be learned. Indeed, there is a great deal to discuss
about one’s methods, and, in particular, the data analysis. One outcome of
this edited volume, we hope, is to draw attention to the limitations of space
placed on qualitative research in journal articles. Another outcome is to
encourage qualitative researchers to be explicitly clear with the space they do
have about what their processes actually entailed.
With these related, broader outcomes in mind, we aim with this edited
book to move beyond the common practice of merely providing citations for
terms such as “open coding” and “the constant comparative process” as suf-
ficient explanation of methodological processes. Instead, we argue that quali-
tative social science researchers must be granted the space to move beyond
the simple statement of “themes emerged from the data” to show readers how
they actively and consciously arrived at their themes (Richards & Morse,
2013), or what Richards (2020) refers to as “making data” (not to be confused
with “making up data,” which a graduate student in one of Áine’s classes once
thought it meant!). This edited volume offers behind-­the-­scenes narratives
from researchers that show the complexity and messiness that they dealt with
as they attempted to make sense of thousands of pages of transcribed inter-
view data, multiple data sources, diverse types of data (e.g., observations,
interviews, photos, and diary entries), and blind reviewer feedback. In doing
so, we respond to Sharp et al.’s (2014) call to expand and improve the terrain
of qualitative research—by “having more qualitative research available for
consumption in publication and expanding the types of paradigmatic fram-
ing, research designs, and qualitative analyses” (p. 154).
4 M. Elise Radina and Áine M. Humble

Intended Audience
This edited volume gives qualitative researchers the opportunity to go into
detail about their qualitative data analysis, something rarely seen in pub-
lished work. Like the earlier edition, it serves as a resource for both experi-
enced and novice qualitative researchers. This book may also be useful as a
supplemental textbook for graduate-­level research methods courses in a
variety of social science disciplines studying the range of human experiences
across the life span (e.g., family science, sociology, gerontology, psychology,
nursing, anthropology, public health, and social work).
This edited volume builds on what students may be learning in research
methods courses and moves their understanding to a new level. In our experi-
ence in teaching qualitative methods, students are often desperate for examples
and detailed recipes for how to conduct qualitative research (Radina & Sharp,
2019). Readers can engage with the authors through their accounts of their
processes and reflections on their craft. This occurs through the narratives pre-
sented in each carefully crafted chapter along with the supplemental materials
(e.g., examples of handwritten coding) provided by the authors and located on
the book’s companion website (eResource; see tab “Support Materials -
Ancillaries”). Instructors may want to assign some or all of the original pieces
of work for which the chapter authors are describing their data analysis. For
example, students could be provided with an example of Rude et al.’s (2021)
article, “Rights-­in-­between: Resident perceptions of and accessibility to rights
within restricted housing units,” that appeared in Law & Society Review for
which the data analysis process is described in Chapter 9. Additionally, this
book may provide advanced scholars with ideas for mentoring their students
in qualitative data analysis. It may also encourage more established scholars to
venture out of their tried-­and-­true methodologies to try new ones.
Indeed, as a result of our first edition, both of us were inspired to try arts-­
based research for the first time. Influenced by chapters by Weaver-­Hightower
(2019) and Miller (2019), and a “dialogue” chapter about arts-­based meth-
ods (Healey et al., 2019), Elise wrote the text for a children’s book based on
her qualitative study of parents raising children with the rare congenital
condition known as lymphedema (Radina, 2020). Áine was also inspired to
try her hand at poetic inquiry (Humble, 2021) as a result of engaging with
these same chapters. It is exciting to see how we can expand our repertoires
at any stage of our careers!

Book Structure Overview

For the first edition, we read widely and searched through databases to find
authors of journal articles to invite to contribute a chapter. For this expanded
edition, we created a call for chapter proposals, resulting in several dozen
 Introduction 5

submissions. We chose our contributors because they represented a range of

disciplines that focus on the human experience, and we anticipated that they
would have the ability to share the story of their analyses based on how they
talked about their work and what visuals they could provide to really show
readers their processes. Although some authors were familiar to us, most
were not, and consequently, we again had the pleasure of meeting authors
for the first time during this process. We especially enjoyed our process of
discovering and working with those authors who were new to us. As men-
tioned earlier, we also set out to expand the diversity of methodological
approaches with this edition. We used this diversity to help us organize the
order in which the chapters are presented, from least positivistic to most
positivistic epistemological perspective. That is, we begin with Underhill’s
chapter (Chapter 1) about autoethnography and end with two mixed meth-
ods studies—the first being qualitatively driven (Chapter 12: Shannon-­
Baker, Fisher, and Greer) and the second one being quantitatively driven
(Chapter 13: Thrower, Dorsch, Knight, and Harwood).
Overall, we feel that we were successful in meeting our multifaceted diver-
sity goals. We note, however, that it was impossible to include every method-
ology, data analysis technique, or discipline. Table 0.1 shows the order of the
chapters, the variety of publications on which each chapter is based, and the
academic discipline of the first author.

TABLE 0.1 
Summary of Chapters, Primary Study Publication Sources, and First
Authors’ Academic Disciplines

Chapter/Author Publication Source Academic Discipline

Chapter 1: Underhill Journal of Human Sexuality Family Science

(2021)
Chapter 2: Ferdinands Social Science & Medicine (2021) Health Promotion
and Raine
Chapter 3: Schadler Sexualities (2021) Sociology
Chapter 4: Hackett and Death Studies (2020) Gerontology
de Medeiros
Chapter 5: Erasga, Community, Work & Family Sociology
Cleofas, Andrada-­Poa, (2022)
and Jabal
Chapter 6: Boone Journal of Child & Family Studies Human
Blanchard, Coard, (2019) Development and
and Mereoiu Family Science
Chapter 7: Family, Systems, and Health Human
Edwards-­Bianchi (2021) Development and
Family Science
(Continued)
6 M. Elise Radina and Áine M. Humble

TABLE 0.1 (Continued)

Chapter/Author Publication Source Academic Discipline

Chapter 8: Fort, Santos, Journal of Health Care for the Public Health
de los Angeles Poor and Underserved (2020)
Villaverde, and Moore
Chapter 9: Rudes, Law & Society Review (2020) Criminology
Magnuson, and Ingel
Chapter 10: Vogl, Book chapter in New Parents in Sociology
Schmidt, and Zartler Europe: Couples in between
Norms and Work-­Care (2019)
Chapter 11: Grummert International Review for the Education
and Haslerig Sociology of Sport (2020)
Chapter 12: Shannon-­ Two journal articles: Special Education
Baker, Fisher, and Education Research, Policy &
Greer Practice (2021) and The Journal
of Special Education (2022)
Chapter 13: Thrower, Frontiers in Psychology (2019) Exercise and Health
Dorsch, Knight, and Science
Harwood
Note: For the full references, see the “Publications discussed in chapters” file at eResource.

Unique Contributions
Whereas other books focus on qualitative research and detail various
analysis techniques, this edited volume allows readers to be seemingly “in
the room” with the researchers and getting their own hands vicariously
dirty with the data. As Cuthbert et al. (2012) and DeLyser (2008) argue,
the teaching and learning of qualitative research methods require the use
of hands-­on activities almost more so than more traditional learning that
relies on the use of targeted readings about techniques. Similarly, Breuer
and Schreier (2007) suggest that learning how to conduct qualitative
research should be very much viewed as “learning a ‘craft’ … [that] ulti-
mately requires students to cooperate in carrying out related activities”
(p. 5). This book is offered from that perspective. Although the actual
apprenticeship of learning a craft at the foot of a master is beyond the
scope or ability of any book, this edited volume provides an intimate
approximation of this experience for the reader.
The narratives we have collected and curated here are further enhanced
by the supplemental materials available on the book’s companion website,
referred to in this book as the eResource. At the eResource, readers will see
examples of tables, figures, and documents that are clearly referred to in the
chapters. These extra files are included as examples with the aim of helping
readers visualize how the researchers made sense of their data, organized or
 Introduction 7

synthesized their understandings, and presented their data in visual ways.

Our hope is that these examples will be useful to researchers in their own
work and provide them with some possible ways they may want to approach
their own analyses.

Conclusion
We thank our contributors for their vulnerability in sharing their stories and
thoughts with us, and we also thank them for their patience with us in
responding to our multiple requests for “more detail” and “specific exam-
ples” (insert smiley face here!). Our goal has been to demystify the research
process by providing these researchers’ narratives—to pull back the curtains
on how “themes emerge.” We hope we have achieved that.

References
Breuer, F., & Schreier, M. (2007). Issues in learning about and teaching qualitative
research methods and methodology in the social sciences. Forum Qualitative
Sozialforschung/Forum: Qualitative Social Research, 8(1). [Link]
[Link]/[Link]/fqs/article/view/216/477
Cuthbert, D., Arunachalam, D., & Licina, D. (2012). ‘It feels more important than
other classes I have done’: An ‘authentic’ undergraduate research experience in
sociology. Studies in Higher Education, 37(2), 129–142. [Link]
075079.2010.538473
DeLyser, D. (2008). Teaching qualitative research. Journal of Geography in Higher
Education, 32(2), 233–244. [Link]
Dickie, V. A. (2003). Data analysis in qualitative research: A plea for sharing the
magic and the effort. The American Journal of Occupational Therapy, 57, 49–56.
[Link]
Healey, G. K., Miller, E., & Weaver-­Hightower, M. B. (2019). Dialogue about arts-­
based research. In A. M. Humble & M. E. Radina (Eds.), How qualitative data
analysis happens: Moving beyond “themes emerged” (pp. 209–212). Taylor &
Francis. [Link]
Humble, A. M. (2021, June 21–24). “Just two people expressing their love for each
other”: Creating research poetry. In K. S. Barss (Chair), Queering methods. The
Qualitative Analysis Conference [Online].
Joaquin, J. J., & Tan, R. R. (2021). The lost art of short communications in aca-
demia. Scientometrics, 126, 9633–9637. [Link]
­04192-­7
Goldberg, A. E., & Allen, K. R. (2015). Communicating qualitative research: Some
practical guideposts for scholars. Journal of Marriage and Family, 77(1), 3–22.
[Link]
Miller, E. (2019). Creating research poetry: A nursing home example. In A. M.
Humble & M. E. Radina (Eds.), How qualitative data analysis happens: Moving
beyond “themes emerged” (pp. 18–33). Taylor & Francis. [Link]
10.4324/9781315171647-­2
8 M. Elise Radina and Áine M. Humble

Radina, M. E. (2020). Ethnographic creative non-­fiction: The creation and evolution

of Lily’s Lymphedema. In P. Hackett & C. Hayre (Eds.), Handbook of ethnogra-
phy in healthcare research (pp. 438–451). Routledge.
Radina, M. E., & Humble, A. M. (2019). Introduction: Real stories of how this vol-
ume happened. In A. M. Humble & M. E. Radina (Eds.), How qualitative data
analysis happens: Moving beyond “themes emerged” (pp. xix–xxix). Taylor &
Francis.
Radina, M. E., & Sharp, E. A. (2019). Being lost and becoming reoriented with the
massive landscape: Students and instructors’ reflections on family science qualita-
tive research methods courses. Family Science Review, 23(1), 54–77. [Link]
[Link]/?s=radina
Richards, L. (2020). Handling qualitative data: A practical guide (2nd ed.). Sage.
Richards, L., & Morse, J. M. (2013). Readme first for a user’s guide to qualitative
methods (3rd ed.). Sage.
Rudes, D. S., Magnuson, S., Ingel, S. N., & Hartwell, T. N. (2021). Rights-­in-­between:
Resident perceptions of and accessibility to rights within restricted housing units.
Law & Society Review, 55(2), 296–319. [Link]
Saldaña, J. (2014). Blue-­collar qualitative research: A rant. Qualitative Inquiry, 20(8),
976–980. [Link]
Sharp, E. A., Zvonkovic, A., Humble, A. M., & Radina, E. (2014). Cultivating the
family studies terrain: A synthesis of qualitative conceptual articles. Journal of
Family Theory & Review, 6(2), 139–168. [Link]
Weaver-­Hightower, M. B. (2019). Analyzing self and other in autoethnography:
Telling secrets about one’s stillborn child. In A. M. Humble & M. E. Radina
(Eds.), How qualitative data analysis happens: Moving beyond “themes emerged”
(pp. 3–17). Taylor & Francis. [Link]
1
HOW AUTOETHNOGRAPHY BEGINS
AND NEVER ENDS
A Tracing of the “Self” in Relation to #MeToo
and Higher Education

Angela Underhill

There are moments from my life that return to me when I least expect it.
Some of these moments are, at the time, completely common or mundane.
Yet, they return to me, perhaps through a familiar sight or sound, and I may
pause to reminisce before moving on. Other memories, though, are stickier.
Sticky memories hold me in place. Sometimes, I am brought back to a time
and place where the intensity from the past is hot enough to burn my face in
the present. My past self’s recognition that this moment matters can be strong
enough to freeze my body, allowing my mind to work overtime to figure out
why it mattered, and why it still matters. And, once I determine the why, I
can consider the what next? Tracing these moments can become scientific
endeavors with the power to open possibilities far beyond the self. The ques-
tion then is, how? One of the ways I found to deeply understand these sticky
moments came in the form of autoethnography and critical theory.
In this chapter, I revisit the active, iterative autoethnographic process I
used to examine the forces (described in the New Materialism section) com-
ing together to create moments of shame and otherwise in pedagogical
spaces in my “#MeToo: (Re)Shaping sexual research pedagogies through a
new material autoethnographic analysis” article (Underhill, 2021; hence-
forth, “#MeToo article”). I recount how I came to write about this topic and
briefly highlight the theoretical frameworks that deepened my analysis. I
then trace my data collection and analysis process, encapsulating (a) my
memories; (b) empirical data; (c) artifacts that became meaningful through-
out the process; (d) feedback from supervisors, mentors, conference goers,
and reviewers; and (e) reading and writing about theory alongside my expe-
riences. Throughout, I emphasize the ethical and relational considerations

DOI: 10.4324/9781003254041-2
10 Angela Underhill

I made. I conclude with reflections on what I originally viewed as a caution-

ary note from my #MeToo article, but which I now see as a pillar of auto-
ethnography and an invitation for future possibilities: Autoethnographic
analysis, like all rigorous science, never has a clear end. As I stated in my
article, “By the time this is read, my words and reflections may carry new
meanings” (Underhill, 2021, p. 173).

Where to Start?
In my #MeToo article, I took up the challenge of unpacking sticky moments
that seemed to be reoccurring in different ways throughout my life as a sex-
ual health and rights researcher and educator. (If I am honest with myself,
as autoethnographers need to be [Weaver-­Hightower, 2019], I began taking
up this challenge when I chose to pursue a career in sexual health and rights.)
As the hashtag suggests, the moments centered my confrontations with the
topic of sexual assault at different levels of my academic career: First as an
undergraduate student, then as a graduate student in a course I co-­developed
with my peers and instructor (henceforth, graduate experience), and, finally,
as a sessional instructor (a position where I may have appeared as powerful
to my students, when in reality I had little actual power in the academy;
henceforth, instructor). In each classroom, I experienced different, intense
emotional and physiological responses that transformed my teaching and
learning process: nausea… a racing heart… and shame so powerful I blacked
out. Although I cannot remember every single detail of each event (as I note
in sections of my #MeToo paper), I do vividly recall the feeling of this mat-
ters. I also recall wondering what to do with that mattering.
For years, I replayed my undergraduate student experience in my head. I
would go over the details and wonder what to do with the shame that plagued
me—shame that grew from experiences much earlier than the ones I had in
the classroom. I wondered if I could really be a sexual health educator when
I could not talk about my own experiences. Later, I enrolled in a graduate-­
level critical theory course and was once again frozen when #MeToo went
viral on social media and took over our class. Then, I taught an undergraduate-­
level course that included a lecture on the #MeToo movement. I was
reminded once again of the earlier classes that made me freeze, burn, and
feel the past in the present. From those past experiences (and running them
back in my mind over and over), I had learned what was useful and/or chal-
lenging in the classroom for me, and I used these lessons to inform my teach-
ing. Exploring alternate ways of knowing, feeling, and being in a pedagogical
space made me feel a bit more comfortable with difficult topics, but it did not
feel like enough. Then, I encountered autoethnography and new material-
ism. In my #MeToo article, I shared how this theory and methodology
 How Autoethnography Begins and Never Ends 11

shaped my understandings of what is needed to facilitate productive peda-

gogical spaces focusing on sexuality. In this chapter, I share my story of how
to do autoethnography.

Seeing Academia Differently

To do autoethnography, you first need to be exposed to the richness and power
of such a method. Autoethnography tends to be reserved as supplemental read-
ing in many university courses, rather than the focus (apart from specialized
methods classes). It differs from “traditional” science. It is overlooked.
In 2019, I informally audited another graduate-­level critical theory course
while delving into new materialism for my doctoral research. Through this
course, I was introduced to McPhail et al.’s (2017) autoethnographic analy-
sis. Using critical fat studies as a basis for their article, McPhail and col-
leagues offered a powerful reflexive argument for a deliberate acknowledgment
of how an educator’s embodiment (their physical body, bodily history, how
their bodies are read, and more) comes to shape—and be shaped—in peda-
gogical spaces. I was struck by McPhail et al.’s honesty, emotion, and insight.
Such work is rare in academia.
McPhail et al. (2017) resisted the standard ways of performing academia
and made themselves the site of the research, reflecting the reality that their
physical bodies were already the site of scrutiny in the classroom due to
(problematic) societal discourses, their academic disciplines (dietetics), their
gender, and their body sizes. They rendered the invisible judgment they felt
into an explicit acknowledgment of their vulnerability and productive
meaning-­making for themselves and their students—and the readers, of
course. It was right around this time that I was also confronted with a timely
social media “memory” of my #MeToo post (see Figure 1.1) and reminded
of how my own bodily history took up so much invisible space in the class-
rooms I had occupied. It was this combination of theory, witnessing vulner-
ability in academia, and being reminded of my own that I finally felt like I
had something to do with that feeling of this matters.

[SEE FIGURE 1.1 at e­Resource: The Viral Hashtag].

Theoretical Shaping
Theoretical shaping, as opposed to framing, reflects the idea that theory and
knowledge is, or should be, constantly evolving alongside analyses. Shaping
is used to denote that theory is an active force, changing with new evidence.
Although researchers should use theory to guide their work, a rigid frame
(as opposed to an active force) can limit the analysis and therefore restrict
12 Angela Underhill

the possibilities of transformation. In other words, if we keep looking for

answers to social world issues in the same places, with the same tools, we will
keep finding the same answers.1

Autoethnography
In my pre-­ autoethnography processing, I was not finding meaningful
answers. I was confined within the rigid bounds of the academy and the
prescriptive norms within. Autoethnography was a site for new answers. The
site of autoethnography, though—the self—makes it risky (Spry, 2001). The
very process challenges much of what is valued in academic (post-­positivist/
objective) spaces. The autoethnographic process makes visible what we often
purposefully make invisible in research. In research, we are trained to keep
our feeling knowledge to ourselves. We are told that our experiences make us
biased, less credible. We are told that not knowing or having experienced
something reduces or eradicates bias, rather than simply shifts the source.
We are trained to “research down,” to pick topics and populations that need
“saving,” and to focus on “marginalized” experiences to better our world.
Yet, the result is often that we are the ones marginalizing Others, and we
perpetuate stigmatizing and harmful norms through our distance from the
lived experience (Fraser & Jarldorn, 2015; Tuck, 2009). Autoethnography
can challenge all these accepted norms, which in turn can lead to fear of
reprisal, critique, and rejection, and the actual experience of those actions
(Davies, 2002; Spry, 2001). Imagine becoming vulnerable for the world and
having academic reviewers reject you with the rationale that your analysis—
your story—is not valid or reliable. Risky.
Autoethnography is the result (written, performed, or otherwise) of pro-
cessing memories to understand broader social phenomena. Autoethnography
is not simply reflection, critical reflection, or reflexivity. From an epistemic
point of view, some scholars argue that reflection (which suggests mirroring
reality), critical reflection (which demands consideration of one’s assump-
tions and biases), and reflexivity (the process of reflecting and identifying
ways to act anew) all suggest that someone can find truth in the world by
simply stepping back and examining the situation (Bozalek & Zembylas,
2017). The process and outcome of these practices can therefore reduce
knowledge in a simplistic way and imply an impossible objectivity and linear
path for moving forward. In comparison, autoethnography troubles the idea
of a singular truth. An underlying assumption is that there is no single or
clear answer. Instead, autoethnography uses one site of experience to open
possibilities for difference.
Another key difference between autoethnography and reflection is that
some scholars argue reflection leads to partial accounts that center on the
 How Autoethnography Begins and Never Ends 13

knower (Bozalek & Zembylas, 2017; Davies, 2002). Autoethnography differs

from these practices by interrogating the self in relation to O/others, culture,
social norms, and more (Davies, 2002; Spry, 2001), while maintaining that
there are multiple truths depending on one’s situatedness, which affects their
way of knowing. These requirements trouble the very idea of scientific
objectivity. These requirements also align with the theoretical framework I
used to inform my autoethnographic process: new materialism.

New Materialism
New materialism was essential in guiding my thinking beyond my self. In
short, new materialism is a term used to describe a collection of scholarship
underpinned by the assumption that nature and culture (like other binary
understandings, such as reason/emotion, discourse/material, and male/
female) are neither mutually exclusive nor hierarchical categories (Fox &
Alldred, 2017). Instead, they actively shape one another and matter in differ-
ent ways. According to new materialist thinkers, everything in our world—
from meaning-­ making to bodily histories, to physical objects and
beyond—can be considered matter, and all of this matter is interrelated and
in a constant state of becoming (Barad, 2003). Nothing is fixed or final (Roy
& Subramaniam, 2016). Following this line of thinking, the term “forces,”
an active word, is used to describe the people, practices, and objects (and
more) that come together to create any observed outcome. For this reason,
new materialist perspectives demand that everything measured is done with
explicit consideration of the contextual forces coming together to produce
the observed outcome. We cannot simply state, this is the way it is, because
each situation changes depending on the forces coming together. We can
state, however, that in these circumstances, this thing happened, and here is
the relational network of forces we observed as contributing. This network
can be referred to as an assemblage (Deleuze & Guattari, 1980/1987). In
many ways, assemblages that include (but are not limited to) one’s own
memories represent the autoethnographic requirement to consider the self
in relation to broader social life, including relationships, social norms, and
beyond.
Because all matter is constantly in flux and interacting with other matter,
new materialist thinkers actively consider what forces (human and non, in
the moment and in a different time) are coming together to produce differ-
ence (Barad, 2003; Bozalek & Zembylas, 2017; Feely, 2020; Grosz, 1994). It
is not about simply reflecting on the happenings or trying to determine what
someone’s true intentions were, but rather considering how we (and those
involved) know, and how this materializes into difference. From this perspec-
tive, “difference” is seen affirmatively, acknowledging value in multiple
14 Angela Underhill

perspectives (in difference) and value in tracing the forces to see what else
could be contributing to the differences observed. One aim of those using
new materialism is to use their tracings to imagine more positive or produc-
tive effects that could have been or could be if the forces were or are changed.
In short, new materialist thinkers practice social justice (Barad, 2003).
When I began my autoethnographic process, I could readily identify
many meaningful forces contributing to my classroom experiences. My rela-
tion to the topic being discussed marked me as an insider. I could not engage
with the topics without remembering my own experiences. My insider
knowledge provoked powerful, uncontrollable physiological responses such
as a racing heart. Upon further reflection, there were also forces that were
more covert yet still vital to the effects. I started thinking about how the
room itself constrained me as a student in my undergraduate class. I was
closely surrounded by strangers, and I worried about drawing attention to
myself due to my flushed skin. In each classroom, I also desired to do well.
This desire limited how I could (dis)engage with the material. I could not
simply leave my classes; doing so could negatively impact my grade as an
undergraduate student, my reputation among my peers and advisor as a
graduate student, and my students’ learning when I was an instructor.
Unpacking that further, now, I also wonder what leaving would have sig-
naled to my students if they too had experiences that would position them
to say, “me too.” I began to think more purposefully about the other bodies
in each experience and how they too affected—and likely were affected by—
what occurred in the space. The purpose of these educational spaces was to
teach and learn. Yet, in those moments, there were forces coming together
that prevented me from learning and later teaching in the traditional, “ratio-
nal” academic way. How could my felt knowledge in those moments offer
productive learning assemblages for others?

Writing
With all of this in mind, I began to write. At the time, producing the written
account was low risk: I was required to write a final paper for my independent
study that would be seen by only two people (my instructor and a second
reader). For me, this reduced some of the fear I felt around being honest and
vulnerable as autoethnography requires (Spry, 2001; Weaver-­Hightower, 2019).
Little did I know that this would turn into an accepted poster abstract for a
conference that tends to accept “objective” research. In contrast to the normal,
post-­positivist conclusions, I constructed the results section of my poster (pre-
sented in Figure 1.2) as a series of guiding questions to prompt the conference
attendees to think through their own classrooms. My poster attracted more
attention than past oral abstracts I had presented. Following the conference,
 How Autoethnography Begins and Never Ends 15

one of the organizers reached out and asked me to submit a full paper for
review. In truth, I did not anticipate that the paper would be read(!), let alone by
well-­established researchers who would then invite me to submit a chapter for
their upcoming edited book about qualitative data analysis. It occurs to me that
the desire for felt knowledge in academic spaces may be stronger than I expected.

[SEE FIGURE 1.2 at eResource: An Early Depiction of an Ongoing Analysis].

When the “Research” Is the Easy Part

I started my analysis with what I thought would be easy, yet still important:
setting the stage with the political and social climate. In any new material
analysis, scholars need to explore the interaction of the contributing forces,
including the political and social discourses that inform one’s experience
(and vice versa). I drew from past lectures I developed that focused on inter-
disciplinary scholarship and advocacy in relation to the #MeToo movement.
Some of the articles focused on the (problematic) social discourses that
dominate our society and revolve around sexual assault and victim blaming;
others focused on the statistics, reporting the incidences and general rates of
sexual assault. The advocacy materials offered powerful videos, counter-­
narratives, and comics challenging the dominant discourses.
As I reviewed the lectures and associated sources, I started to consider
how these discourses and numbers acted in each classroom. Both new mate-
rialism and autoethnography require careful examinations of the researcher
(or self) in relation to other forces. Aligned with these perspectives, I consid-
ered: How did I relate to these research findings and advocacy movements then,
and now? I knew in my graduate and instructor experiences that my position-
ality as a young, cis woman aligned with the statistics of one group often
targeted in sexual assaults (Sex Information and Education Council of
Canada [SIECCAN], 2015). This provoked another question: How did oth-
ers in the room relate to the research findings?
My power and privilege mattered. I was (and am) educated, able-­bodied,
white, and generally read as heterosexual. Given the generally high preva-
lence of sexual assault in our society, I assumed (as an instructor) there
would be students in the room who shared that experience. A similarity.
However, I recalled never conceiving of this similarity when I was an under-
graduate student. I also recognized the varying social locations my students
held and how these locations potentially complicated their experiences in
ways that differed from my own. For example, like many others who have
experienced sexual assault, I never reported my experiences because it did
not feel important to me (SIECCAN, 2015). For those with different social
locations, their ability and decision to report (or not) may have been
16 Angela Underhill

compounded by the layered aspects of institutional violence, such as police

violence instigated by racism (Linder, 2019), from which my white privilege
protected me. I also considered how my responses as an undergraduate stu-
dent (freezing, turning flush, and feeling alone) differed from my responses
when I was the instructor who chose the content, who witnessed and affirmed
the emotional responses of others, and who demonstrated vulnerability to
my students. From this, I wondered, how did my positions in relation to others
occupying these spaces facilitate or constrain learning? These moving inqui-
ries followed the new materialist assumption that all matters affecting a situ-
ation are in constant flux.

Disclosures
Eventually, I had to fully turn to the experiences of being in different class-
rooms, focusing on the same topics where I differed in position, in expertise,
and in power. Each experience shared the central theme of discussing sexual
assault in higher education, yet each was markedly different in how I felt and
engaged in the space and how others engaged in the space (I had no way of
knowing how they felt). Some of the forces contributing to those moments
were the same across time and place, including the binding institutional
norms that result in large undergraduate classes, for instance, as well as the
ideal standard of “rationality” and “objectivity” in academic spaces. Some
of the forces were radically different, such as my self. With each additional
year of interrelating with new knowledge, theory, people, and more, I was
different. I was different when I began to analyze and different when the
analysis moved from a course paper to an abstract, to the final publication.
I am different now that I am looking back.
The first time I wrote about each of the experiences, I rushed. I was honest,
raw, and my stomach turned at times. Still, at a conscious level, I knew I was
holding back. On my second review, I implored Chadwick’s (2017) embodied
methodology and asked myself, what did it feel like? I quickly typed words like
“hot,” “nausea,” and “exhausted.” I thought about how my wearable fitness
tracker vibrated against my wrist as I lectured, letting me know my cardio work-
out was successful; I added “racing heart.” When I reread the scenarios, the fear
of vulnerability temporarily swept over me. I considered removing moments
that made me look weak, unprepared, or emotional. This was supposed to be
an academic paper. Did the reader really need to know that I felt overwhelmed
during my graduate studies or that I was balancing a myriad of extracurricular
obligations? Did they need to know my voice cracked as I lectured?
Spry (2001) asserts that autoethnography “is an embodied practice; it is
an intensely sensuous way of knowing” (p. 180). My body was acutely aware
of the history I lived with in those moments; of the potential scrutiny from
 How Autoethnography Begins and Never Ends 17

my peers, and later students regarding my reality; and how dangerous those
moments could be given the privileging of the mind over the body in higher
education. As uncomfortable as it was, I reminded myself that my discom-
fort in those moments is what mattered. It was, after all, what prompted me
to write. If my stomach turned while I read the experiences, perhaps readers
would feel something too.

Complicating Vulnerability Further

Once I had an initial draft of each classroom situation, my next step—
informed by new materialism and autoethnography—was to map out the
forces contributing to those moments of discomfort beyond my bodily his-
tory and observations. This process relies on situating oneself in the social
world and identifying discursive and material forces that may be coming
together to produce the observed effect(s). Therefore, unpacking the forces
in each scenario became part of my analytic process. To deepen the descrip-
tions, I continuously relied on my theoretical grounding and asked myself:
What contributed to this moment?
To organize my thoughts, I made a table, in a Word document, based on the
Material/Discursive Continuum that Feely (2020) constructed in his assemblage
analysis of sexual surveillance in long-­term care homes (see Table 1.1). Some of
the headings I originally included aligned with Feely’s (i.e., architecture, eco-
nomic capital, and site-­specific rules and policies). Others I selected based on
what made me pause as I read and revised my scenario depictions over and over
(i.e., physical objects, such as desks, institutional discourses, societal discourses,
and the instructor/student at individual and relational levels). To thicken the
descriptions and really begin analyzing the forces, I brainstormed questions
under each heading to prompt me to think further about each potential force.
For instance, under the architecture heading, I wondered, “Are there gender
neutral/accessible washrooms?” (Underhill, 2021, p. 173). Under institutional
discourses, I asked, “What formal roles and relative hierarchies are there? …
Who is responsible for what?” (Underhill, 2021, p. 173). Each question led to
memories. I documented the endless forces that I started to recall in my table,
and I added these forces into the descriptions of what happened. By reading
Feely’s analysis, I began to understand the potential vitality of even the smallest
details, tangible and otherwise.

[SEE TABLE 1.1 at eResource: Prompts for Analysis].

Every detail began to take on a new meaning to me. I reviewed my table and
started to wonder, how do these forces interact? I began with obvious interac-
tions: How did each of the unique forces act in relation to each scenario? I
18 Angela Underhill

remembered, for instance, the creaky seating in the undergraduate classroom

and the silent chairs on carpeted floors in the graduate room. This sparked
more: I remembered how disruptive it could be to my peers if I chose to leave
in the middle of a lecture when I was an undergraduate student, not only
because of the close, creaky seating but because it was winter and I would
have to pick up all of my outdoor clothing as well. It seemed significant,
then, to consider the weather and time of year. Similarly, despite the quiet
seating in the graduate classroom, there were fewer than a dozen people in
the room. Leaving discreetly was impossible. I also recalled being exhausted
in the graduate classroom. Was that just because it was the end of the day,
or was it because of everything else I was juggling? Likely, it was both. I went
through the details and added, removed, and edited the descriptions to
ensure each sentence included a force that, if interacting differently with the
other forces, could lead to difference. I wanted each sentence to provoke the
reader to think about what could matter to the experiences of others.

How to Understand the Other(s)

The next step in my process was to begin the more formal analysis. I returned
to some of the original questions I had when I began: Who is protected by
these forces? Who is put at risk? Despite what the name may imply to those
unfamiliar with the methodological practices, autoethnography is far from
an individual reflection of what happened. In my process, it was impossible
to write solely from my experience without considering my Others: those
who were also researchers, educators, students, those who could also say
#MeToo, and all the overlapping identities of these categories and beyond.
How might their experiences interact with my own in these spaces?
I went back to my descriptions and the list of forces I had begun to accu-
mulate in my original table. I regrouped the forces and made new headings
as needed. For instance, physical objects became physical/environmental to
allow for inclusion of outdoor clothing and weather (see Table 1.2).
Meticulously, I went through each one and jotted notes in response to who
was protected and who was at risk. Aligned with new materialism, I then
asked: What force(s) can be shifted to allow for more flourishing? In imagin-
ing a socially just classroom, this was what mattered most.

[SEE TABLE 1.2 at eResource: Original, Condensed Summary of Forces].

One of the things I have come to know about autoethnography is that read-
ing and writing are perhaps the most important parts of the process (next to
feeling). Augustine (2014) similarly shares that their new materialist analysis
consisted of the “intermingling of data and theory after focused reading and
 How Autoethnography Begins and Never Ends 19

copious amounts of writing” (p. 752). Each theorist and researcher I read
from became a force affecting my understanding of my own experiences. I
drew from my social justice training; critical disability, queer, fat, and race
studies; articles I read in my independent study; and materials I engaged
with through each of the courses I reflected on. I went back to the disem-
bodied statistics that encapsulate the silent stories of many within neat,
organized categories. I thought about how all these hidden people might
exist in any room with me, with their own varying levels of power and their
own histories, and I wondered what forces might create a productive learn-
ing effect for them.

Confession
I feel like this is the time for me to confess that the process was far from lin-
ear, reproducible, or easily traceable.
Like Weaver-­Hightower (2019) notes, much of the autoethnographic pro-
cess happens in one’s head. As I already shared, I had begun consciously
analyzing my experiences as an undergraduate student years before I ever
considered writing about them. Parts of my experiences had been replayed
in my head so often I am confident, years later, in each detail. Other parts,
though, were further away. The prompting questions I created helped me
recall some of these finer details (like the building’s layout that could make
the space safer for some than others). Perhaps this is a limitation of autoeth-
nography—though, of course, it is a limitation of all research that relies on
the past. Whether it is personal experience, retrospective surveys, or archival
data, our interpretations are influenced by time, culture, and the practices
that they are produced within. Perhaps it is a limitation, then, or perhaps
what sticks tells us something valuable. As I share in my #MeToo article,
“The night that stands out to me most was a night where I retained little to
no knowledge at all” (Underhill, 2021, p. 174). The content of the night was
not what mattered. Reproducing the content from the night would not have
generated the new knowledge and insights I gained from the autoethno-
graphic process. Part of what mattered was the feeling it left me with and all
the pieces that I could still recall.

Struggling with the Reconceptualization of Data

Still, part of me was set on a singular truth. I wanted to have evidence and
documentation for my claims, convinced that this would somehow reduce
the vulnerability I was experiencing. Fortunately, autoethnography and new
materialism both allow for multiple forms of data. I began searching for
artifacts that I had not even realized I was collecting. I typed each course
20 Angela Underhill

code into my email and found long-­lost letters from peers and students. I
filtered these emails to focus on #MeToo, or specific dates that aligned with
the syllabi I had saved from each class. I found notes of gratitude and regrets
from students who took my class when I lectured about #MeToo. One stu-
dent notified me that they posted their notes from the #MeToo lecture onto
the course discussion board to support their peers who could not attend. Ah,
I thought. Would someone reach out to their peers for the material if they
elected to miss a difficult class, or would they fear the peer would wonder why
they were not there like I feared when I was in their position? This question
highlighted potential forces coming together: a student’s embodied experi-
ence; their attendance or ability to catch up, or not; how societal discourses
and past interactions might influence one’s actions despite the academic
stakes. Once again, I was grounded in the autoethnographic principles of
acknowledging the possibilities of multiple truths. I went back to thicken my
description of lecturing.
I returned to the syllabus of my graduate-­level course and fondly reviewed
the care that was put into co-­developing the space, recognizing that even the
most tender of pedagogical practices was not powerful enough to remove
negative emotions (now, I wonder if that should ever be the goal). Rereading
some of the assigned articles for the course brought me back to those late-­
night theoretical discussions and conjured up some of the theoretical con-
nections we, as a class, had discussed the night #MeToo took over. Indeed,
my first, unpublished title (#MeToo and other things we shouldn’t have to
teach about) seemed to reflect the frustration that many of us felt in needing
a #MeToo movement despite the widespread (and ignored) knowledge that
sexual assault was a problem. I reread Britzman’s (1998) paper again, one of
the required readings from my graduate course, and felt ironically connected
to others by her assertion that embodied learning can be lonely, yet transfor-
mative. I found myself lost in Knoll’s (2009) article detailing a feminist dis-
ability studies pedagogy and wondered how to facilitate comfort for those
encountering difficult, embodied knowledge in higher education. These
articles once again shifted my analysis, giving me material to weave into my
historical account and reassemble what was important about my experiences
and how they were situated in the broader social context of academia.

Including Others in the Process

In this tracing process, I recalled designing my sessional course and meeting
with mentors to discuss course design and potential safety measures. One
mentor suggested a disclaimer about the course content to give students
agency in deciding if they were in a place to take such a course; another sug-
gested offering optional test questions so students could opt out of some
 How Autoethnography Begins and Never Ends 21

content. A third suggested leaving out test questions about sexual assault
altogether. All suggestions prompted changes to the forces in pedagogical
spaces, and all required careful reflection on what could contribute to the
pedagogical assemblage. These conversations shaped my initial course struc-
ture, which in turn shaped each lecture and the way my students were con-
strained within the bounds of my course. Each conversation ultimately
influenced my autoethnography.
Once I started looking for forces, I spiraled into the possibilities. Like
Augustine (2014), I was starting to identify new associations and make con-
nections between theory, data, and Others. The process was rich and brought
on a new sense of vulnerability: How much more should I be doing as an
educator? What had I missed as an instructor? Who had I (un/intentionally)
privileged or constrained?
Later, I invited more people into the analysis process, with varying
embodiments and histories: my independent study evaluators, conference
abstract reviewers, conference attendees, and journal article reviewers. Given
my process (term paper, to abstract, to poster, and to manuscript), I wel-
comed an abundance of feedback. Opening the emails that contained the
reviewers’ feedback was one of the most difficult parts of the process. This is
not necessarily unique to autoethnography, but the personal nature of the
paper sparked a different kind of fear. Would my credibility be questioned
because of the personal nature of my work? Would reviewers consider this
science? Would they consider it valuable? Would I simply be rejected?
Fortunately, these concerns were moot. My first evaluators (for my term
paper) focused on the more traditional elements of a paper, reminding me, for
example, to add in a limitations section and asking for me to clarify some of
my analytic steps. The conference attendees raised interesting points for me to
tease out in the final manuscript. One person, for instance, questioned the
benefits of content warnings (one of several forces I suggested people consider
adding to their courses), citing researchers who found content warnings
increased anxiety for students more than the actual content. This prompted
me to think through where this anxiety might come from, and it allowed me to
further consider the interactions of forces in pedagogical spaces. To me, it sug-
gested that content warnings were not enough. Indeed, they are quite useless
if the other forces in the room are not conducive to student safety and flour-
ishing. I expanded two paragraphs about this in my final manuscript to address
other potential reviewer comments: “If we expect ‘bravery’, we also need to
facilitate a space that promotes such an ask before students are ever put in a
position to [perform bravery]” (Underhill, 2021, p. 177).
Some of the most meaningful changes I made to the analysis were in
response to reviewer comments. In the first version of my #MeToo paper
(intended for my independent study course), I compiled a table with reflective
22 Angela Underhill

questions for the reader to consider in constructing a pedagogical space.

I wanted readers to consider their own embodiments and what they could
change in their own classrooms, something I felt I could not tell them to do as
it would shift with every classroom assemblage. I named various forces that
could be considered at the top of the table and listed reflective questions. For
instance, under Physical Space, I asked, can students easily enter/exit the room?
One reviewer asked me to expand and outline the potential pedagogical shifts
more clearly. In reading this, it occurred to me that my original reflective ques-
tions would be meaningless to anyone who had not been exposed to the same
analysis, experiences, and knowledge as me. I adapted the table and explained
each prompting question alongside potential ways to mitigate forces that
could inhibit the teaching and learning process (see Figure 1.3). The final
table still allowed for flexibility, but it provided more direction for those who
may have needed it.

[SEE FIGURE 1.3 at eResource: Moving from Prompts to Analysis].

The value of others reading and responding to the paper served as an addi-
tional layer of analysis. Again, it occurs to me that autoethnography may be
a misnomer. It is not simply about the self, and it does not need to be done
in isolation.

Never Ending
I am the type to write and rewrite every sentence multiple times before mov-
ing on to the next. Sometimes, this means I only have one full, final draft, but
every sentence within that final draft has been agonized over hundreds of
times before I finally share it. With autoethnography, I found the sentence-­
by-­sentence review endless. Every new paper I read, every memory I had,
and every artifact I found offered a new opportunity to dig deeper. Eventually,
I was saved from the analytic possibilities by the very hard deadline for my
term paper. When it returned to me, I spiraled again. I was eventually saved
by the conference abstract deadline. My surprise at the abstract’s acceptance
to a national, empirically leaning conference forced me to spiral again, try-
ing to ensure the arguments were sound and supported. Later, the journal
deadline forced me to conclude the analysis once more. I could keep analyz-
ing to this day, and my interpretations would shift just as I have over time,
with exposure to new ideas, conversations, and circumstances.
To date, I have four distinct drafts of my #MeToo paper. My positional-
ity, knowledge, and experiences at each editing point shifted the analysis.
Having now reviewed the article many more times in preparing this chapter,
I find myself getting stuck at different parts of the autoethnographic account
 How Autoethnography Begins and Never Ends 23

that previously held less, or different, meaning. At its core, autoethnography

is about vulnerability. Yet, the risks of vulnerability for instructors vary
based on each person, class, and institution. Since publishing my #MeToo
paper, I have engaged with literature highlighting the perils of social justice
work in institutions that have not structurally changed (Ahmed, 2021;
Elwood et al., 2019). I now find myself pausing at my recommendations and
considering the potential risks to contracted (as opposed to tenured) instruc-
tors who offer lenience to institutional norms (such as absences and equity-­
based assignments). Are these individuals put at risk if they do so? What
made it more possible for me to engage in some of these risky practices, over
others, despite my own contract status? We have also now collectively lived
through years of a global pandemic. I have now taught multiple sexuality
courses online asynchronously. In this mode, it occurs to me that many of
my recommendations are irrelevant. An entirely different analysis would be
needed to center this learning environment. Autoethnography, like all sci-
ence, should not be thought of as final even after it is written down.

Conclusion
So how does one do autoethnography?
By now, I hope this question is met with a range of responses. This chap-
ter is far from a recipe that can or should be reproduced step-­by-­step.
Instead, it is a mix of ingredients that may or may not be used, paired, or
introduced in the same way. It is an ongoing resistance to the invulnerable,
impersonal researcher bias, and finding others who are doing the same. It is
an acknowledgment that feeling matters, and it is valuable to explore. It is a
search to understand and identify the many different, possible forces hold-
ing us in sticky moments. It is a reexamining of each force, looking at each
one anew, imagining them in relation to different contexts and bodies and
theories and discourses, and discovering how they change shape, color,
meaning, and fit. It is an invitation to engage, reflect, ask questions, and
imagine differently. It is a personal engagement with the broader world that
will never end. There might be a final paper, but it is always a draft. The
analysis continues, moving into the back of the mind, perhaps one day find-
ing itself at the start of a new page.
As I revisit my new materialist, autoethnographic journey of my #MeToo
paper, I am relieved. Parts of the paper still feel true, relevant, and impor-
tant. However, that is not the relieving aspect. I am relieved that, in reread-
ing the paper, I have identified new areas to explore, analyze, and reimagine.
For instance, in my #MeToo paper, I conceptualized my experiences as
shame assemblages (Underhill, 2021). Now, years later, I have encountered
new material, grown as an educator and scholar, and engaged with the
24 Angela Underhill

ever-­evolving social and pedagogical landscapes. I wonder if that shame

assemblage might be part of a bigger assemblage informing my own peda-
gogical values, one piece to a much more expansive network. I know, now,
that the new insights and potential avenues for exploration do not signal a
fundamental flaw in the analysis. Rather, it signals a response to, and inte-
gration of new, interdisciplinary data that includes (rather than ignores)
material that makes analysis uncomfortable, messy, and impartial. It signals
an ever-­evolving change in what matters.

KEY WORKS GUIDING MY DATA ANALYSIS

Chadwick, R. (2017). Embodied methodologies: Challenges, reflections and strategies.

Qualitative Research, 17(1), 54–74. [Link]
Chadwick’s text offers a guide for “embodied methodology.” Although
Chadwick focuses on narrative analysis with participants, the prompts and strate-
gies proved useful for deepening my own description of events, which in turn
affected what material I had available to analyze.
Feely, M. (2020). Assemblage analysis: An experimental new-­materialist method for
analysing narrative data. Qualitative Research, 20(2), 174–193. [Link]
10.1177/1468794119830641
New materialism is a growing field with many different approaches to data
analysis. Feely’s assemblage analysis offers one approach and worked as an
example of using new materialism as a methodology. This paper was founda-
tional for me because it offers possibilities for what material and discursive
forces can come together to create an assemblage. So often we think deeply
using one specific lens, but Feely offers examples to think from multiple per-
spectives and bring them together.
McPhail, D., Brady, J., & Gingras, J. (2017). Exposed social flesh: Toward an embodied fat
pedagogy. Fat Studies, 6(1), 17–37. [Link]
Without McPhail, Brady, and Gingras’s article, I do not know if I ever would have
considered autoethnography. Their paper offered the vulnerability, richness, and
insight I was craving to better understand how to use embodied and felt knowledge
in pedagogical spaces. The value of their paper also gave me motivation to continue
despite the fear I had throughout the process.

Note
1 Thank you, Dr. Carla Rice, for sharing this phrase to discuss the value of nontra-
ditional methodologies. It has transformed my research approach.

References
Ahmed, S. (2021). Complaint! Duke University Press.
Augustine, S. M. (2014). Living in a post-­coding world: Analysis as assemblage.
Qualitative Inquiry, 20(6), 747–753. [Link]
Barad, K. (2003). Posthumanist performativity: Toward an understanding of how
matter comes to matter. Signs: Journal of Women in Culture and Society, 28(3),
801–831. [Link]
 How Autoethnography Begins and Never Ends 25

Bozalek, V., & Zembylas, M. (2017). Diffraction or reflection? Sketching the contours of
two methodologies in educational research. International Journal of Qualitative
Studies in Education, 30(2), 111–127. [Link]
Britzman, D. (1998). Lost subjects, contested objects: Toward a psychoanalytic inquiry
of learning. SUNY Press.
Chadwick, R. (2017). Embodied methodologies: Challenges, reflections and strate-
gies. Qualitative Research, 17(1), 54–74. [Link]
Davies, C. A. (2002). Reflexive ethnography: A guide to researching selves and others
(1st ed.). Routledge.
Deleuze, G., & Guattari, F. (1980/1987). A thousand plateaus: Capitalism and schizo-
phrenia (B. Massumi, Trans.). University of Minnesota Press.
Elwood, J., Andreotti, V., & Stein, S. (2019). Towards braiding. Musagetes.
Feely, M. (2020). Assemblage analysis: An experimental new-­materialist method for
analysing narrative data. Qualitative Research, 20(2), 174–193. [Link]
10.1177/1468794119830641
Fox, N. J., & Alldred, P. (2017). Introduction. In N. J. Fox & P. Alldred (Eds.),
Sociology and the new materialism: Theory, research, action (pp. 3–13). Sage.
Fraser, H., & Jarldorn, M. (2015). Narrative research and resistance: A cautionary tale.
In L. A. Brown & S. Strega (Eds.), Research as resistance: Revisiting critical, Indigenous,
and anti-­oppressive approaches (2nd ed., pp. 153–175). Canadian Scholars’ Press.
Grosz, E. (1994). A thousand tiny sexes: Feminism and rhizomatics. In C. V. Boundas
& D. Olkowski (Eds.), Gilles Deleuze and the theater of philosophy (pp. 187–210).
Routledge.
Knoll, K. R. (2009). Feminist disability studies pedagogy. Feminist Teacher, 19(2),
122–133. [Link]
Linder, C. (2019). Power-­conscious and intersectional approaches to supporting stu-
dent activists: Considerations for learning and development. Journal of Diversity
in Higher Education, 12(1), 17–26. [Link]
McPhail, D., Brady, J., & Gingras, J. (2017). Exposed social flesh: Toward an embod-
ied fat pedagogy. Fat Studies, 6(1), 17–37. [Link]
1142813
Roy, D., & Subramaniam, B. (2016). Matter in the shadows. In V. Pitts-­Taylor (Ed.),
Mattering: Feminism, science, and materialism (pp. 23–41). New York University
Press.
Sex Information and Education Council of Canada (SIECCAN). (2015). Sexual
assault in Canada: Legal definitions, statistics, and frontline responses. https://
[Link]/_files/ugd/1332d5_91cc31fb4c224134a72a47b12774c570.pdf
Spry, T. (2001). Performing autoethnography: An embodied methodological praxis.
Qualitative Inquiry, 7(6), 706–732. [Link]
Tuck, E. (2009). Suspending damage: A letter to communities. Harvard Educational
Review, 79(3), 409–428. [Link]
Underhill, A. (2021). #MeToo: (Re)shaping sexual research pedagogies through a
new material autoethnographic analysis. The Canadian Journal of Human
Sexuality, 30(2), 171–182. [Link]
Weaver-­Hightower, M. B. (2019). Analyzing self and other in autoethnography:
Telling secrets about one’s stillborn child. In Á. M. Humble & M. E. Radina
(Eds.), How qualitative data analysis happens: Moving beyond “themes emerged”
(pp. 1–15). Taylor & Francis. [Link]
2
USING INSTITUTIONAL
ETHNOGRAPHY TO TRACE
THE RULING OF WEIGHT
SURVEILLANCE WORK
Alexa R. Ferdinands and Kim D. Raine

Weight stigma—the blaming and shaming of people based on their body

weight (Calogero et al., 2016)—is a commonly overlooked public health
issue affecting young people in North America (Puhl & Lessard, 2020).
Weight-­ related discrimination remains socially acceptable, with weight-­
stigmatizing policies and practices pervading the social institutions, like
healthcare and education, we move through in our everyday lives (Rubino
et al., 2020). The multiplicative, population-­wide impacts on health and
well-­being of weight stigma (e.g., disordered eating and anxiety) across the
lifespan pose serious threats to public health (Puhl & Lessard, 2020). As a
public health researcher and dietitian, I (Alexa) worried about how I and
those in my profession(s) were contributing to weight stigma via interven-
tions targeting obesity (Ramos Salas, 2015). I conducted a study to better
understand how weight stigma happens in young people’s lives, hoping to
identify public health levers to reduce it (Ferdinands et al., 2021). Weight
surveillance—one prominent form of weight stigma experienced by young
people in this study—later became our investigative focal point. This surveil-
lance included activities like weight monitoring on scales, comparing their
body to others around them, and spending time scrutinizing themselves in
the mirror. Guided by my PhD supervisor (Kim), I chose institutional eth-
nography (IE) as my method of inquiry. Through this study, I showed how
young people’s experiences of growing up in their bodies were socially orga-
nized by dominant, biomedical weight and health discourses enacted in vari-
ous domains of their lives, such as at home, at school, and in healthcare
settings. In this chapter, I share how I analyzed the data generated in this
study, reflecting on challenges encountered along the way.

DOI: 10.4324/9781003254041-3
 Institutional Ethnography and Weight Surveillance 27

Positionality
I am a cis white, able-­bodied, thin woman trained in dietetics and public
health. My research is located within broader debates in weight-­related
research and practice (Pausé, 2022), such as arguments over whether obesity
is a disease or not. My weight history, professional background, and per-
sonal relationships with people in larger bodies continuously shape how I
situate myself within these debates. As a former weight management dieti-
tian, I sometimes hesitate to disclose my professional background, fearing I
will be grouped into the camp of biomedical “obesity alarmists” framing
fatness as a problem to be fixed (Gard, 2010, p. 10). Reading Gard’s scholar-
ship (Gard, 2010; Gard et al., 2022; Gard & Wright, 2005), wherein he
emphasizes the lack of certainty regarding taken-­for-­granted truths across
different perspectives of weight-­related scholarship, has helped me find my
footing. Similar to Gard, I try to work respectfully with people who think
differently than me, and I avoid painting everyone in a particular camp with
the same brush. I recognize my ability to do so is steeped in thin privilege.
Currently, I position my work at the intersection of critical obesity and
weight scholarship, questioning assumptions within obesity science and
holding interest in better understanding stigma and discrimination processes
(Pausé, 2022). This positionality drew me toward IE.

Institutional Ethnography
Emerging in the throes of second-­wave North American feminism, Smith
(1987, 1990a, 1990b, 2005, 2006) designed IE as a sociology to be done for
and with the people whose interests scholars aim to serve (i.e., our “stand-
point informants”). “Standpoint” refers to a particular social location from
which the research begins. Starting within the material events of people’s
lives, IE focuses on how things happen, as opposed to theorizing or abstract-
ing why (Smith, 2005). For this study, we took the standpoint of young peo-
ple in larger bodies. In IE, institutions are not brick-­and-­mortar structures,
but, rather, clusters of social relations coordinating people’s activities across
time and space around particular functions, such as education or healthcare
(Smith, 2005). Smith (1990a) refers to such social relations as ruling rela-
tions, focusing attention on how power is linked with institutional complexes
and their affiliated knowledges, whether they be bureaucratic, managerial,
scientific, legal, medical, or otherwise.
Discourses, work, and texts are three foundational IE concepts (Smith &
Griffith, 2022). Discourses point to how social relations organize individuals’
local activities via texts. Discourses are not abstractions; they are empirically
investigable practices in which people participate (and their participation, in
turn, reproduces the discourse). For example, in my study, informants
28 Alexa R. Ferdinands and Kim D. Raine

enacted a healthy weight discourse, which suggests everyone has an ideal

weight at which good health (and morality) can be achieved (Ferdinands
et al., 2021). This discourse was found in informants’ talk through references
to concepts such as the body mass index (BMI). Work is interpreted broadly
to include any activity (physical or mental) requiring time, energy, and
intent, directing researchers’ attention to often unacknowledged everyday
practices (Smith & Griffith, 2022). For instance, weight-­ related work
included activities like exercising, weighing oneself on the scale, and using
calorie-­tracking apps. Texts are anything that can be replicated across time
and space, as key coordinators of people’s everyday work. Researchers seek
to identify texts in informants’ talk when they implicitly describe text–work
sequences (Smith & Turner, 2014). I examined texts such as clinical obesity
guidelines, school anti-­bullying policies, media images (e.g., Instagram posts
from fitness influencers), and weight-­related YouTube videos. Additionally,
to understand how these foundational concepts are used in IE, I frequented
the methodology chapters in IE dissertations, especially those from Smith’s
former students (e.g., Khayatt [1987], McCoy [1999]), which tend to include
more detailed methodological descriptions than published manuscripts.
These chapters helped me learn how to develop my research problematic.

Developing the Research Problematic

A research problematic in IE uncovers latent questions about the everyday
world (Smith, 2005). Problematics differ from research questions or objec-
tives (Campbell & Gregor, 2008). They are grounded in disjunctures between
“official” and experiential knowledge, wherein standpoint informants’ expe-
riences are inaccurately accounted for, or overlooked altogether, within
dominant discourses (Rankin, 2017b).
I attended IE workshops led by prominent scholars, including Dorothy
Smith and Susan Turner, to help formulate my problematic. There, I con-
nected with other novice institutional ethnographers who were equally con-
cerned about doing IE “correctly.” Smith and Turner encouraged us to write
about what we already knew about our topic of interest and how we came to
know it. They emphasized the importance of drawing on our personal
knowledge and questioning our assumptions before formally collecting
data. I naively hoped I would depart the workshop with a completed prob-
lematic in hand, but little did I realize that it would take many, many months
to fully elucidate this. I wrote pages of personal reflexive journal notes, often
in a stream-­of-­consciousness manner, about how my dietetics experiences
evoked my interest in weight stigma. Reflecting on my personal and profes-
sional encounters with people in larger bodies, I discerned a disjuncture
between experiential and authorized knowledges of fatness. For example,
 Institutional Ethnography and Weight Surveillance 29

clinical weight management guidelines are often couched in individualism,

assigning personal responsibility to people for their weight. But the patients
I counseled did not match the fat stereotypes portrayed in medical textbooks
and the media.
At the IE workshops, I wrangled with my “institutional capture,” includ-
ing how I failed to question familiar terminology used in dietetics (Bisaillon,
2012a). IE demanded I interrogate the shells within institutional discourses
(Smith, 2005). Shells are materially empty terms, like stigma and obesity,
that readers and listeners instinctively fill in with assumptions about every-
day experience. I leaned on reflexive writing to avoid using such theoretically
loaded terms when explaining people’s practices. However, I struggled (and
still do) with the contested issue of selecting appropriate weight-­related ter-
minology. As a dietitian, talking and writing in biomedical language rooted
in the BMI, like overweight and obesity, was second nature to me. But using
IE required me to stay grounded in the actualities of informants’ lives. In
contrast with conceptual shells like obesity, fatness is tangible. However,
despite fat activists’ reclamation of the term “fat” as a neutral descriptor of
body size, some informants felt it remained a hurtful word that continued to
be weaponized against them. Based on conversations with informants and
the literature (e.g., Hart et al., 2020), I landed upon the phrase “living in
larger bodies” to refer to body weight and size.
Eventually, I decided the purpose of my research was to produce useful
knowledge about the social organization of young people’s work of growing
up in larger bodies. I learned about this work by listening to informants’
descriptions of their lives and identifying the difficulties they faced. The
problematic became increasingly clear with time and critical reflection, as I
moved back and forth between conversations with informants and the litera-
ture. Because there were no other IEs on this research issue, to my knowl-
edge, the potential for mapping out social relations in this field was enormous.
Many institutions such as healthcare, education, fashion, media, and the
home arose in our conversations, and it would have been impossible to
untangle them all in a single analysis. Consequently, I discussed how to place
manageable boundaries around my study with fellow students in Kim’s lab.
As I talked about my early insights, my student colleagues pointed out the
separation they saw between the work tied to healthcare and education and
the work tied to media and fashion. Therefore, for the manuscript discussed
herein, I focused on informants’ work as it relates to healthcare, education,
and the home. But the work of growing up in a larger body, even when lim-
ited to these three institutions, remained too vast of a topic. In helping me
to narrow my focus, Kim noted that the most prominent sub-­categories of
work reported by informants could be subsumed under the category of
weight surveillance.
30 Alexa R. Ferdinands and Kim D. Raine

Data Generation
Each informant’s story helped me follow the thread of inquiry (weight sur-
veillance) and see more clearly how the problematic arose institutionally. I
used two qualitative data-­generating strategies in this study: in-­person one-­
on-­one interviews and group interviews. I recruited 16 informants (14 women
and 2 men) between the ages of 15 and 21 who self-­identified as having
grown up in a larger body. From March to May 2019, I conducted one to
two interviews per week, with a new informant each time, for a total of 14
weeks. I scheduled the interviews, each approximately one hour long, in this
way to allow time for preliminary analyses between interviews and to modify
my questions as needed for subsequent interviews, basing each interview on
what I had learned in previous ones.
The grand tour question (Spradley, 1979) I posed to informants was “How
did it feel to grow up in your body?” Follow-­up questions involved asking
them to walk me through a typical day of activities, starting from when they
woke up to when they went to sleep. Their responses unveiled a range of
weight-­related tasks, including those seemingly mundane, that may have oth-
erwise remained unexamined (e.g., comparing one’s body size or shape to
others around them, measuring food portions on a scale, and researching diet
or exercise plans). The interview guide was open-­ended, enabling me to probe
about informants’ work and map out the social organization of people’s
interconnected activities across multiple sites. DeVault and McCoy (2006)
noted that Mykhalovskiy describes interviewing as involving much more
than just asking questions; it also demands on-­the-­spot analysis. Running
early analytic thoughts past informants during the interview can help move
the analysis forward. Similar to Mykhalovskiy, I tried to think analytically
during the interviews. Drawing on nutrition counseling skills such as para-
phrasing was helpful as I tried to quickly make sense of what I had heard and
determine what follow-­up questions needed to be asked.
Some informants asked me why I was doing this research, perhaps skepti-
cal of my intentions. It appeared they were requesting insight into my moti-
vations for undertaking this research, wanting to ensure that I would
represent them and their stories in a way in which they were comfortable.
Upon sharing my rationale, one informant replied, “I don’t want to say I’m
surprised but you are like a skinnier person and it’s nice to see that this is an
interest for people that aren’t just [larger], you know what I mean?” Based on
informants’ body language, follow-­up comments, and subsequent participa-
tion in the group interviews, I believe they perceived my interest as genuine,
though I cannot know for sure. Sometimes, my intersecting social identities
overlapped with those of my informants; sometimes, they did not. For exam-
ple, as a woman, I found myself nodding along to women informants’
descriptions of appearance-­related pressures. Many informants were also
 Institutional Ethnography and Weight Surveillance 31

university students, like me. I wavered in terms of how much was appropriate
to share with informants, though. Weight stigma is a topic that nearly every-
one has personal experience with, but I wanted to keep the conversation cen-
tered on my informants. I took detailed interview notes immediately following
each interview, including general impressions of the conversation, reflections,
emotional reactions, and new or persistent questions. Engaging in reflexive
practices, such as maintaining a personal reflexive journal, helped me stay
accountable to the ethical commitments I had made to informants and inter-
rogate how my positionality affected my relationship with them, the research
direction, and my overall approach to data generation and analysis.
After the individual interviews were completed, I invited women infor-
mants to participate in five repeated group interviews between September
2019 and January 2020 (approximately one per month, ranging from one to
two hours each). I limited the group to women, given the gendered nature of
weight-­related social relations that had become apparent in our earlier inter-
views; informants may have felt less comfortable in sharing personal experi-
ences had other genders been included in the group. Five women participated
in these sessions, assisting us in clarifying our initial understandings of the
individual interview data.
I audio recorded all individual and group interviews and transcribed
them verbatim, with the aid of free software called Express Scribe. I did not
have a transcription foot pedal, but I suspect that would have sped up the
process. I enjoyed transcription; it was a pleasant break from deeper analytic
thinking. I completed transcription in two rounds, with the second round
serving as an accuracy check. Transcription also helped me familiarize
myself with the data.
As I began data analysis, sometimes I had to seek out additional informa-
tion to answer emergent questions. For instance, I initiated several informal
email and in-­person conversations with translocal informants (i.e., those not
situated at the standpoint location) to glean further details about particular
work processes. For example, I spoke with a few education professors to
learn about health curricula for pre-­service teachers, which may shape how
teachers implement health lessons in their classrooms. I took notes during
and after these conversations.

Analysis
IE analysis differs from conventional approaches to qualitative data analy-
sis. Rather than connecting to or generating theory, institutional ethnogra-
phers aim to trace exactly what is happening. Having a supervisor who “got”
IE (Kim had used IE in her PhD) was beneficial to me during analysis. I
heard stories from fellow students in less-­than-­ideal circumstances whose
supervisors were not as receptive to IE’s alternative paradigm and
32 Alexa R. Ferdinands and Kim D. Raine

questioned their students’ avoidance of theming and abstractions. As this

study was my first qualitative project, I also had the advantage of never hav-
ing analyzed qualitative data before; I had fewer pre conceived ideas to
unlearn about what qualitative analysis should look like.
There is not one analytic strategy or writing format universally deemed
correct for doing IE. IE researchers are driven by the overarching goal of
keeping the institution and people in view, understanding there are various
ways of achieving that goal. There are a handful of IE publications dedi-
cated to unpacking IE analysis, with each author operating from a unique
vantage point. Three examples are as follows:

• Chapter 5 in Mapping Social Relations (Campbell & Gregor, 2008, pp.

83–102)
• Part II: Analysis in IE as Practice (Smith, 2006, pp. 89–162)
• Rankin’s (2017a, 2017b) two-­part series on IE analysis
I also read as many IE studies as I could get my hands on, attempting to
grasp the range of possibilities (e.g., DeVault, 1991; Khayatt, 1994; Ross
et al., 2018; Smith & Smith, 1998). Grounded in this learning, I used five
overlapping, iterative strategies for analyzing the data: (a) reading, listening,
and deliberating; (b) indexing; (c) writing narrative accounts and “analytic
chunks;” (d) mapping; and (e) talking with critical friends and colleagues.

Reading, Listening, and Deliberating

The first, and ongoing, strategy I used was what Bisaillon (2012b) referred to
as “reading and deliberating” (p. 141). To Bisaillon’s strategy I add “listen-
ing.” Listening to informants’ voices and tone brought me back to our inter-
views, offering another opportunity to garner insight into the data that may
not have been captured from reading the transcripts alone. I relistened to the
audio recordings while going for walks outside, trying to immerse myself in
the data in myriad ways. Sometimes, I paused in my walk to make notes on
my mobile phone; other times I waited until returning home, where I quickly
jotted ideas down on scrap paper. I call these early analytic stages “playing
with the data” because it felt playful, creative, and fun. The notion of play
helped me avoid entering analysis with blinders on to interesting threads.
My problematic was intended to guide, not constrain, me.
As I read and reread my data (16 individual interview transcripts, 5 group
interview transcripts, notes from interviews and informal conversations, rel-
evant documents like clinical obesity guidelines, and reflexive journal notes),
I asked myself questions like, “What do these data tell me about how a par-
ticular event (e.g., weight teasing at home) happens as it does?” and “How is
it that informants are saying what they are saying?” McCoy (2006) describes
 Institutional Ethnography and Weight Surveillance 33

engagement with interview transcripts as akin to doing an interview with the

data itself, as the researcher asks themselves questions to understand the
intricate details of informants’ work. For example, what knowledge or skills
do informants need to do their work? And how is their work interconnected
with others’ work?
The analysis involved moving back and forth between scrutinizing infor-
mants’ talk and the contexts that produced that talk. Points of tension, such
as how questionable strategies used to deliver school nutrition curricula
(e.g., requiring students to use calorie-­tracking apps for assignments) some-
times contributed to disordered eating, shaped the problematic. I reviewed
my collection of data regularly, recognizing I could piece together the emerg-
ing big picture and make new discoveries through further readings. In addi-
tion to the generated data, I reread seminal IE works countless times, each
time seeing something new in their pages.

Indexing
Like Rankin (2017b), I found that indexing was a valuable early analytic
step for working with and organizing the initially overwhelming amount of
data. Smith (2006) describes indexing as a strategy for organizing data in a
way that supports making work processes visible. Some institutional eth-
nographers use the term “coding” (more commonly used among qualita-
tive researchers) to describe this strategy. However, I intentionally used
“indexing” to help me stay focused on IE’s materialist foundations. In
alignment with IE’s ontology, I did not want to develop abstract themes or
categories through analysis; I wanted to empirically trace linked practices
going on in the world to unveil practical information that people could use
to make change.
At first, I found it difficult to conceptualize what indexing looked like in
practice. To resolve this lack of clarity, I turned to Bisaillon’s (2012b) dis-
sertation, wherein she thoroughly describes her indexing process. I fol-
lowed her approach for annotating textual data, which involved
color-­coding the data during successive readings (Bisaillon, 2012b). I
printed the textual data and manually highlighted them, assigning one
color for each of the following five categories: work (what people were
doing and their interactions with others), talk (language people used),
texts (documents revealed in informants’ talk), people (those informants
interacted with), and institutions (Bisaillon, 2012b). Document 2.1 shows a
transcript excerpt from this study that I color-­coded and marked up with
preliminary thoughts.

[SEE DOCUMENT 2.1 at eResource: Example of Color-­Coded Interview Transcript].

34 Alexa R. Ferdinands and Kim D. Raine

Although this process was a useful analytic entry point, this particular classifi-
cation system was not very fruitful for moving my analysis beyond data famil-
iarization. Some IEs, like Bisaillon’s dissertation, involve reams of documents
and other texts to be analyzed. Mine, however, included just a few. I suspect
Bisaillon’s annotation technique may be more valuable for IEs that involve
many texts. Further, while I initially thought that I would prefer paper copies of
the data for indexing, I found the volume of hard copy data to be quite over-
whelming. After the first round of indexing hard copy transcripts, I returned to
marking up Microsoft Word copies using the comment function. Document 2.2
provides an example of an indexed Microsoft Word transcript excerpt.

[SEE DOCUMENT 2.2 at eResource: Sample Indexing of an Electronic Interview

Transcript].

Microsoft Word’s “find” feature was handy for navigating the data. Although I
had taken an NVivo course with the intention of using that software in my
PhD, conversations with trainees at the IE workshops convinced me otherwise.
Qualitative analysis software can be used as a data management tool for index-
ing, but it seemed like using such software might overcomplicate what could be
easily accomplished through Microsowpft Excel and Word. That said, I can see
how larger teams involving multiple researchers analyzing data simultaneously
could benefit from such software, which may streamline data management.
Because I was doing my PhD, in my case, this was more of a solo endeavor.
After completing the initial round of annotating, I employed a (mostly)
inductive indexing approach based on general topic headings associated
with work processes. I tried to concentrate on people’s work practices
(whether they be physical, mental, social, or emotional), keeping work pro-
cesses and the social at the analytic forefront. I used Microsoft Excel for
organizing my indices and quotes (Table 2.1), and I continued to update this
same Excel document rather than create newer versions over time. In future
research, however, I will save separate versions every few days to better track
how the index changed over time.

[SEE TABLE 2.1 at eResource: Sample Index Headings].

It is important to note that indexing was not an endpoint, but, rather, a pre-
liminary step. Some indexed headings did persist all the way through to the
final analysis, though, and are labeled as such in the manuscript. Through
indexing, I was able to observe overlaps across indices. For example, quotes
pertaining to “using the BMI” index also pertained to the “quantifying
weight and bodies” index. Documenting these overlaps in my analytic writ-
ing (see next section) helped me to see where and how work processes and
practices might be linked.
 Institutional Ethnography and Weight Surveillance 35

Writing Narrative Accounts and “Analytic Chunks”

When I began this research, I was perplexed by the notion of writing as
analysis. How was I to know what to write until I had completed the analy-
sis? This confusion dissipated as I indexed and wrote in an iterative fashion.
I now appreciate the crucial contribution of writing to deepening analytic
thinking. The first more surface-­level stage in writing was developing narra-
tive accounts. I wrote a single-­spaced, one-­to two-­page summary for each
informant, narrating their life as reported and beginning to piece together its
social organization. As an example, Document 2.3 summarizes the life story
of 15-­year-­old Jenny, who struggled with disordered eating and participated
in a weight-­loss summer camp. Writing these accounts helped to expose
latent work processes that have been indexed.

[SEE DOCUMENT 2.3 at eResource: Jenny’s Informant Summary].

The second stage, writing analytic chunks, involved picking an event as

reported by informants and detailing how it was coordinated, further con-
necting local to translocal social relations (Rankin, 2017b). As I indexed, I
maintained a running list of thoughts that I titled “things I found interesting
and worth exploring.” These broad points of interest (Table 2.2) that refer-
enced informants’ practices became the analytic chunks of focus.
TABLE 2.2 Things I Found Interesting and Worth Exploring

• Participating in gym class: being last in sports competitions (pretending like

you’re not trying or don’t care); getting changed in the washroom rather than
the change rooms (preferring a closed door for privacy)
• Facing weight-­related discrimination while trying to find a romantic partner
• Weight teasing from family members and gender differences (e.g., fathers’
comments vs. mothers’ comments)
• Interacting with adult authority figures: standing up for yourself, but also
respecting your elders
• Weight cycling
• Dieting
• Learning about body positivity on social media (e.g., reclaiming the term “fat”)
• Knowing and using the BMI from a young age
• Hearing, reading, and using fat talk
• Noticing a gradual increase in the prevalence of fat bodies in the media
• Quantifying and obsessing about weight
• Fearing the presence of weight scales in healthcare settings, schools, gyms, and
bathrooms
• Observing clinicians dodge the topic of obesity, especially when working with
children
• Hesitating in choosing weight-­related terminology
• Shopping, not being able to find clothes that fit: being excluded from wearing
conventional “little girl” clothes
36 Alexa R. Ferdinands and Kim D. Raine

I wrote informally about each of these points, supported by informants’

quotes. For example, one analytic chunk revolved around fat-­shaming at
home. I was troubled by how consistently informants described their parents
as key perpetrators of fat-­ shaming in their lives. Informants described
frankly shocking instances of parental bullying but quickly switched gears
to protect their parents, reassuring me their parents only said these things
because they wanted the best for them and were concerned about their health
and well-­being. I reflected in my journal and discovered that my upbringing
differed quite drastically from that of my informants. My parents almost
never discussed weight or dieting in front of me. Comparing and contrasting
the social organization of informants’ experiences with my own, and with
those of other informants, helped me identify ruling relations.
During this stage, the lines between informal reflexive journal writing and
more formal analytic writing became blurred. Personal journal writing had
been essential to developing my problematic. But the deeper I ventured into
analysis, the less time I spent writing in my personal journal. Instead, I dedi-
cated more time toward formal analytic writing. For a while I got stuck in
writing mere topic summaries; it was difficult to analytically leap into the
translocal. But writing helped me work through that sticky point. Writing
enabled me to see the social relations that I needed to see and discover what
I wanted to show others. I was writing my way to clarity.

Mapping
Like writing, mapping was another technique that moved me through analytic
ruts (Rankin, 2017b). Similar to my confusion around the concept of “writing
as analysis” I struggled to know how to start mapping. Writing was more com-
fortable for me than drawing diagrams. Kim advised me to start sketching out
my findings early on in my analysis, but admittedly, I put it off for a few months.
As soon as I put (literal) pen to paper, however, I readily saw the value in this
approach. Mapping forced me to shift my gaze from the realm of abstract ideas
to the materiality of the ruling relations. In this way, mapping helped to identify
mystical connections—a term that Smith (1990b) adopted from Marx and
Engels to emphasize situations where theoretical concepts are reified to explain
how something works, despite being detached from actual people’s experiences.
My job as an institutional ethnographer was to empirically track ruling rela-
tions from the local work of standpoint informants into the translocal work of
other people. I needed to display what was happening in a diagram that depicted
how work processes are linked and coordinated. I referred to Turner’s (2006)
work for mapping guidance. Turner is well renowned for the intricate mapping
techniques she developed during her PhD.
I did all my mapping work by hand on paper and whiteboards. Figure 2.1
illustrates a paper copy of a map depicting the social organization of young
 Institutional Ethnography and Weight Surveillance 37

people’s weight surveillance work, which I used to organize my writing of the

findings. I also used these maps to direct my conversations with critical friends,
particularly via whiteboard discussions in Kim’s lab. Because these friends were
not immersed in the weight stigma literature like me, they were less prone to
institutional capture and helped to identify gaps in my thinking. For example,
in the process of sketching out and explaining earlier versions of these maps,
my friends posed questions like, “How do you know that?” When I could not
effectively answer based on the data, I knew that my analysis was not (yet)
empirically grounded.

[SEE FIGURE 2.1 at eResource: Mapping the Social Organization of Young People’s
Weight Surveillance Work].

Talking with Critical Friends and My Supervisory Committee

Conversations with others are often mentioned in passing as part of qualita-
tive analysis, but these were so vital in my PhD that they warrant explicit dis-
cussion here. I took notes while talking with critical friends and my supervisory
committee, which I continuously integrated into the analysis. Sometimes,
these conversations were structured in nature, such as those with my supervi-
sory committee where I arrived with an agenda and specific questions. Other
times, these conversations were less formal, and I simply described where I was
at in the analysis and what was interesting to me at the time. In conversations
with friends, I related informants’ knowledge to not just the broader literature
but also to my personal and professional experiences with people living in
larger bodies. Because weight stigma is so pervasive, nearly every day some-
thing came up that seemed relevant to the analysis, such as conversing with a
friend about a family member who constantly used fat talk.
To make progress in my analysis, I relied heavily on oral and written conver-
sations with two fellow PhD students. These relationships were mutually benefi-
cial as I served the critical friend role for their respective analyses. Our
conversations took place in person as well as via Skype, voice memos (WhatsApp
and Vox), text, and email. I met one friend, Erika Ono, at an IE workshop; we
assisted each other in unpacking sociological terms (e.g., what does discourse in
IE actually mean?) and checking that our analyses aligned with an IE approach.
I met my other friend, Paige Reeves, at a qualitative methodology conference.
She is using critical ethnography in the field of disability studies; there are inter-
esting overlaps in both our methodological approaches and topic areas. Her
sharp eye forced me to question my assumptions about IE. In the literature, IE
is often framed as being radically different from other qualitative approaches.
But that framing has since evolved and is more nuanced now than it was when
IE was first being developed, as other feminist sociologies too have advanced
and become more commonplace. Both Erika and Paige read drafts of my ana-
lytic writing and critiqued my maps.
38 Alexa R. Ferdinands and Kim D. Raine

More formally, I met with Kim and my supervisory committee members

(Kate Storey and Tara-­Leigh McHugh) regularly throughout the analysis. I
had collected reams of data, much more than could be used in a single anal-
ysis. I sometimes got sidetracked in mapping out social relations extraneous
to the thread I was attempting to follow, and my committee helped me return
to my focal point. Their written feedback on drafts were particularly valu-
able for ensuring the rigor of my research.
As Bisaillon (2012b) explains, an IE study can be deemed rigorous when
the researcher has produced a clear and accurate depiction of how things
actually work (in regard to the focal point of study) in people’s everyday
lives. Toward the end of my analysis, my committee also helped me work
through the issue of knowing when I was done, that is, when I had generated
a useful understanding of weight surveillance work. Likely related to my
status as a novice qualitative researcher, this endpoint seemed fuzzy. My
committee reassured me that I did not need to tackle every issue that infor-
mants raised in this manuscript and that I still had the rest of my career to
unpack how weight stigma works.

Reflections on IE Imposter Syndrome

While moving through the analysis, I encountered a range of challenges
associated with my feelings of being an IE imposter. For instance, through-
out my research, I was perplexed by how to engage with other theories, like
intersectionality, when using IE. In retrospect, I believe my analysis could
have been strengthened by engaging with more critical and feminist theories.
However, I hesitated to do so because of IE’s complicated relationship with
other theories like intersectionality (Campbell, 2016) and scholars’ warnings
about the risk of falling into the sociological trap of objectifying and catego-
rizing people’s embodied experiences. But in my reading of Smith’s work, it
seems that she advises researchers to neither start nor end in theory—not to
entirely avoid engagement with other theories. Smith designed IE to be flex-
ible in understanding a range of social problems and encouraged continuous
innovation, which demands novel theoretical engagement (Smith & Griffith,
2022). I discussed this topic with my friend Erika, who is expanding on the
concept of intersectionality with an IE lens in her dissertation. Erika explains:

[T]he analytic framework of intersectionality [can be] utilized in order to

acknowledge the intersecting relations of ruling, drawing on Smith’s
((2005) conceptualization of institutions as “the intersection and coordi-
nation of more than one relational mode of ruling” (p. 225). Just as an
intersectional lens advances institutional ethnography, this method of
inquiry strengthens the concept of intersectionality because it investigates
 Institutional Ethnography and Weight Surveillance 39

how relations of ruling actually happen through our actions and interac-
tions with others. Institutional ethnography can uncover how we do inter-
sectionality and keep us from abstracting and reifying embodied
experiences by understanding how ‘the social’ is organized through these
intersecting ruling relations.
(Ono, 2023, p. 60, emphases in original)

My fear of boundary monitoring by other IE scholars (Murray et al., 2021)

was not unfounded. A few years ago, I presented preliminary IE findings at
a conference, excited to get feedback on my work. Despite earlier remarks at
the conference from scholars about how disheartening it is to be told that
your work is “not IE,” I was told just that by an audience member. In the
moment, I swallowed my tears, nodded, and thanked them for their feed-
back. I flew home from the conference feeling like I was doing IE wrong.
Since then, I have witnessed colleagues endure similar situations. I whole-
heartedly agree with the importance of methodological coherence and
maintaining IE’s integrity; my IE knowledge has certainly evolved from
reflecting on mistakes in my analyses presented at that conference. I had suc-
cumbed to “analytic drift” (McCoy, 2006, p. 109), focusing too much on
individuals and their experiences rather than the institutions coordinating
those experiences. But there is a difference between monitoring (exclusive)
and mentoring (inclusive), the latter being preferable. We can still promote
rigorous IE research while prioritizing kindness. For example, asking genu-
ine questions, either during or after the conference session, can prompt
trainees to think differently about their work in a way which is ontologically
compatible with IE.
I continue to ponder the question of who gets to decide what is, or is not,
IE (or any other qualitative approach). Are there circumstances where
boundary monitoring is appropriate and others where it is not? These
debates are not new (see Saldaña [2014] for a humorous take on qualitative
gatekeeping). In my experiences as a novice institutional ethnographer with
an inclination toward pragmatism, returning my gaze to IE’s core social
ontology (McCoy, 2021; Smith, 2005) and original intent of producing use-
ful knowledge that can help people address social problems helped me miti-
gate some of my worries about IE orthodoxy during data analysis.
When publishing Ferdinands et al. (2021), one reviewer did question the
merits of this study as IE because it did not adhere to conventional IE for-
mulae. For example, this study did not involve significant textual analysis—I
focused more on analyzing informants’ talk. I also had not formally inter-
viewed translocal informants (e.g., parents, healthcare providers, and educa-
tors), a common second phase of an IE study. I responded to, and satisfied,
the reviewer by elaborating on how I had modeled my study design upon
40 Alexa R. Ferdinands and Kim D. Raine

other reputable published IE works (e.g., Smith & Smith, 1998). The beauty
of IE is that there are countless ways to approach data generation and
analysis.

Conclusion
The goal of analysis in IE is to show how ruling relations are enacted to
shape the activities and experiences of people in their everyday lives (McCoy,
2006). In this chapter, I have described how I used five iterative steps (read-
ing, listening, and deliberating; indexing; writing narrative accounts; map-
ping; and talking with critical friends and colleagues) to analyze IE data
from a study examining young people’s weight surveillance work. By sharing
my experiences, the challenges, and the breakthroughs, I hope to promote
transparency in IE data analysis and highlight the utility of IE for research-
ers interested in examining a range of social problems. The very nature of
IE—an alternative way of looking at the world—prohibits the publication
of a step-­by-­step, how-­to guide. But in detailing how I analyzed data in this
IE and reflecting on lessons learned in the process, I hope that the insights
gleaned will be useful for other researchers interested in better understand-
ing and using IE on their own terms.

KEY WORKS GUIDING OUR DATA ANALYSIS

Bisaillon, L. (2012a). An analytic glossary to social inquiry using institutional and politi-
cal activist ethnography. International Journal of Qualitative Methods, 11(5), 607–
627. [Link]
This paper includes a detailed glossary of IE-­related terms. It is a great reference
for designing and conducting an IE.
Campbell, M., & Gregor, F. (2008). Mapping social relations: A primer in doing institu-
tional ethnography. University of Toronto Press.
This is an excellent, concise introductory text for those who are new to, but are
interested in, using IE. Chapter 5 (pp. 83–102) is dedicated to analyzing data in IE.
Smith, D. E. (2005). Institutional ethnography: A sociology for people. Rowman Altamira.
This book provides the theoretical backbone of IE, offering updated arguments
from Smith’s (1987) book, The everyday world as problematic. We consider it essen-
tial reading for anyone conducting an IE.
Smith, D. E. (2006). Institutional ethnography as practice. Rowman Altamira.
Part II of this book focuses on analysis, with chapters from renowned IE scholars,
including Marie Campbell, Liza McCoy, Alison Griffith, and Susan Turner. Practical
examples of how to work with interview data and map and how to discover institu-
tional discourses are shared.
Turner, S. (2023). Mapping for Change. [Link]
Susan Turner developed a useful website with resources, including mapping
worksheets and instructional videos, for IE mapping.
 Institutional Ethnography and Weight Surveillance 41

References
Bisaillon, L. (2012a). An analytic glossary to social inquiry using institutional and
political activist ethnography. International Journal of Qualitative Methods, 11(5),
607–627. [Link]
Bisaillon, L. (2012b). Cordon sanitaire or healthy policy? How prospective immigrants
with HIV are organized by Canada’s mandatory HIV screening policy [Unpublished
doctoral dissertation]. University of Ottawa. [Link]
10393/20643
Calogero, R. M., Tylka, T. L., & Mensinger, J. L. (2016). Scientific weightism: A view
of mainstream weight stigma research through a feminist lens. In T. A. Roberts,
N. Curtin, L. E. Duncan, & L. M. Cortina (Eds.), Feminist perspectives on building
a better psychological science of gender (pp. 9–28). Springer. [Link]
10.1007/978-­3-­319-­32141-­7_2
Campbell, M. L. (2016). Intersectionality, policy-­oriented research and the social
relations of knowing. Gender, Work & Organization, 23(3), 248–260. [Link]
org/10.1111/gwao.12083
Campbell, M., & Gregor, F. (2008). Mapping social relations: A primer in doing insti-
tutional ethnography. University of Toronto Press.
DeVault, M. L. (1991). Feeding the family: The social organization of caring as gen-
dered work. University of Chicago Press.
DeVault, M., & McCoy, L. (2006). Institutional ethnography: Using interviews to
investigate ruling relations. In D. E. Smith (Ed.), Institutional ethnography as
practice (pp. 15–44). Rowman & Littlefield.
Ferdinands, A. R., McHugh, T. L. F., Storey, K. E., & Raine, K. D. (2021). The rul-
ing of weight: An institutional ethnography investigating young people’s body
weight surveillance work. Social Science & Medicine, 289. [Link]
socscimed.2021.114404
Gard, M. (2010). The end of the obesity epidemic. Routledge.
Gard, M., Powell, D., & Tenorio, J. (Eds.). (2022). Routledge handbook of critical
obesity studies. Routledge.
Gard, M., & Wright, J. (2005). The obesity epidemic: Science, morality and ideology.
Routledge.
Hart, L., Ferreira, K., Ambwani, S., Gibson, E., & Austin, S. (2020). A roadmap for
addressing weight stigma in public health research, policy and practice. Strategic
Training Initiative for the Prevention of Eating Disorders (STRIPED). https://
[Link]/wp-­c ontent/uploads/sites/1267/2020/06/STRIPED-­
Roadmap-­Delphi-­Study-­Weight-­[Link]
Khayatt, D. (1987). Gender role conformity in women teachers (NL 39200).
[Unpublished doctoral dissertation]. University of Toronto. ProQuest
Dissertations Publishing.
Khayatt, D. (1994). Surviving school as a lesbian student. Gender and Education,
6(1), 47–62. [Link]
McCoy, L. (1999). Accounting discourse and textual practices of ruling: A study of
institutional transformation and restructuring in higher education (NQ45788).
[Unpublished doctoral dissertation]. University of Toronto.
42 Alexa R. Ferdinands and Kim D. Raine

McCoy, L. (2006). Keeping the institution in view: Working with interview accounts
of everyday experience. In D. E. Smith (Ed.), Institutional ethnography as practice
(pp. 109–125). Rowman & Littlefield.
McCoy, L. (2021). Materialist matters: A case for revisiting the social ontology of
institutional ethnography. In P. C. Luken, & S. Vaughan (Eds.), The Palgrave
handbook of institutional ethnography (pp. 35–46). Springer Nature Switzerland.
Murray, Ó. M., Ablett, E., & Delucchi, A. S. (2021). Between orthodoxy and open-
ness: A book review essay on The Palgrave handbook of institutional ethnography.
International Journal of Social Research Methodology, 26(3), 353–362. [Link]
org/10.1080/13645579.2021.1984738
Ono, E. (2023). The violence of “best interests of the child”: Social work in the ruling
relations. [Unpublished doctoral dissertation]. University of British Columbia.
Pausé, C. (2022). Devil pray: Fat studies in an obesity research world. In M. Gard, D.
Powell, & J. Tenorio (Eds.), Routledge handbook of critical obesity studies (pp.
77–84). Routledge.
Puhl, R. M., & Lessard, L. M. (2020). Weight stigma in youth: Prevalence, conse-
quences, and considerations for clinical practice. Current Obesity Reports, 9(4),
402–411. [Link]
Ramos Salas, X. (2015). The ineffectiveness and unintended consequences of the
public health war on obesity. Canadian Journal of Public Health, 106(2), e79–e81.
[Link]
Rankin, J. (2017a). Conducting analysis in institutional ethnography: Analytical
work prior to commencing data collection. International Journal of Qualitative
Methods, 16(1), 1–9. [Link]
Rankin, J. (2017b). Conducting analysis in institutional ethnography: Guidance and
cautions. International Journal of Qualitative Methods, 16(1), 1–11. [Link]
org/10.1177/1609406917734472
Ross, C. A., Jakubec, S. L., Berry, N. S., & Smye, V. (2018). “A two glass of wine
shift”: Dominant discourses and the social organization of nurses’ substance use.
Global Qualitative Nursing Research, 5, 1–12. [Link]
2333393618810655
Rubino, F., Puhl, R. M., Cummings, D. E., Eckel, R. H., Ryan, D. H., Mechanick, J.
I., Nadglowski, J., Ramos Salas, X., Schauer, P. R., Twenefour, D., Apovian, C.
M., Aronne, L. J., Batterham, R. L., Berthoud, H.-R., Boza, C., Busetto, L.,
Dicker, D., De Groot, M., Eisenberg, D., … & Dixon, J. B. (2020). Joint interna-
tional consensus statement for ending stigma of obesity. Nature Medicine, 26(4),
485–497. [Link]
Saldaña, J. (2014). Blue-­collar qualitative research: A rant. Qualitative Inquiry, 20(8),
976–980. [Link]
Smith, D. E. (1987). The everyday world as problematic: A feminist sociology.
University of Toronto Press.
Smith, D. E. (1990a). Texts, facts, and femininity: Exploring the relations of ruling.
Routledge.
Smith, D. E. (1990b). The conceptual practices of power: A feminist sociology of
knowledge. University of Toronto Press.
Smith, D. E. (2005). Institutional ethnography: A sociology for people. AltaMira
Press.
 Institutional Ethnography and Weight Surveillance 43

Smith, D. E. (Ed.). (2006). Institutional ethnography as practice. Rowman &

Littlefield.
Smith, D. E., & Griffith, A. I. (2022). Simply institutional ethnography: Creating a
sociology for people. University of Toronto Press.
Smith, G. W., & Smith, D. E. (1998). The ideology of “fag”: The school experience
of gay students. The Sociological Quarterly, 39(2), 309–335. [Link]
10.1111/j.1533-­8525.1998.tb00506.x
Smith, D. E., & Turner, S. M. (Eds.). (2014). Incorporating texts into institutional
ethnographies. University of Toronto Press.
Spradley, J. P. (1979). The ethnographic interview. Holt, Rinehart and Winston.
Turner, S. M. (2006). Mapping institutions as work and texts. In D. E. Smith (Ed.),
Institutional ethnography as practice (pp. 139–161). Rowman & Littlefield.
3
WRITING A NEW MATERIALIST
ETHNOGRAPHY ON POLYAMOROUS
PARENTS
Cornelia Schadler

This chapter explains, in detail, how I conducted the analyses for a paper I
published in Sexualities on polyamorous parents (Schadler, 2021).
Polyamorous parents are parents who have more than one intimate partner
as well as children with one or more of these partners. In the paper, I made
a difference between three types of parents in consensual non-­monogamous
relationships:

(a) poly-­nuclear parenting (simulating two-­parent families with biological

and legal children), (b) hierarchical parenting (a network of at least three
parents that differ on their degree of involvement), and (c) egalitarian
parenting (where three or more members of a polyamorous network are
perceived as equal parents).
(Schadler, 2021, p. 2)

The data for this project included interviews with at least one parent of 13
networks (N = 18) and podcasts, websites, and self-­help books the parents
read or listened to (Schadler, 2021). The interviews were the focus of the
analysis.
The study was situated in the fields of family science and gender studies.
Family science is a field that studies the basic structures of family living
(Williams, 2020), such as the economic structures and legal basis of family
living (Cooper & Pugh, 2020), the impact of gender roles on family pro-
cesses and their divisions of tasks, and the evolvement of family as a form of
living (Schadler, 2016; Smock & Schwartz, 2020). In recent years, family
science has also extended its scope to theoretical developments in gender

DOI: 10.4324/9781003254041-4
 Writing a New Materialist Ethnography on Polyamorous Parents 45

studies (Few-­Demo & Allen, 2020; Reczek, 2020). New Materialism (Barad,
2007; Dolphijn & van der Tuin, 2012), the theory I worked with in the study
on polyamorous parents, is a materialist–poststructuralist theory that
emerged in the field of gender studies and science and technology studies
and is now increasingly also used in other fields of research. When applied
to the topic of parenting, this theory assumes that material–discursive
boundaries are built between the different parenting practices, such as spe-
cific forms of housing, specific organizations of tasks, and specific modes of
thinking which in turn produce specific types of parents (Schadler, 2021).
During this study, I also worked on translating new materialist theories
into a methodological framework and methods for analysis, which I pub-
lished in Qualitative Research (Schadler, 2019). In this chapter, two modes of
analysis, tagging and referencing, will be explained in greater depth to show
how the findings of the study in Sexualities were established. However, some
personal and institutional circumstances (becoming a parent and changing
departments) also influenced the direction of the analysis of this paper, and
these contexts will get some attention in this chapter as well. Finally, new
materialist analysis works on eroding the boundaries between analysis and
writing. Therefore, this chapter discusses how writing for specific journals
influenced the analysis for the paper.

Analysis from the Perspective of New Materialism

I found New Materialism in the midst of a research crisis. I was analyzing
data with interpretative methods and felt that the findings were incomplete.
There was a “more” to the data that I could not grasp. New Materialism
provided me with a theory that emphasized the material conditions, physical
processes, and environments without overlooking meaning and discourse.
New Materialisms emerged from poststructuralist theories with an
emphasis on material differentiation processes (Dolphijn & van der Tuin,
2012). The most important feature of New Materialisms is that they are so-­
called anti-­dualist process ontologies (Braidotti, 2006). Anti-­dualist process
ontologies are theories that assume the world consists of ever-­differentiating
material–discursive processes (Barad, 2007), and within those processes,
material–discursive boundaries are built and transformed constantly. These
processes produce all the entities we can know and name, from humans, to
inorganic matter, to thoughts. New Materialisms distinguish themselves
from theories that set a specific entity a-priori to the process, such as mean-
ing, reasoning, structure, genes, or anatomy (Braidotti, 2002). They rather
track how these entities came into existence and how they are all connected
with each other. Therefore, they distinguish themselves from interpretative
theories, which are often used as resources for qualitative research. This
46 Cornelia Schadler

nexus of theories also includes a concept of humans who are produced in

these processes with a specific body and a specific reasoning (Barad, 2007;
Haraway, 1988). So, the bodies and reasoning of humans do not precede
their family practices, but family practices produce specific humans with
specific reasoning. Therefore, my study on polyamorous parents looked for
family processes that are differentiated by types of family processes, which in
turn, include specific humans, with specific tasks, that explain their family
practices in a specific way (Schadler, 2021).
This shift in theoretical perspective also resulted in a modification of the
analysis methods I used. When I started my research with these theories, there
was hardly any translation of new materialist theories into analytical qualita-
tive methods. After this, my work was also dedicated to develop a New
Materialist Ethnography (Schadler, 2019). Because qualitative tools of analysis
focus on condensing unstructured data (content analysis) or extracting latent
meaning from the data (interpretative analysis), I needed to modify my tools
of analysis to focus on latent material–discursive processes that produce the
boundaries of parenting practices (Schadler, 2019). From a new materialist
perspective, analysis is a process of intra-­action with the data, which assumes
that processes of analysis co-­produce the analyzing researcher and the data
(Barad, 2007). This type of analysis refers to Barad’s (2007) concept of exteri-
ority within, which theorizes the configuration of an outside position within a
process. Temporal cuts assign specific positions to an entity (e.g., the research
subject), and they consequently define a researcher as the creator of the whole
research process and the object as the focus of the process. Theoretically, how-
ever, all entities (e.g., the researcher, the research object, and the method) are
established simultaneously, although they seem temporally apart and separate
entities. Therefore, the production of results happens within a conglomerate
of processes, which become hidden when a research result appears.
Although theoretically the researcher is not the sole producer of the anal-
ysis, empirically, because of our production in these processes of exteriority
within, researchers are in the position to experience the findings they make
because of their actions. In the current world, researchers are deemed
responsible and held accountable for the findings they bring forward.
This way of analyzing may sound strange to readers new to this theory.
The advantage of this perspective is that the analysis is not looking for single
drivers, such as meaning, or rational reasoning, or a specific structure in the
data. None of these entities are considered the foundation of the world’s
becoming or human behavior. Nevertheless, all these entities and more can
and will be identified in the process of analysis in very specific combinations
(Schadler, 2019), while none of the entities are considered causing or deter-
mining the other entities. Therefore, New Materialist Ethnography helped
me look more openly to the data than other modes of analysis.
 Writing a New Materialist Ethnography on Polyamorous Parents 47

On the foundation of New Materialism, I built three modes of analysis:

tagging, referencing, and rebuilding worlds as integral methods of the meth-
odological framework of New Materialist Ethnography. Important to all
methods of analysis is a new materialist definition of boundaries. Boundaries
are the borders of all entities, which can be the limits of an object as well as
the limits of a concept. The boundaries are established in ever-­ongoing
material–discursive processes (Barad, 2007). These boundaries are revised
and reworked in every process; hence, boundaries are fluid, even when the
resulting entities seem fixed and stable over time (e.g., the boundaries of a
specific human). These boundaries can also shift depending on the perspec-
tive and focus. Research can investigate the intra-­action between a parent
and a child, perceiving both of them as distinct people with boundaries.
Research can investigate “families” that lump several people into a conglom-
erate of several people (and maybe also things and animals). Research can
also investigate parts of the body and draw a boundary within an entity that
is established as a whole in other research. New materialist research is con-
sidered a process of intra-­acting with the boundaries that are produced not
only in the field of interest but also during the research and analysis.
Accordingly, I define the analysis as a process of intra-­action with boundar-
ies and a process that is producing new boundaries while rebuilding the
researched worlds.
Tagging is a tool to analyze data on the manifest level, and it involves
responding to and intra-­acting with boundaries in the data. From the per-
spective of New Materialism, we already live in a world with existing bound-
aries, and these boundaries are intra-­acting with us (Barad, 2007). Going
through the data is such a way to intra-­act with already-­existing boundaries,
and analysis is a process to recognize and transform them. The researcher is
engaging with boundaries in the data, which is called intra-­acting, or kicking
back, as Barad (2007) also calls it. However, analysis is also not defined as
discovering (universal and ahistorical) boundaries but as a process of intra-­
action with the boundaries in the data, which transforms the researcher and
the boundaries in the data. For me and other new materialist researchers, it
is important to not call the produced boundaries codes (Augustine, 2014; St.
Pierre & Jackson, 2014), because the name “codes” would refer to a research
practice that is connected to a positivist (condensing content or discovering
boundaries) or interpretative (codes that represent latent meaning) vocabu-
lary and research framework, and both of these frameworks are considered
representational frameworks. The process was named tagging for method-
ological and theoretical clarity, because I was interested neither in a conden-
sation of the data (reductive coding) nor in the interpretative analysis of the
data (inductive coding). I used the denotation tags for the boundaries pro-
duced in the analysis of the manifest level (Schadler, 2019).
48 Cornelia Schadler

During the empirical praxis of tagging, everything in the researcher’s

mind and body that is produced in this moment can become a tag. Therefore,
strictly, there is no filtering out inadequate or irrelevant thoughts or feelings
because they are all produced in this process and may be a product from this
process. The practice of tagging consists of reviewing the data quickly and
associating the data with words, sentences, or figures that come to mind
without judging what comes to mind or body (a thought, a name, a feeling,
or a bodily perception). I usually do this process in a qualitative data analy-
sis (QDA) software program. I used [Link] for my analyses until 2018, at
which point I switched to MAXQDA when I started using a MAC laptop (in
my opinion, MAXQDA worked better with MAC), but one could use any
software that lets them annotate data or pen and paper.
A second tool of analysis I developed, which I used before, next to, or
after tagging, was referencing. Based again on the theoretical concept of
exteriority within (Barad, 2007), it is again, in theory, defined as a process of
analysis, where in intra-­action with the data, the findings and the researcher
are formed. I developed this tool to look at the latent level of the data, but it
is not defined as interpretation of the data because, in theory, no active
entity is doing the interpretation and looking for meaning. Rather, the
researcher is transformed in the process of analysis with new knowledge,
which presents itself in the mind and body of the researcher. Again, for
methodological clarity, the term “interpretation” was not appropriate any-
more, and the term referencing seemed to fit the process better. The reconfig-
ured knowledge is a reference to the process of analysis after all.
Barad (2007) uses the term “sedimented history” to illustrate latent
boundaries in material data:

As the rings of trees mark the sedimented history of their intra-­actions

within and as part of the world, so matter carries within itself the sedi-
mented historialities of the practices through which it is produced as part
of its ongoing becoming—it is ingrained and enriched in its becoming.
(p. 180)

In the process of analysis, data and findings are defined as such entities that
carry the sedimented historialities of its production. This concept of sedi-
ment history in the materiality of the data also allows for a concept of hid-
den or latent processes that are engrained in a specific product such as data.
The process of collecting data (interviewing and observing) is actualizing
these boundaries, and the processes of making observation protocols or
transcripts are redefining the materiality of the data again. In this data, we
find the sedimented histories of the processes of research itself and the pro-
cesses of the world the researcher and the researched are a part of. Analysis
 Writing a New Materialist Ethnography on Polyamorous Parents 49

is engaging with the data by asking for these sedimented processes: “What
are the processes that could produce a specific boundary in the data?” By
doing so, the researcher is already a part of a joint transformation process
that is producing knowledge. It is not about interpretation—finding latent
meaning in the data and words for this hidden meaning—but about intra-­
acting with or becoming with the data and gaining knowledge in the process.
Because all entities, including the data and the researcher, are made in the
differentiation processes of the world, references to their boundaries are
possible. I called this process referencing (instead of interpretation) because
in this process, references to the latent boundaries become manifest.
In the practices of analysis, referencing is looking for latent processes by
going to a piece of data sequence by sequence and asking, for example,
“which material-­discursive processes need to happen to make this piece of
data possible?” The result is a memo/note I usually attach in the QDA soft-
ware, which includes all the processes I can identify.
However, analysis does not stop at tagging and referencing because I con-
sider writing or illustrating the data for a talk or publication as a part of the
analysis. The processes and boundaries produced in tagging and referencing
are revised again when rebuilt into a piece of writing or an illustration of
these boundaries. On the one hand, I consider other components of the
practices of dissemination such as the required writing styles of journals or
publishers, the writing and publishing habits of a specific scientific commu-
nity, or a researcher’s writing education as additional players that intra-­act
with the data and co-­shape the written or illustrated result. On the other
hand, the process of writing or illustrating itself is rebuilding the pieces of
analysis into a world that inhabits the boundaries and processes and that
carries specific atmospheres of a specific environment, which may be not
directly expressed at the manifest level. Research therefore also makes and
shapes researched worlds.
Next to these procedures of analysis, from the perspective of New
Materialism, there are other conditions for research that are also a part of the
processes of analysis, such as time constraints for the research, team struc-
tures or career rationales, and many more. I theorize that they are a part of
the analysis but are excluded in the productions of findings so that the out-
come of the research can be related more clearly to their producing entities,
which are methods and researchers. In this chapter, I include some of these
conditions to show how these conditions may also shape the analysis.

The Project: From Chosen Family to Polyamorous Parents

The project on parents in consensual non-­monogamous relationships had
three major purposes: (a) to develop a theoretical notion of family from the
50 Cornelia Schadler

perspective of New Materialism (Schadler, 2016), (b) to configure a method-

ology (Schadler, 2019), and (c) to research families that differ from each
other as much as possible. I collected data on various forms of relationships
(non-­monogamous relationships, couples, nuclear families, humans with
pets, nomads, singles, and roommates and workmates). I started with one
form of relationship, and after a period of data collection and analysis, I
moved on to the next form of relationship or family. This process led me to
a focus on non-­monogamous relationships and families early in the project,
which was unexpected and initially unintended. The focus on polyamorous
relationships and families was highly influenced by the interest in this work
by the scientific and polyamorous community. When I issued the call for
interviews, many more people than usual wanted to be interviewed (although
I assumed the number of people living polyamorous is smaller than, for
example, heterosexual couples or people living with pets). Also, when I men-
tioned this group at conferences, the interest seemed to be higher in this
group than in the other types of family. I was soon invited into research
networks focusing on consensual non-­monogamous relationships, and these
collaborations led to more publications on the topic than on other forms of
chosen family. Because my previous research already had a focus on parent-
hood, I also soon had a specific interest for the parents in my data pool,
which included parents that seemingly live in heterosexual standard families
with an additional secondary partner separated from family life, but also
had networks of three, four, or five people raising children together as equal
partners.
Although this chapter focuses on the analysis of the data on polyamorous
parents, it is important to note that this analysis was co-­produced by the
interest of the community researched and the scientific community as well.
From a new materialist perspective, the analysis is embedded in contexts
such as funding and communities, which are usually excluded from the pic-
ture in the written paper.

Tagging Data
After the call for interviews, I interviewed 18 individuals from 13 polyam-
orous networks in which at least one member had children (Schadler, 2021).
The interviews were accompanied by additional data collections, such as
podcasts, books, or websites mentioned by the interviewed people. I initially
stored this data in [Link] to break them down into smaller chunks for
analysis and to get a sense of the themes that are discussed in the data. The
first step of analysis in my work on polyamorous parents was to tag the data.
Because the process of tagging is theoretically framed as a process of trans-
formation for all involved entities, there is no need for a specific justification
 Writing a New Materialist Ethnography on Polyamorous Parents 51

for the identified tags. If there was a rationale or a motivation for a specific
tag, I sometimes wrote a note to capture this rationale. From the perspective
of New Materialism, however, this rationale is also a product of the process
of tagging and not a reason for a tag. Everything that is known or felt by me
can become a tag and is included at this stage, even when it feels unclear or
unreasonable. Therefore, tags are defined as co-­built categories rather than
the product of active and intentional interpretation of data. However, the
developing tags are not completely contingent on the analysts’ individual
intra-­action with the data, but the theory is that the intra-­action with bound-
aries brings them to the surface.
The actual process of tagging consisted of reading through the data and
tagging everything that comes to mind at a certain place in the data. For exam-
ple, as depicted in Figure 3.1, the following quote was tagged with “parents’
free time,” “separation primary/secondary,” “children’s own space,” “separated
spaces children and parents,” “poly-­friendly house,” “children old enough to be
alone in room,” “hiding poly relationship from outsiders,” “economic resources
for space,” “freedom from care work,” “do children really stay in their rooms?,”
“relationship hidden from children,” and “feeling: too easy.”

Then it was parents’ free time and the children went to their room.
Whether they played or slept was their concern. And so they did their
thing and we did our thing. The generous space [of the house] and the
spatial layout made this easy.
(Leo, poly-­nuclear parent)

[SEE FIGURE 3.1 at eResource: Tagging].

The last two tags were accompanied by notes that discussed emerging feel-
ings of doubt: Was it really that easy to send kids to their rooms at a certain
time? Do they really need nothing from their parents, or does the mother
take up the caring tasks while Leo spends time with his secondary partner
(outside of the house or in another part of the house)? Therefore, I did not
take Leo’s words for granted, and I recognized that there were hints that
separate times and spaces were needed to live polyamorous and have chil-
dren with only one partner.
From the tags, I concluded that with the separation of time and space,
there was a separation of parenthood and relationship time with a second-
ary partner. The narration of “parents’ free time” was a rationale and maybe
also a specific performance of keeping times, spaces, and relationships sepa-
rate from each other. Later, I named these parenting practices poly-­nuclear,
because parents were simulating a nuclear family for their children and liv-
ing their polyamorous lives in secret or separate from the primary family.
52 Cornelia Schadler

This name came to me when I arranged and sorted the produced tags visu-
ally in [Link]. After I tagged the available data, I revisited the tags either
by looking through the tags or by visualizing them by transferring them to a
network view in [Link]. Figure 3.2 shows an example of this, in which I
drew a visual network of tags for poly-­nuclear families. As seen in Figure
3.2, the tags are in rectangular shapes, and the network view allows me to
move them around and create relationships between the rectangles. This
often helps me to see a pattern or structure (boundaries) in the data. In this
case, I saw that the issues of separating time, space, and relationships were
tagged, especially in cases where parenting was the task of the primary part-
ners. The worlds of the people involved in these relationships seemed to be
structured by these separations, whereas other polyamorous networks were
not tagged that much with words or feelings that hinted at separating times
and spaces that clearly, or the lines of separation ran in another direction.

[SEE FIGURE 3.2 at eResource: Network of Tags].

It became apparent to me that the processes of parenting in polyamorous

networks varied a lot. There were parents who basically lived like monoga-
mous couples with children, but they had other secondary partners who were
not involved in parenting. In these cases, parenting was done solely by the
primary partners. These narrations seemed to differ clearly from other cases
where more than two parents were involved in caring for children and where
there was less of a hierarchy between the partners in the relationships.
I had identified a difference between poly-­nuclear parenting and the two
other forms of polyamorous parenting practices already in this round of
tagging. The parents talked about activities that ensured that the secondary
partners were separated from their lives with the children, such as meeting
the secondary partners at other places or only at specific times of the day or
month. However, I could not grasp the boundary between hierarchical and
egalitarian polyamorous parenting at this stage of research. At this stage,
there were only two types of parenting that could be clearly distinguished:
(a) the parents who enacted activities that simulated nuclear families; and (b)
the parents who shared caring practices among three, four, or five partners,
who all were defined as parents or co-­parents. I tagged activities, such as
“sharing care work,” “negotiating care work,” and “freedom from care
work” for both types of parenting. The boundary between hierarchical and
egalitarian parenting was established later in the process I call referencing.

Referencing Data
After a round of tagging and the establishment of some first boundaries, I
started to reference the data. To reference latent material–discursive pro-
cesses, I looked for the processes that are a part of this specific text, the
 Writing a New Materialist Ethnography on Polyamorous Parents 53

mentioned activity or entity. I searched for the boundary-­making processes

and entities I imagined were a part of the production of a specific utterance
in the interview or in a picture or text. I tried to build a world in which some-
thing like this (my data) could have happened. During this process, I usually
worked with my interview data. I went through the data mostly line by line,
but sometimes also paragraph after paragraph, and occasionally I focused
on short sentences or half of a sentence in the interview.
In my research project tagging and referencing could happen simultane-
ously (I immediately start the process of referencing after creating a tag) or
at separate times. Analysis can happen in circles, whereas referencing can
lead to more tagging and then to referencing again. However, for the paper
on polyamorous parents, the processes of tagging and references were rather
separate activities because I gathered a large amount of data in a rather
short period of time, and tagging provided me with an overview of the
boundaries in the data.
To illustrate the referencing process and its difference from tagging, I
return to the example quote from Leo (see Figure 3.1). During the tagging
process, I identified parts of the data that hinted at a separation of the
nuclear family and the parents’ activities with their secondary partners. The
referencing process breaks the data down into smaller chunks and dives into
possible configuration processes.
For the referencing process, I exported the data into a MAXQDA file, the
program that I was now using. Most of the tags were excluded so I could work
with a clean file to analyze the interviews again. I used the tags to export spe-
cific parts of the data that were interesting to me—for this example, all parts
that were tagged with “poly-­nuclear.” When I was starting the referencing pro-
cess, I already had the paper on polyamorous parenting in mind. I was already
interested in parenting types and differences in the organization of work and
care time. But now it was time to take a closer look at the data. For this paper,
I did not reference the whole interview but specific parts of the interviews that
discussed family organization. As illustrated in Figure 3.3, I edited the sections
of interest from the interviews by separating paragraphs into sentences or
smaller chunks. I also separated Leo’s paragraph on parents’ free time into four
sentences and attached a note for each sentence in the software. Figure 3.3
shows some of the references for the following comment from Leo: “Then it
was parents’ free time, and the children went to their room.”

[SEE FIGURE 3.3 at eResource: Referencing: Example 1].

I asked myself: “What are the processes that configured the boundaries in
this sentence?” and “Which processes made it possible for Leo to say this
sentence in the interview?” I now started to speculate about latent material,
physical, social, discursive, structural, legal, and individual processes that
54 Cornelia Schadler

could be present in this sentence. In a memo (see Figure 3.3), I wrote about
how I imagined how free time from family is established. Family is often the
time we spend our free time from work, but Leo also seems to be part of a
second kind of free time, a free time from family (and work), like another
sphere of life (it is not only work–family balance, but work–family–relation-
ship balance). So, I speculated a lot in the note about this other layer of time
and space, and I finally called it “privacy from privacy,” and how this second
level of privacy has to be built by separating time and space, similar to sepa-
rating work by time and space from family time. To separate the space, either
Leo or the family must “go away” for a certain amount of time. Here we see
the family moving out of the way. I doubted that sending the kids to bed
always worked and that it was happening that way for real (my notes circle
around my feelings of doubt for about half a page). However, it was also not
relevant for my findings if the narrated practice really happened as described,
but that Leo and his family are building a boundary, that many other fami-
lies do not have: that second level of privacy. I found this second level of
privacy again in other poly-­nuclear families, but not so much in the families
with three or more present parents.
Asking for the processes producing this account, I configured a world in
which children, primary partners, and secondary partners could also be
separated in different spaces. There are activities to separate people and
their functions from each other. The parents might sit at the dinner table
with their children or spend time with them in the living room. I now made
pages of references (analytic text as shown in Figure 3.3) on the separation
of rooms, how the separation of rooms and bodies, historically, became a
thing, and how architecture is reconfiguring this concept of the separation
and collectivization of family members. I made a list of parenting practices
that might lead to the situation that the children do as they are told (stay in
their room). I slowly configured a world in which material environments—
rooms, tables, and furniture and their histories—have their contribution,
but also discourses and norms about families and parenting practices, next
to concepts about childhood and skills at a certain age. These processes and
entities produced a partial world in which parenthood could be set aside
spatially (sending children to their room and they also stay there), but also
habitually (now everybody can do what they want) and in the parents’ iden-
tity (I am not a father anymore, but I am now the partner). I called that
“privacy from privacy” or “privacy level II” in the process of referencing.
This latent material–discursive boundary showed itself in the process of
referencing reconfigured as a finding in the process of analysis. Without
referencing latent processes, I would not have been able to rebuild this
boundary in writing.
 Writing a New Materialist Ethnography on Polyamorous Parents 55

Rebuilding Worlds: Starting to Write and Pausing Research

I started to write the first version of the paper halfway through the referenc-
ing process with the Journal of Family Issues as the target in mind. However,
at this point of analysis, another personal issue became entangled with the
process of analysis and writing. When I wrote the paper, I was already preg-
nant, and other papers had to be finished more urgently before the birth of
the baby. After I returned from leave, I finished the empirical analysis of the
project, but keeping up with my publications was hard due to severe sleep
deprivation. Additionally, after the project’s running time finished, I applied
for and was offered a (permanent) teaching position at another department.
Although the modes of analysis remained unchanged, I am sure that the
second half of the analysis for the paper and the writing of the paper would
have been different without these events. I was looking at the data with fresh
eyes. I had less time, but I was also less stressed and under less pressure
because I now had a permanent academic position.
Then, in 2020, Christian Klesse put together a group of people already
working on polyamorous parents for a Special Issue in Sexualities. I gave my
paper a last chance to be published and to be part of the Special Issue. I kept
my tags untouched, but I started over with referencing in MAXQDA. I had
already a rather clear understanding of the poly-­nuclear model, but I also
revisited these parts of the data (as visible in Figure 3.3 and illustrated in the
section on referencing). In Figure 3.4, I illustrate how analysis produced a
boundary between hierarchical and egalitarian parenting practices, for exam-
ple, when I revisited the part of the interview where this quote was located:

This, for example, is our joint calendar [shows something on the display of
her smartphone]. This is my calendar, this is Flora’s calendar, this is our
joint calendar, this is Fabian’s calendar, and this is when the child is with us.
(Fiona, an egalitarian parent)

[SEE FIGURE 3.4 at eResource: Referencing: Example 2].

When I revisited the data with fresh eyes after a pause in analysis, I saw dif-
ferences in routines between groups of people with a hierarchical parent
structure and egalitarian parents. Groups of parents who had a clearer divi-
sion of tasks developed routines that did not need any negotiation anymore.
It seemed clear who did what and when, even when they had to find their
division of tasks through processes of trial and error in the past. In the case
just presented, tasks are constantly renegotiated; however, there are routin-
ized modes of negotiation and communication. I saw a boundary between
56 Cornelia Schadler

the worlds of people with routinized tasks and people with routinized nego-
tiation of tasks.
The first draft of my manuscript, written before the pause in research, did
not include the boundary between hierarchical and egalitarian parenting
practices. Revisiting the data with fresh eyes now allowed me to intra-­act
with another boundary and to build differences that I had not put in words
before. When I revisited my references, this boundary was already there in
some descriptions of feelings or doubts or hints at something I could not
express in words. The pause in research might have allowed me to reconfig-
ure this boundary in a written ethnography of different parenting practices.

Revision of the Paper and Analysis

The larger part of my published paper focused on three types of parenting,
and the last section in the paper illustrated that families could switch between
those family types, sometimes even within hours. This section emphasized
that the involved people enacted a certain parenting practice but did not
(fully) control their actions with a specific type of identity. Consequently,
their rationalizations changed in specific situations when they switched their
parenting mode. Reviewers of the paper requested that I strengthen this part
in the paper to emphasize the new materialist theoretical perspective that
perceived the involved parents as people who enacted their parts in complex
processes. I revisited my data again and more thoroughly described these
aspects of the theoretical perspective guiding the analysis in greater detail.
Luckily, the editors allowed me to add another 500 words to the text, which
was only possible because one invited paper for the special issue was not
completed, and the space left was divided among the other authors. Table 3.1
shows a comparison of a paragraph of the paper before and after the
revision.
The revision made the transitions from one type of family to others for a
specific polyamorous network more visible. It foregrounded that from the
perspective of New Materialism, parents enact a specific position in specific
processes, and their identities do not (fully) control these enactments. It
shows more clearly that there may be a specific identity as egalitarian poly-
amorous family, but the enactments of family types vary in specific situa-
tions, and polyamorous families become invisible at certain places and
reappear again transitioning to other places.
In this section, I want to illustrate that rebuilding the worlds I encoun-
tered and analyzed during my research is also a part of analysis. In the first
part of my paper, I draw rather clear types of family, while this part of the
paper questions the durability and stability of these types. How I draw the
 Writing a New Materialist Ethnography on Polyamorous Parents 57

TABLE 3.1 Comparison of Text before and after the Revision

Before Revision
While the above-­described parenting assemblages are stable for a longer time, a
closer analysis of the interviews shows that in specific situations, all parents
deviate from the style of parenting they identify with. For example, Roger
describes a family gathering at Rita’s parents. Relatives do not understand their
concept of family and parenting and seem to believe that Roger is the ex-­husband
and Robert is the new partner. The relatives perceive the polyamorous family in
the light of serial monogamy and heterosexual relationships and cannot perceive
anything outside of these concepts. Rita became tired of explaining; therefore,
they just let their relatives believe what they want. Therefore, they become a
(poly-­nuclear) family that is simulating a heterosexual partnership.
After Revision
While the above-­described parenting types are stable for a longer time, a closer
analysis of the interviews shows that in specific situations, all parents deviate
from the style of parenting they identify with. For example, Roger describes a
family gathering at Rita’s parents. Relatives do not understand their concept of
family and parenting and seem to believe that Roger is the ex-­husband and
Robert is the new partner. Even Roger’s partner, who is not involved in parenting
practices, fits in this picture. The relatives perceive the polyamorous family in the
light of serial monogamy and heterosexual relationships and cannot perceive
anything outside of these concepts. Rita became tired of explaining; therefore,
they just let their relatives believe what they want. Consequently, they become a
(poly-­nuclear) family that is simulating a heterosexual partnership. It is important
that from a new materialist perspective within these hours they are a serially
monogamous couple sharing parenthood. They enact this family model, and
their (secret) egalitarian polyamorous relationship does not come to matter in
this moment. It is nonexistent in this moment and is reactivated on the way home
by changing (material and discursive) contexts. They become a part of practices
that rebuild the egalitarian parenting model on their way home from this family
gathering: for example, in the car, when they discuss and reflect on the family
event and then the upcoming week and family matters, and when they come
home to their specific living situation (all living in the same house across two flats
sharing all the work) where they enact their usual family routine again. Therefore,
when they go to a family gathering, they travel between two worlds in which they
are two different forms of family (Schadler, 2021, pp. 13–14).

difference between the types rebuilds the boundaries between the types of
families differently. Writing for a specific audience reshapes the data in
another very specific way. Further, in the case of this paper, writing and
analysis did not just take alternating turns but melted into each other
because writing pauses and opportunities guided my focus in analysis, and
analysis reconfigured the worlds I was building through words.
58 Cornelia Schadler

Conclusion
In my 2021 paper on polyamorous parents in Sexualities, I illustrated three
types of parenting practices. In this chapter, I revisited my enactment of the
processes of tagging and referencing to build the boundaries between these
types of parenting. I also discussed how other contexts such as the research
community, publishing conventions, and personal events influenced the
analysis. I tried to explore the world that allowed the analysis and the paper
to happen.
During this and other projects, I developed methods of analysis (tagging,
referencing, and rebuilding worlds) that translate new materialist theory into
a set of intra-­actions with data, one focusing on rather manifest boundaries
of the data (tagging) and one speculating on latent material–discursive pro-
cesses (referencing). These methods are two of many possible modes of
analysis from the perspective of New Materialism. They are a contribution
to research methods for when researchers want to focus on the production
of boundaries or on complex processes without foregrounding a specific
issue that drives these processes.

KEY WORKS GUIDING MY DATA ANALYSIS

Braidotti, R. (2006). Posthuman, all too human: Towards a new process ontology.
Theory, Culture & Society, 23, 197–208.
I was in the first year of my dissertation and was stuck with some preliminary
data (interviews with parents at the transition to parenthood). I had the feeling that
there was more to my data, but I could not reach it with the methods and theories
at hand. So, I gave up and got back to reading theory. Then, at that specific after-
noon, I read this article, which discusses the work of Donna Haraway, and I became
a new materialist researcher.
Barad, K. (2007). Meeting the universe halfway: Quantum physics and the entanglement
of matter and meaning. Duke University Press.
This book contains the figure of exteriority within, which was most influential for
the development of a New Materialist Ethnography and the methods of tagging and
referencing. It provides a thorough discussion of posthumanist process ontologies.
Froschauer, U., & Lueger, M. (1992) Das qualitative Interview: zur Analyse sozialer
Systeme. WUV.
I received a thorough training as a hermeneutic researcher by Ulrike Froschauer.
And although I do not use her method anymore for my research, it heavily influ-
enced my perspective on latent processes.
 Writing a New Materialist Ethnography on Polyamorous Parents 59

References
Augustine, S. M. (2014). Living in a post-­coding world. Analysis as assemblage.
Qualitative Inquiry, 20(6), 747–753. [Link]
Barad, K. (2007). Meeting the universe halfway: Quantum physics and the entangle-
ment of matter and meaning. Duke University Press.
Braidotti, R. (2002). Metamorphoses: Towards a materialist theory of becoming.
Polity Press.
Braidotti, R. (2006). Posthuman, all too human: Towards a new process ontology.
Theory, Culture & Society, 23(7–8), 197–208. [Link]
0263276406069232
Cooper, M., & Pugh, A. J. (2020). Families across the income spectrum: A decade in
review. Journal of Marriage and Family, 82(1), 272–299. [Link]
jomf.12623
Dolphijn, R., & van der Tuin, I. (2012). New materialism interviews & cartographies.
Open Humanities Press. [Link]
Few-­Demo, A. L., & Allen, K. R. (2020). Gender, feminist, and intersectional per-
spectives on families: A decade in review. Journal of Marriage and Family, 82(1),
326–345. [Link]
Froschauer, U., & Lueger, M. (1992). Das qualitative Interview: Zur Analyse sozialer
Systeme. WUV.
Haraway, D. (1988). Situated knowledges: The science question in feminism and the
privilege of partial perspective. Feminist Studies, 14(3), 575–599.
Reczek, C. (2020). Sexual-­and gender-­minority families: A 2010 to 2020 decade in
review. Journal of Marriage and Family, 82(1), 300–325. [Link]
jomf.12607
Schadler, C. (2016). How to define situated and ever-­transforming family configura-
tions? A new materialist approach. Journal of Family Theory & Review, 8(4), 503–
514. [Link]
Schadler, C. (2019). Enactments of a new materialist ethnography: Methodological
framework and research processes. Qualitative Research, 19(2), 215–230. https://
[Link]/10.1177/1468794117748877
Schadler, C. (2021). Ever more parents in polyamorous families: A new materialist
typology of parenting practices and division of work. Sexualities. Advance online
publication. [Link]
Smock, P. J., & Schwartz, C. R. (2020). The demography of families: A review of
patterns and change. Journal of Marriage and Family, 82(1), 9–34. [Link]
org/10.1111/jomf.12612
St. Pierre, E. A. S., & Jackson, A. Y. (2014). Qualitative data analysis after coding.
Qualitative Inquiry, 20(6), 715–719. [Link]
Williams, K. (2020). The decade in review. Journal of Marriage and Family, 82(1),
7–8. [Link]
4
READING BETWEEN
THE LINES OF AFTER DEATH
COMMUNICATION STORIES
Using Narrative Analysis to Make
the Implicit Explicit

Sara E. Hackett and Kate de Medeiros

In the spring of 2019, Sarah was a second-­year doctoral student enrolled in

Kate’s graduate course, Advanced Qualitative Research Design and Analyses,
at Miami University. As part of the course, the main assignment for students
was to re-­examine qualitative data that they had previously collected. The pur-
pose of this assignment was to apply a new qualitative lens to our data so that
we could not only practice using a different qualitative methodological tech-
nique but also appreciate the magnitude of ways that we could view and inter-
pret our rich, qualitative data. In addition to reanalyzing our data, we were
tasked with writing a methods and results section of a manuscript as a means
of encouraging us to publish an original manuscript.
For my class project, I conducted a secondary analysis of data that I col-
lected for a pilot qualitative descriptive study during the summer of 2018. In
that study, I combined the continuing bonds framework, a theory that
explores why and how living persons maintain connections to deceased per-
sons (Silverman & Klass, 1996), with the Hierarchical Mapping Technique
diagram, a diagram of three concentric circles meant to represent the closest
individuals in a person’s life (Kahn & Antonucci, 1980), with the specific
intent to examine the potential presence of deceased individuals in the social
convoys of bereaved persons (Stemen, 2020b). For the initial pilot study, I
used an abductive approach where my analysis and findings were guided by
both theoretical frameworks and the data itself (Peterson et al., 2021).
Although Hierarchical Mapping Technique-­related research largely argues
that deceased persons fall out of and are not part of surviving persons’ social
convoys (Antonucci et al., 2013; Carstensen et al., 1999; Fuller et al., 2020),
the findings from this study suggested that most of my participants viewed

DOI: 10.4324/9781003254041-5
 Reading between the Lines of After Death Communication Stories 61

deceased, self-­reported loved ones as key members of their present-­day social

network. Participants shared that the deceased belonged in their present-­day
convoy because they perceived that they continued to provide them with
important types of support. For example, several individuals discussed how
they received ongoing emotional support from the deceased when it came to
tasks such as decision-­making, while others talked about how tangible items,
like shared resources and assets, gave them the support they needed to con-
tinue to live their desired lifestyle despite the death of their loved one.
As I reflected back on this data and my findings, I remembered that sev-
eral participants talked about the continued support of the deceased by
sharing examples of After Death Communication (ADC) instances when
they believed that the deceased communicated with them through an action,
such as sending a sign, showing their presence in a dream, or hearing their
voice (Guggenheim & Guggenheim, 1997). I was intrigued by how my
study’s findings fit in the midst of the conversation about the dynamic nature
of grief because the grief literature points to a dominant cultural narrative
of bereaved persons needing to move on from people in their life who have
died (Breen & O’Connor, 2011). Consequently, I wondered if participants
felt compelled to couch their ADC narratives in a particular way. That is, I
questioned if people shared ADC narratives in a manner that suggested that
they were concerned about how they would be perceived by others for shar-
ing an experience largely viewed as peculiar and, in some instances and cul-
tures, taboo.
Drawing from my specific research interest, coupled with Kate’s expertise
in the realm of narrative analysis, I used this class project as an opportunity
to gain insight into how ADC narratives unfold. The objective was to inves-
tigate if (and if so, how) narrative patterns existed in the ways that bereaved
persons told stories about ADC experiences. Our belief was that through
identifying any potential narrative patterns in the way ADC experiences are
described, researchers and practitioners would be better equipped to truly
hear how bereaved persons perceive their life without the deceased. The
manuscript that resulted from our research and later published in Death
Studies found that even though new models of bereavement have surfaced,
which normalize continued connections between living and deceased per-
sons (Klass, 2006), bereaved persons continue to signal how they share their
ADC narratives that they are concerned about how listeners will perceive
them for discussing such experiences (Stemen, 2020a). We hoped that this
new knowledge may be able to assist with the development of resources and
supports that could help people cope with life after loss.
This chapter provides a behind-­the-­scenes look at how Kate and I worked
together during my time in her class to analyze my data using a different and
underutilized qualitative methodological approach, narrative analysis. In
62 Sara E. Hackett and Kate de Medeiros

particular, we will explain how we utilized a structural narrative model to

conceptualize stories for a secondary data analysis, describe how we coded
interview data using a narrative approach and the aforementioned narrative
model, and detail analytical tools available in Dedoose software to support
narrative analysis. Our hope is that by sharing about this innovative yet rig-
orous qualitative technique, more researchers will be encouraged to take the
next steps in exploring the deeper meaning behind what their research par-
ticipants share with them so that more people feel heard and understood.

Identifying Narrative Data

Although there are many definitions of what comprises a “narrative,” narra-
tives typically refer to stories or the process of reinterpreting and translating
past experiences from memory into stories through language (oral or written)
or symbols (e.g., art or dance) (de Medeiros, 2014; Polkinghorne, 1988;
Riessman, 1993). Whether the retelling of an event is “true” or “accurate” is
irrelevant. What is important is that the narrator has merged events in a way
to create something meaningful to them, which in turn becomes the focus of
narrative inquiry. In this chapter, the focus is on oral narratives as they are
introduced and told by research participants during qualitative interviews.
Unlike discourse, which broadly describes socially meaningful language use
such as words that comprise a grocery list as opposed to random words
(Maschler & Schiffrin, 2015), a defining feature of a narrative is emplotment,
which is the selection and ordering of certain events in the telling of the story
in a way that leads to a series of causes and effects (Becker, 1997). In an oral
narrative, the teller not only talks about something from the past but, in doing
so, also imbues meaning on the events and details that they have selected
(Riessman, 1993). The notions of meaning and emplotment are important
because they become central elements in the analysis of narratives.
We used Labov and Waletzky’s (1967) well-­known structural approach to
narrative analysis. This approach focuses on key syntactical elements within a
given story (Labov, 2001; Labov & Waletzky, 1967) and involves examining
narratives through their correspondence with particular components of the
story: (a) abstract orientation, (b) complicating action, (c) ending, (d) evalua-
tion, and (e) coda. Essentially, the narrative itself can be understood in terms
of what details are revealed and how they are told. How the story is introduced
is the abstract (e.g., “I’m going to tell you about the time I witnessed a terrible
fire”). The characters that are introduced form the orientation or the way in
which the speaker helps orient the listener to the setting (e.g., “My two cousins
Bill and Bob were visiting from Idaho and we decided to go to the park”). The
complicating action refers to the way in which various actions are revealed to
move the story forward (e.g., “There we were, minding our own business near
the ranger shed when suddenly we saw flames coming from the windows. We
 Reading between the Lines of After Death Communication Stories 63

then heard cries for help so we knocked down the door”). There can be several
complicating actions that lead to the story’s ending. The evaluation refers to
how the speaker makes sense of the story (e.g., “That day taught me never to
never be afraid to spring into action”). The coda points to how the speaker
transitions from the story to the present (e.g., “And that’s why I never leave
home without my cell phone in case I need to call for help”). How we used this
framework to analyze our data is discussed in further detail in the subsequent
sections of this chapter.

Our Narrative Approach

We drew from data that was collected from a larger qualitative descriptive
exploratory pilot study. As part of the study, I conducted 18 interviews with
individuals who experienced the death of a self-­identified loved one (Stemen,
2020b). Given that previous studies that utilized the Hierarchical Mapping
Technique diagram were generally exclusive of deceased persons, the p ­ urpose
of the pilot was to test my new methodology that blended the Hierarchical
Mapping Technique diagram with the continuing bonds ­framework (Klass
et al., 1996). My exploratory research objective was to learn about whether
(and if so, how) participants considered their close deceased loved ones as
members of their present-­day social network.
Using the transcripts from the exploratory pilot study (outlined previ-
ously), we conducted a secondary narrative analysis. We selected a narrative
approach because it provides researchers with a lens to explore how stories
are told (Riessman, 1993) in addition to what stories say. Furthermore, nar-
rative analysis examines the deeper meaning behind the specific and pointed
choices that individuals make in deciding which parts of a story to share
with listeners and which parts to leave out (de Medeiros, 2014). Therefore,
this study gave us an opportunity to utilize a systematic method that would
allow us to make implicit messages shared by research participants explicit
through the identification of specific narrative markers, a process that will be
shared in more detail in the rest of this chapter.
Our data analysis relied on the use of a qualitative deductive approach
(Azungah, 2018). When using a deductive approach, the researcher is pri-
marily guided by a theoretical framework to develop their findings, as com-
pared to having findings that are driven purely by their interpretation of the
data (Azungah, 2018). Therefore, to utilize this technique, our first task was
to determine how we were going to conceptualize narratives in the data.
After reviewing the narrative literature, we chose Labov’s (2001) model of
natural narrative. As outlined previously, Labov explains that generally sto-
ries are told in a particular and chronological manner, from abstract to coda,
though he noted that not all components must be included to determine
which stories do or do not qualify as narratives.
64 Sara E. Hackett and Kate de Medeiros

With Labov’s (2001) definition of what a narrative is in mind, I started the

narrative analysis process by printing out hard copies of the 18 participant tran-
scripts and looking for any stories that participants shared during their respec-
tive interviews. I carefully read each transcript line by line to identify any
narratives in the data. To be systematic in my approach, I marked narratives
that I noticed contained at least three elements of Labov’s (2001) model of nat-
ural narrative. On printed-­out copies of the transcripts (we were required to
work with hard copies in the class), I drew attention to the narratives in the data
by penciling in bracket marks to those sections of the transcripts. Afterward, I
shared those transcripts, complete with my handwritten notes, with Kate for
review and discussion. Given that this was my first time using narrative analysis,
and this approach requires more objective than subjective interpretation, it was
imperative to draw from Kate’s expertise to verify that I had identified passages
that qualified as narratives, according to Labov’s framework. If narratives were
missed or incorrectly identified, this would negatively impact our ability to elicit
the study’s findings and say anything about how ADC narratives unfold.
Precision was key. Therefore, Kate read through the 18 transcripts, along with
my notes, to verify that all narratives were accounted for in the data so that we
could start a new dataset that consisted of only narratives. Kate did not identify
any new narratives at this stage of the analysis.
Now that we had a new dataset, the next step was to focus on narratives
that contained our specific area of interest, ADC experiences. This dataset
started with all stories from all of the study’s 18 participants. However, after
carefully rereading the participants’ narratives, I found that only 10 tran-
scripts contained information about ADC experiences, and so we only
focused on those. At this point, we had a more focused dataset that exclu-
sively showcased ADC narratives.
To move forward with the narrative analysis of ADC experiences, I
uploaded the 10 transcripts to Dedoose software. We chose to use Dedoose
for this study because this was the software that I used to analyze the data
for the study (Stemen, 2020b) from which this secondary data analysis was
drawn. Further, we both found Dedoose to be intuitive to use for coding the
data and inclusive of all the analytical tools that we would need to carry out
our research (described later in this chapter). Furthermore, it was accessible
for a project that required the involvement of multiple researchers, allowing
Kate and me to work simultaneously on the data file. Another important
feature of Dedoose is its low cost (around $15 per month) compared to
other qualitative software packages. In addition, users are not required to
commit to a long-­term subscription or site license, and they can cancel their
subscription at any time, with their data retained in the software.
At this point, I created a preliminary codebook that contained only codes
that mapped onto Labov’s (2001) model of natural narrative (e.g., abstract or
orientation) and their respective definitions. Following this, I read each ADC
 Reading between the Lines of After Death Communication Stories 65

narrative one by one, multiple times. Then, I read each isolated ADC narrative
line by line and began to ask myself questions, such as “Which elements of
Labov’s (2001) model appeared in the narrative (if any)?” and “How do I know
that this element is truly one element or another one?” If an element from
Labov’s (2001) model was present in the narrative, I used the coding function in
Dedoose, shown in Figure 4.1, to highlight that section of the text and isolate
it as a specific data point that spoke to a particular code in the codebook.

[SEE FIGURE 4.1 at eResource: Example of Coded Text in Dedoose].

Again, narrative analysis at this stage requires acute precision. To be able to

address how stories unfold structurally, which was one of our main study
objectives, one must accurately differentiate the elements that appear in the
narrative. Although some elements were easy to identify, such as the o ­ rientation,
other elements had more overlap that made them trickier to differentiate.
For example, at the start of the analysis process, I remember spending a
whole night coding data only to realize the next day during class, when I was
having a discussion with Kate about my progress, that I was blurring the lines
between the narrative’s resolution (i.e., ending) and evaluation (i.e., main take-
away). Admittedly, as an emerging qualitative scholar who had routinely uti-
lized approaches such as thematic analysis, a technique much more open to
interpretation, I struggled to lean into the precision and uniformity between the
ADC narratives that I needed. This was an area where Kate’s knowledge was
invaluable. She shared tips on what to look for in narratives, along with key
words and phrases that could help me tell the difference among the elements of
Labov’s model. For example, she challenged me to pinpoint the final sentence of
the complicating action—an element that is nearly present in all narratives—to
find the resolution. From there, she shared how phrases like “since then,” “mov-
ing forward,” and “ultimately” are likely to signal the narrative’s evaluation.
After recoding the data, Kate and I met several times where she challenged
me to defend each code I applied to the transcripts. During our meetings, we
would read aloud through the interview transcripts together, and I would
review the codes that I applied to the different parts of the identified narratives.
As part of this process, Kate asked me to explain why I coded each section of
text in the particular way that I did. Although this was a tedious process, it
helped me to learn the data inside and out, which resulted in increased confi-
dence as I entered the next phase of the narrative analysis process.
Once all the ADC narratives were coded to account for the elements of
Labov’s (2001) model of natural narrative, I used several Dedoose functions
to determine structural patterns in the ways that participants described their
ADC experiences. Again, this played to our major objective of gaining
insight into how narrative stories unfold and if storytellers signal anything
to suggest a cultural awareness of how ADC experiences are perceived. To
66 Sara E. Hackett and Kate de Medeiros

meet this objective, the first step of the process was to review which of
Labov’s elements appeared in the ADC narratives. To assist, I utilized the
code presence tool in Dedoose. As illustrated in Table 4.1, the code presence
tool parses out each code and provides information about whether a particu-
lar code appears within an individual interview transcript. If a code was
applied in an interview transcript, the Dedoose code presence output shows
a red box with the number one inside of it. If a code was not used, the box
in the output does not turn red and contains a zero.

[SEE TABLE 4.1 at eResource: Output from Dedoose’s Code Presence Tool].

After I used the code presence tool to see which elements of Labov’s (2001)
model of natural narrative appeared in the ADC narratives, I then employed
Dedoose’s code application tool. The output from the code application tool,
presented in Table 4.2, is helpful because it provides the researcher with a
numeric count of how many times each code in their codebook is utilized
within each data point (i.e., interview transcript) and across their entire
dataset (i.e., all interview transcripts).

[SEE TABLE 4.2 at eResource: Output from Dedoose’s Code Application Tool].

By using Dedoose’s code presence and application analytical tools, Kate and
I were then able to review the codes from the codebook and make compari-
sons among the codes that were found within the ADC narratives that par-
ticipants shared during their respective interviews. For example, using these
outputs, we could easily identify which elements of Labov’s (2001) model of
natural narrative were routinely applied to sections of the text and which
elements did not appear in the data. We found that by and large most ADC
narratives contained several of the elements of narrative that Labov (2001)
described; however, we also noted that very few participant narratives were
inclusive of an abstract (i.e., hint that a story will be told). Rather, the ADC
narratives tended to include and began with an orientation (i.e., introduc-
tion of setting and characters). Though again, Labov (2001) is clear in his
model that not all elements need to appear in a story for it to be considered
a narrative, this structural finding appeared to suggest that participants were
mindful of how they started to share information about ADC experiences
with listeners because they jumped right into their ADC experiences as
opposed to providing lengthy background information that would suggest
to a listener that an ADC story was about to be shared. Building from this
finding, we determined that the next step of our narrative analysis would be
to investigate what was unique about how participants shared their ADC
experiences.
 Reading between the Lines of After Death Communication Stories 67

Similar to before, I began this phase of the narrative analysis by rereading

the ADC narratives line by line in Dedoose. As I re-­examined the narratives and
looked at the codes that were applied from Labov’s (2001) framework, I began
writing memo notes in a journal that I was required to keep for the course and
also in Dedoose. The purpose of writing the memo notes was to document my
thought process and have transparency about how I was interpreting the data.
Here, I used the structural findings that we identified previously and thought
more deeply about why the ADC stories were told in a particular way. Although
our structural findings required systematic objectivity, this part of the process
required reflexivity and reflection. My memo notes documented observations
that I made about the stories shared by individual research participants as well
as observations that seemed to cut across participants and suggested a specific
pattern for how ADC narratives tended to unfold in the study’s sample. For
example, in the memo note shown in Figure 4.2, I recorded an observation
about how participants seemed to value sharing their feelings while storytelling.
Along with documenting my observations, I also used the memo notes to record
my theories about why I observed what I did and why the ADC stories unfolded
in a particular manner. Using the example of the previous memo note I wrote,
I theorized that sharing feelings while storytelling both enhances understanding
between the storyteller and listener and cues the listener into the personality,
values, and personal beliefs of the storyteller.

[SEE FIGURE 4.2 at eResource: Example of Memo Note].

After I carefully reread each ADC narrative, I noticed an overarching pattern

that spoke to how participants revealed their ADC experiences. In brief, I noted
that most of the participants commented on the likelihood and believability of
their ADC experience. To capture this idea, I added the code “Qualifier” to
account for this phenomenon within the codebook. I then went back to review
the other transcripts again and recoded all the ADC narratives to capture all of
the sections of text where qualifiers appeared within participants’ stories. Like
before, I looked at Dedoose’s code presence and application tools to see how
often the “Qualifier” code was used within and between participants’ interview
transcripts. Using these tools, I found that the code “Qualifier” was consis-
tently applied within and between participants’ ADC narratives. To take the
narrative analysis a step further, I started looking at the specific placement of
the code within the ADC narratives. During this process, I discovered that the
code “Qualifier” often appeared in different places throughout the ADC narra-
tives. As the code of “Qualifier” permeated throughout the ADC narratives,
Kate and I identified the concept of qualifying as a major, overarching theme
in our findings that described why the ADC narratives were told in a specific
manner by participants—as an outlet for participants to signal perception.
68 Sara E. Hackett and Kate de Medeiros

The next step of our narrative analysis was to look at all the sections of text
that spoke to the code “Qualifier” and to determine if any smaller ideas or
patterns existed within the concept of qualifier. To complete this step of the
data analysis, I looked for key phrases (e.g., “I’m sure of,” “I don’t know”) that
signaled the participants’ perception of their ADC experience. In doing so, I
interpreted that the sections of text with the qualifier code applied to them
were generally presented in one of two ways. The participants either verbal-
ized that (a) their ADC experience was real or (b) their ADC experience was
inexplicable. To account for these two separate belief systems, I once again
adapted our study’s codebook. For instances when participants voiced any-
thing that resembled disbelief or judgment about their experience, I continued
to use the “Qualifier” code. To capture instances when participants voiced
their belief of their ADC experience, I added and utilized a code that I named
“Certainty.”
After modifying the codebook, I went back through the interview tran-
scripts in Dedoose, re-­examined the ADC narratives line by line and recoded
the data to detect any passages that expressed the new code of “Certainty.”
As before, I used the code presence and code application tools in Dedoose to
examine how the codes were applied to the data. Adding to these tools how-
ever, I then used the packed code cloud tool and the code co-­occurrence
tool. The packed code cloud tool, presented in Figure 4.3, is useful because
it provides a visual representation of how codes are applied in a dataset. For
example, if a code was applied heavily, that code would appear very large in
the word cloud. Using the packed code cloud, I could quickly observe how
the “Qualifier” code was used much more frequently than the “Certainty”
code. This data provided further evidence that the idea of qualifying should
be a major theme of the study.

[SEE FIGURE 4.3 at eResource: Output from Dedoose’s Packed Code Cloud Tool].

[SEE TABLE 4.3 at eResource: Output from Dedoose’s Code Co-­Occurrence Tool].

The second tool that I utilized was the code co-­occurrence tool. Using this
tool, I learned which codes were routinely applied within the same sections
of the interview text. For example, as can be observed in Table 4.3, I could
see that the codes “Qualifier” and/or “Certainty” appeared in every ADC
narrative. Additionally, the tool also helped us identify structural patterns in
the way ADC narratives were told by participants. For example, the output
of the code co-­occurrence tool illustrated that the code “Certainty” often
coincided with the codes “Evaluation” and “Coda.” By looking at the spe-
cific placement of the “Certainty” code in the narrative, we learned that par-
ticipants who voiced strong beliefs that their ADC experience was real most
often shared this sentiment at the end of the telling of their ADC narrative.
 Reading between the Lines of After Death Communication Stories 69

To illustrate this pattern, sections of the text that were coded as “Certainty”
are italicized within an excerpt from one participant who said:

[Husband’s name] and I used to come to basketball games. And there was
one day where I was just standing there, and I just felt there was some-
body standing in back of me. And I could just feel a warmth coming. I
felt a calmness coming over me. There was nobody in back of me, but I felt
that. I just feel it [his presence] sometimes.

The code co-­occurrence tool also helped us confirm that the “Qualifier”
code appeared throughout the different elements of Labov’s (2001) model of
natural narrative, though often at the beginning and ending of the ADC
story. To illustrate this pattern, sections of text that were coded as “Qualifier”
are italicized within an excerpt from one research participant who stated:

I was sitting quietly, and a couple odd things happened. And I looked up. We
had a clock in our bedroom that was a gift when we got married. And it was
a battery clock. And the batteries were worn out … it was not keeping time.
And this was 5:30 in the morning, really early. And I looked up and the clock
said the time on it was 2:55. And that was the time she died. So, what made
me look at that clock at that moment? I don’t know. Maybe it’s coincidence.

Now that we had further identified what was structurally happening within
the ADC narratives that the research participants told, we could investigate
more closely why these participants were sharing their stories in such a way.
As shown in the ADC narrative examples provided previously, participants’
messaging about their ADC experiences was largely implicit. Participants
appeared to tiptoe around, providing definitive statements about their ADC
experiences. This was well documented in the “Qualifier” and “Certainty”
codes that were applied across many of the ADC narratives. Guided by nar-
rative analysis, which prompted us to look closely at the specific application
of these codes, we began to think about why these beliefs were regularly
presented in the ADC narratives shared by the participants. As part of this
process, I reread those areas of the text to identify the suggested root of
these phrases. For example, looking back at the participant excerpts included
in this chapter, I noted that the phrase “maybe it’s coincidence” appeared to
signal participants’ insecurity about the ADC they experienced, whereas the
phrase “There was no one in back of me, but I felt that” appeared to suggest
confidence in the ADC experience. Therefore, Kate and I determined that
the root or function of both of these phrases was to voice a perceived belief
in the ADC experience. This then begged the question: Why did participants
feel compelled to share their perceived validity of their ADC experience?
70 Sara E. Hackett and Kate de Medeiros

By noting the implied emphasis that participants placed on whether they

believed their experience, which became explicit through the narrative analy-
sis they conducted, we interpreted that participants may have been under the
impression that their story may be perceived as unbelievable by a listener.
Thus, by sharing their own personal feelings and beliefs about the validity of
their experience within the ADC story instead of waiting for the listener’s
reaction to their story, participants seemed to be able to gain some control
of the narrative they were presenting. Further, by disclosing their personal
attitudes, which either left room for doubt or expressed their confidence in
the ADC experience, participants appeared to be attempting to manage how
they might be perceived by the listener.
Whether the participant doubled down on their belief in their ADC experi-
ence or communicated an air of uncertainty about their ADC experience, they
seemingly conveyed a message that portrayed their understanding of how
ADC experiences are generally viewed in society, which is as unusual. No one
voiced anything to the effect of “I know you will not believe me” when sharing
their ADC narrative, but the objective structural findings of the study shed
light on ADC narratives unfolding differently, with storytellers appearing to
tow a line in revealing this hidden part of their life. By using these objective
findings as a springboard, we could then build from the structural findings to
reflect on the reasons why participants spoke about their ADC experiences in
such a manner—to attempt to gain ownership of and, in a way, control the
perception that the listener may have of them. With this newfound knowledge,
our study’s findings further identified how foundational ideas about grief, such
as the need for bereaved persons to relinquish ties to the deceased, continue to
influence how people share about their lives after loss.

Rigor and Trustworthiness in Narrative Analysis

As with any qualitative approach, it is important to consider rigor and
trustworthiness in the analysis. However, it is also important to remember
that there is no “right” answer that can be easily verified by outside means.
Although, as discussed at length in this chapter, precision and objectivity
are paramount in identifying the particular order that narratives unfold
from a structural standpoint (especially when utilizing an established narra-
tive model), subjectivity exists in the interpretation of why certain narrative
elements are included in a story or not. In short, the researcher must use the
objective data to infer the reasons behind the choices a storyteller makes in
sharing a narrative. Therefore, the key to rigor and trustworthiness is in the
transparency of the approach. As mentioned earlier, one should be clear
about how narrative is being defined (e.g., what counts as a narrative in this
work), how are narratives being identified (e.g., response to a question or
discourse marker), and what analytical approach is being used (e.g.,
 Reading between the Lines of After Death Communication Stories 71

Labovian structural approach or thematic analysis) (Loh, 2013; Phoenix

et al., 2010).
The notion of “thick description” is also important (Mishler, 1986). A
thick description involves providing details of context (e.g., when and where
the interview took place or notable features of the interview) and content
(e.g., observations that can be linked to responses, such as a participant cry-
ing while providing a response). Thick description details were added to each
transcript including field notes about the interview and the insertion of any
participant reactions (e.g., hesitations, crying, or laughing). These details
provided important context when we analyzed the interviews. Another
important approach was to look for verification within the larger interview.
For example, in a study that used thematic analysis to explore the meaning of
“place” in the small stories of research participants, researchers were able to
link “place” to other portions of the interview to support their interpreta-
tions (de Medeiros & Etter-­Lewis, 2020). This type of verification is similar to
techniques used by literary scholars when interpreting a particular piece of
work. The strength of the interpretation comes from a close reading of the
narrative and through evidence found within the work itself rather than
through some type of outside confirmation (Swinnen & de Medeiros, 2017).
In addition, it is important to note that although member checking is a com-
mon strategy in other qualitative approaches, it is not particularly useful in
narrative analysis for several reasons. A story told at a given time may have a
different meaning if told at a later time. Because the act of narrating an event
involves in-­the-­moment melding of past experience and current interpreta-
tions, a narrative about the same event may change (de Medeiros, 2014; Loh,
2013). As mentioned earlier, the researcher should not be concerned with
whether a narrative is “true” or “accurate” but rather should view the narrative
as an expression of meaning at that particular time by that particular person.

Conclusion
Narrative analysis is a useful strategy to assist in making implicit messages
from research participants explicit. Rather than looking for overarching
themes within the data, as in many qualitative approaches, narrative analysis
allows the researcher to consider how participants make sense of their expe-
riences by considering the stories they tell and how these stories unfold:
What details are included or excluded? What is the general plot of the sto-
ries? Who are the characters?
In this chapter, we described how we examined ADC narratives using
Labov’s (2001) model of natural narrative. To complete this secondary anal-
ysis, we read through research participants’ transcripts multiple times to
look for specific stories where participants revealed having experienced an
ADC. Once ADC stories were identified, we coded the data using Labov’s
72 Sara E. Hackett and Kate de Medeiros

(2001) model as a guide to look for structural patterns—mainly which ele-

ments of narrative appeared in their stories and which elements were left
out. As detailed in the chapter, our analysis revealed a pattern that few par-
ticipants began sharing an ADC with an abstract (i.e., a hint that they were
about to tell a story). By identifying this pattern, we were able to reexamine
our data to explore the meaning behind why participants’ ADC narratives
unfolded in a particular way. Ultimately, we found that the telling of ADC
narratives was shaped by participants’ perception of how their experience
would be received by a listener, illustrating that foundational ideas about
grief, such as the need for bereaved persons to relinquish ties to the deceased,
continue to influence how people share about their lives after loss.
In conclusion, by identifying and closely examining the structure and
content of the narrative, researchers are able to pay attention to details that
often challenge the dominant cultural narratives of a given phenomenon,
such as how people experience the death of a loved one. Although the objec-
tive, precise, and rigorous nature of narrative analysis may seem to challenge
the reflexive underpinnings of qualitative research, the systematic approach
of examining the structure of interview data, rather than the content alone,
provides researchers with the gift of being able to explore the deeper mean-
ing more intimately behind the rich qualitative data that they collect.

KEY WORKS GUIDING OUR DATA ANALYSIS

de Medeiros, K. (2014). Narrative gerontology in research and practice. Springer.

de Medeiros’ book provides a useful look at how narrative analysis can be uti-
lized to understand the experiences of older adults. Although our study was inclu-
sive of younger adults, the text was most helpful in highlighting the dynamic
process that takes place between storytellers and their audiences. With the knowl-
edge gleaned from this text that listeners play an important role in shaping the sto-
ries that people tell, we conceptualized our study to examine how ADC stories
unfold so as to add to the literature about what is known about ADC experiences.
Labov, W. (2001). Uncovering the event structure of narrative. Georgetown University
round table on language and linguistics (pp. 63–83). Georgetown University Press.
Labov’s model offers a simple and transparent definition of narratives and
each element that may exist within them. By using his model, we were able to
easily identify narratives that appeared in our interview transcripts. Utilizing his
elements of narratives as codes helped us systematically code the ADC narra-
tives in our data and make notes about how the narratives that appeared in our
data were structured.
Riessman, C. K. (1993). Narrative analysis. Sage.
Riessman’s guide provides readers with a rich history of narrative analysis and
what one can and cannot expect from employing this qualitative analytical method.
Her text showcases illustrative examples and detailed step-­by-­step instructions to
help researchers learn how to interpret complex qualitative narrative data. We used
Riessman’s book to guide our data analysis and to help with our development of
themes once we identified the structural patterns that existed in the ADC narratives.
 Reading between the Lines of After Death Communication Stories 73

References
Antonucci, T. C., Ajrouch, K. J., & Birditt, K. S. (2013). The convoy model:
Explaining social relations from a multidisciplinary perspective. The Gerontologist,
54(1), 82–92. [Link]
Azungah, T. (2018). Qualitative research: Deductive and inductive approaches to
data analysis. Qualitative Research Journal, 18(4), 383–400. [Link]
10.1108/QRJ-­D-­18-­00035
Becker, G. (1997). Disrupted lives: How people create meaning in a chaotic world.
University of California Press.
Breen, L. J., & O’Connor, M. (2011). Family and social networks after bereavement:
Experiences of support, change, and isolation. Journal of Family Therapy, 33(1),
98–120. [Link]
Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time seriously:
A theory of socioemotional selectivity. American Psychologist, 54(3), 165–181.
[Link]
de Medeiros, K. (2014). Narrative gerontology in research and practice. Springer.
de Medeiros, K., & Etter-­Lewis, G. (2020). “Place” in the small stories of African
American elders: A narrative case study. The Gerontologist, 60(5), 821–830.
[Link]
Fuller, H. R., Ajrouch, K. J., & Antonucci, T. C. (2020). The convoy model and later-­
life family relationships. Journal of Family Theory & Review, 12(2), 126–146.
[Link]
Guggenheim, B., & Guggenheim, J. (1997). Hello from heaven: A new field of
research—after-­ death communication confirms that life and love are eternal.
Bantam Books.
Kahn, R. L., & Antonucci, T. C. (1980). Convoys over the life course: Attachment,
roles and social support. In P. B. Baltes & O. G. Brim (Eds.), Life-­span develop-
ment and behavior (Vol 3, pp. 253–286). Academic Press.
Klass, D. (2006). Continuing conversations about continuing bonds. Death Studies,
30(9), 843–858. [Link]
Klass, D., Silverman, P. R., & Nickman, S. L. (Eds.). (1996). Continuing bonds: New
understandings of grief. Routledge.
Labov, W. (2001). Uncovering the event structure of narrative. Georgetown University
round table on language and linguistics (pp. 63–83). Georgetown University Press.
Labov, W., & Waletzky, J. (1967). Narrative analysis: Oral versions of personal expe-
rience. In J. Helm (Ed.), Essays on the verbal and visual arts. Proceedings of the
1966 Annual Spring Meeting (pp. 12–44). University of Washington Press.
Loh, J. (2013). Inquiry into issues of trustworthiness and quality in narrative studies:
A perspective. The Qualitative Report, 18(33), 1–15. [Link]
2160-­3715/2013.1477
Maschler, Y., & Schiffrin, D. (2015). Discourse markers language, meaning, and con-
text. In D. Tannen, H. E. Hamilton, & D. Schiffrin (Eds.), The handbook of dis-
course analysis (pp. 189–221). Blackwell.
Mishler, E. G. (1986). Research interviewing: Context and narrative. Harvard
University Press.
Peterson, L. J., Dobbs, D., June, J., Dosa, D. M., & Hyer, K. (2021). “You just forge
ahead”: The continuing challenges of disaster preparedness and response in long-­
term care. Innovations in Aging, 5(4), 1–13. [Link]
74 Sara E. Hackett and Kate de Medeiros

Phoenix, C., Smith, B., & Sparkes, A. C. (2010). Narrative analysis in aging studies:
A typology for consideration. Journal of Aging Studies, 24(1), 1–11. [Link]
org/10.1016/[Link].2008.06.003
Polkinghorne, D. E. (1988). Narrative knowing and the human sciences. SUNY Press.
Riessman, C. K. (1993). Narrative analysis. Sage.
Silverman, P. R., & Klass, D. (1996). Introduction: What’s the problem? In D. Klass,
P. R. Silverman, & S. L. Nickman, (Eds.), Continuing bonds: New understandings
of grief (pp. 3–27). Taylor & Francis.
Stemen, S. E. (2020a). “I can’t explain it”: An examination of social convoys and
after death communication narratives. Death Studies, 46(7), 1631–1640. https://
doi.10.1080/07481187.2020.1825296
Stemen, S. E. (2020b). “They’re in my heart, mind, and cells”: What continuing
bonds reveal about social networks. The Gerontologist, 60(6), 1085–1093. https://
doi.10.1093/geront/gnz188
Swinnen, A., & de Medeiros, K. (2017). “Play” and people living with dementia: A
humanities-­based inquiry of TimeSlips and the Alzheimer’s Poetry Project. The
Gerontologist, 58(2), 261–269. [Link]
5
DATA ANALYTIC STRATEGIES USED IN
A REMOTE PHOTOVOICE PROJECT OF
FILIPINO SINGLE MOTHERS DURING
THE COVID-19 PANDEMIC
Dennis S. Erasga, Jerome V. Cleofas, Mary Rose Jean
Andrada-­Poa, and Ronaldo F. Jabal

Since its introduction by Wang and Burris in 1997, photovoice has become
a widely used qualitative research design for participatory needs assess-
ment using both visual and narrative methodologies. Photovoice extends
an authentic agency unique to the participants’ positionality. It starts with
the interpretive affordance of photographs as both data and a tool to cap-
ture the phenomenon that matters in the lives of people. Overall, the
research participants wield their agency by investing meaning in every
photograph they take, therefore weaving a narrative consistent with how
they genuinely feel about their lives. Photovoice methods have been used to
expose and understand the experiences of single mothers (e.g., Valiquette-­
Tessier et al., 2015) and vulnerable populations affected by crises (e.g.,
Iskander, 2015).
Analytic techniques vary across published photovoice studies. The objec-
tive of this chapter is to provide an in-­depth account of the data analysis
procedure of our photovoice project that explored the experiences of single
Filipino mothers during the time of a global pandemic. We specifically focus
on the steps we took to integrate narrative and photographic data using the
analytic strategy “following a thread” (Cronin et al., 2008). More details on
the background, design, and findings of this project can be found in our
article titled “Single mothering during the COVID-­19 pandemic: A remote
photovoice project among Filipino single mothers working from home”
(Andrada-­Poa et al., 2022), which will henceforth be referred to as “the sin-
gle mothering study.”

DOI: 10.4324/9781003254041-6
76 Dennis S. Erasga et al.

The Remote Photovoice Project and Its Context

Initially, the aim of the single mothering study was to explore the work-­from-­
home (WFH) experiences of Filipino single mothers during the 2019 novel
coronavirus (COVID-­19) pandemic. Cognizant of the vulnerable circum-
stances experienced by single mother-­headed households during crisis situa-
tions, we used photovoice as the design to address our research goal. Our
findings showed that during the time of COVID-­ induced confinement,
Filipino single mothers working from home performed multiple, overlap-
ping, and new roles of motherhood. Within this context, new optics of single
mothering emerged, depicting the quadrifocal mothering characterizations:
(a) increased intimate mothering, (b) shared mothering at home, (c) mother-
ing as working, and (d) mothering as teaching. A more complete discussion
of the findings can be found in Andrada-­Poa et al. (2022).

Doing Research during the COVID-19 Pandemic

Two important contexts shaped the collection and analysis procedures of
this qualitative study. First, this study was done amid the enhanced commu-
nity quarantine (ECQ) implemented by the Philippine national government
in mid-­2021. The Philippines was one of the countries with the longest lock-
downs in the world (March 2020–February 2022) and the poorest rankings
in terms of COVID-­19 resilience (Chang et al., 2021). During ECQ, estab-
lishments that offered non-­essential goods and services had limited opera-
tions. Travel restrictions were imposed. Most office-­based work was relocated
to the home. University campuses were closed, and education transitioned
to fully online learning. Social interactions, such as those inherent in collab-
orative qualitative analysis processes, had to be done online. Hence, our
research team had to innovate our data collection and analysis strategies
beyond traditional means.

Rapid, Frugal, and Coursework-Based Research

The second context that influenced the research team’s methodological and
analytic decisions for the study was its administrative and logistical features.
This project was an academic requirement for a postgraduate degree course
in qualitative research. The first two authors, Mary Rose and Ronaldo, are
doctoral students in sociology under the mentorship of Jerome, the last
author. A PhD course in this program spans 14 weeks. Because the project
only had just over three months from inception to the initial technical report,
it can be considered a rapid qualitative study (Vindrola-­Padros et al., 2020).
Rapid qualitative research studies were common during the early stages of
the pandemic to quickly inform initiatives to address its emergent and
 Remote Photovoice and Filipino Single Mothers during COVID 77

abruptly changing challenges (Vindrola-­Padros et al., 2020). Moreover, this

endeavor was an independent study with no external funding. Hence, we had
to make frugal methodological innovations for the project while ensuring
the trustworthiness of our findings. For example, we used free cloud-­based
office and instant messaging applications (i.e., Microsoft Office, Google
Drive, and Facebook Messenger) to perform analysis because of the lack of
finances to purchase qualitative data analysis software.

Remote Photovoice Data Constitution and Management

In Andrada-­Poa et al. (2022), we vigorously explained and justified how our
remote photovoice protocol veered away from some features of the tradi-
tional photovoice as a practical response to COVID-­19 restrictions. The
data collection had two phases: (a) the photography phase and (b) the inter-
view phase. In the photography phase, the participating single mothers were
asked to take and send via email one photograph of a moment that answered
the question, “what was your highlight for the day as a single mother work-
ing from home?” using their smartphones. They were instructed to submit a
photo once a day for three days. They were also asked to send a short text
description with the photo in the email. In the interview phase, we conducted
online one-­on-­one interviews with each mother via video conferencing,
guided by the photographs they shared and the SHOWeD method (Wang,
1999). The SHOWeD framework of questioning for photovoice participants
asks them to reflect on the images in terms of what they “see,” what they
think is “happening,” how the images relate to their (“our lives”), “why”
these conditions “exist,” and what can be “done” about them.
Because most of our collection methods were digital, we also managed
our data through digital means. All data collected for the study were stored
via cloud storage (i.e., Google Drive). We created a shared Google Drive
folder using our university accounts. Inside the general folder, we had sub-
folders that held specific groups of data. The number, names, and locations
of the folders and files changed as we went through our analysis. Mary Rose
was assigned to organize our files.
We collected four types of data for this project. First, we collected back-
ground information from each of the 15 participating single mothers to
describe the profile of our sample. Details about their age, number, and age of
child/ren, household size, educational attainment, work, and monthly income
were collected via a survey form that they submitted with the signed consent
form. Although the participants’ profiles were not included in the main analy-
sis, knowing the social position of the mothers guided how we analyzed the
findings of the study in the discussion section. A tabular presentation of their
profiles was placed in the Study Participants section of our shared folder.
78 Dennis S. Erasga et al.

The second and third sources of data were the three digital photographs
and the accompanying text descriptions, respectively. The photographs were
downloaded as .jpg or .png files. Each photograph was named using this
format: “Day number, Date submitted, Pseudonym.” The text descriptions
were transferred from the email to a Microsoft Word file. All three photo
descriptions of one participant were saved in [Link] file labeled
“Pseudonym, Photo Descriptions.”
At the start of the analysis, folders were made for each participant, and the
respective photographs and photo descriptions were placed in them. However,
as the analysis deepened and the major superordinate themes were identified,
we reorganized the folders and files. There were four folders, one for each super-
ordinate theme. The photos that represented each theme were copied to the
respective theme-­based folders, likewise with the photo description documents.
This helped in the curation of the data repertoire (more detail on this will be
provided in a subsequent section). Figure 5.1 presents the organizational
scheme in our Google Drive folder during the terminal part of the analysis.

[SEE FIGURE 5.1 at eResource: Screenshot of the Shared Google Drive Folder with
Analysis Files].

The fourth data came from the in-­depth interviews, each lasting from 45
minutes to one hour. Mary Rose and Ronaldo transcribed all interviews ver-
batim into a Microsoft Word file. Each transcript file was placed in its respec-
tive participant-­ labeled folder. Mary Rose and Ronaldo translated the
interview transcripts in the Filipino language to English and then back-­
translated to maintain the overall integrity of the transcripts. Transcripts
were then transferred to Google Sheets (a cloud-­based spreadsheet program
similar to Microsoft Excel). Each turn/response of the participants was
placed in a cell, and each response was labeled with a reference number in
the first column. The subsequent columns were used for coding. One sheet
was initially assigned per participant. Later in the analysis when the higher-­
order categories/themes had been identified, new sheets were created, plac-
ing the transcripts and codes assigned to the same category (i.e., threads)
together to assist in identifying further patterns. This will be further elabo-
rated later in this chapter in the Initial Analysis Per Dataset section.

The Photovoice Participants’ Role in Shaping the “Key Issue”

As with most qualitative studies, we initiated this project with a broad
research objective, meaning that no preset theoretical framework was identi-
fied to guide our research questions. Instead, a sensitizing concept was
agreed upon to help steer the direction of the data collection and analysis
(Anfara, 2008; Hoonaard, 2008). Originally, the key issue that we set for the
 Remote Photovoice and Filipino Single Mothers during COVID 79

study was “working from home experiences of Filipino single mothers dur-
ing the COVID-­19 pandemic.” Because the PhD student members of our
research team were organizational development majors, we were initially
interested in the phenomenon of WFH as the focal point of our analysis,
with single motherhood and COVID-­19 as the contexts that influenced such
experience.
However, photovoice, by its nature, is participatory and participant
driven. Wang (1999) asserts that photovoice participants must be fully
engaged in defining the issues and images that shape public discourse.
During the photography phase, we noticed that the experiences represented
in their photograph submissions portrayed mothering as their main concern
amid pandemic-­induced confinement. This change in the key issue was fur-
ther confirmed during the interviews with the mothers whose sharings char-
acterized their roles primarily as mothers rather than workers or employees.
As directed by the photographic and narrative expressions of our photo-
voice participants, our final phenomenon and analytical focus shifted to
“single mothering while working from home amid the pandemic.”

Employing “Following a Thread” as a Qualitative Integrative

Analysis Strategy
Photovoice studies have used various analytical strategies to draw insights
from the data and report their findings. These strategies can be grouped into
two types. The first type includes photovoice studies whose main emphasis
of qualitative analysis is the narrative of participants (e.g., Suprapto et al.,
2020; Tsang, 2020). In these studies, the insights from the photographs are
mainly derived from the participants’ interviews. The second type involves
photovoice studies that place more emphasis on the photographs taken by
the participants themselves (e.g., Wang & Hannes, 2020).
Our photovoice study borrowed elements from both types, attempting to
strike a balanced analysis across the different sources of data in our project:
photographs, accompanying text descriptions, and interview narratives. To
achieve an integrated analysis among our qualitative datasets, we appealed
to the analytic strategy that Cronin et al. (2008) coined as “following a
thread” (p. 572). This qualitative integrative analysis method strives to
ensure the equal contribution of data across datasets. It is comprised of four
steps: (a) initial analysis using data-­appropriate techniques, (b) initiating a
“thread” from a “promising finding” within a dataset, (c) creation of a “data
repertoire” for the threads, and (d) synthesizing with other threads. The fol-
lowing subsections demonstrate these steps using one of the threads that
emerged from the single mothering study, which is “more intimate moments
with children.”
80 Dennis S. Erasga et al.

First Step: Initial Analysis per Dataset

We analyzed each dataset based on analytic techniques that were appropri-
ate for the method. For this, we used the first three phases of the inductive
reflective thematic analysis by Braun et al. (2019) because of its flexibility
with both narrative and visual data. Inductive reflective thematic analysis
involves the development of higher analytical conceptualizations (i.e., cate-
gories and themes) from qualitative data (both narrative and visual) without
being bound by preexisting theoretical and conceptual frameworks (Braun
et al., 2019). We analyzed the narratives from the interviews first because
these had the thickest volume among the datasets. Then, we examined the
photographs and their accompanying descriptions.
The first phase of reflective thematic analysis is familiarization. We read
and reread all the interview transcripts and photo descriptions. We also lis-
tened to interview audio recordings and reviewed the photographs to notice
interesting features and make notes about the individual data items (Braun
et al., 2019). We documented our reflections on the data in two ways: (a)
making written memos regarding our thoughts about the data and (b) dis-
cussing our reflections through videoconferencing during online classes and
through instant messaging via Facebook.
The second phase of thematic analysis was generating codes. Coding is
the process that aims to systematically organize and reduce data (Braun
et al., 2019). We agreed that the phenomenon would guide our coding. The
initial coders of the data were Mary Rose and Ronaldo (the PhD students).
Early in the course, Jerome (the faculty/mentor) demonstrated the coding
process via videoconferencing. The student research members would be
given a week or so to perform the initial coding for specific types of data.
Jerome would then visit the online file to peer review and make modifica-
tions. The comment function of Google Docs was also used to communicate
coding suggestions. During online class meetings, the reflections and correc-
tions on the coding were then discussed in class. Jerome was cognizant of the
vertical nature of the student–teacher relationship. So, he regularly reas-
sured Mary Rose and Ronaldo that their voices were of equal weight as his
in the analysis, especially because they were more exposed to the data. He
also noted that his role as a mentor was to facilitate, evaluate, affirm, and
challenge their analytic insights.
For the textual narratives from the interview and photo descriptions, the
coding was done via shared Google Sheets. Oftentimes, a full response would
be a paragraph’s length; naturally, we would be able to identify many codes
in one cell. We would clone the row of the response based on the number of
codes that would be identified to ensure that only one code occupies per row.
This would help us later in the retrieval of codes and representative quotes
 Remote Photovoice and Filipino Single Mothers during COVID 81

when they were already being grouped for higher-­level analysis using the
filter function. The codes were placed in the column beside the narrative
being analyzed. Table 5.1 shows sample coding for interview and photo
description narratives.
For photographs, we identified codes using selected analytical questions
by Wang and Hannes (2020) for analyzing photovoice images. Examples of

TABLE 5.1 Sample Coding of Textual Data under “More Intimate Moments with
Children”

Sample Interview Data

Ref No.1 Original Transcript English Code(s) Theme

Extracts Translation

E6 Mahalaga rin yung It’s important Playtime; More

aside sa playtime that aside telling intimate
yung meron din from the stories; moments
kaming parang playtime, we scheduled with
kwentuhan. So, also tell stories bonding children
‘yung makakarelate to one time
ako sakanya, lalo another. So, I
only son lang sya, can relate with
so ‘yun yung him, especially
parang pang relax that he’s an
nya daw kaya tawag only son. So,
nya is bonding time that is how he
so parang pang relaxes. He
relax niya for that calls it
time around 8:30 “bonding
time”—for his
relaxation
around 8:30
[PM]
F4 Every time I get to Every time I get Celebrating More
celebrate their to celebrate daughter’s intimate
birthdays kahit sa [my children’s] birthday moments
simple na birthdays, even with
pamamaraan lang in simple ways, children
small victories na they are
‘yun samin as already small
family yun kasi victories for
yung naranasan my family.
naming noon na
wala kaming pera…
(Continued)
82 Dennis S. Erasga et al.

TABLE 5.1 (Continued)

Sample Textual Description of Photographs

Ref No. Original Transcript Code(s) Theme
Extracts
E4/28 He likes watching Watching TV More intimate moments
“We Bear Bears.” together; doing with children
Our bonding time is housework
watching tv or
movie together as I
am folding clothes.
F4/28 This photo taken Health seeking More intimate moments
taken today, while for sick child; with children
im working, my tending to
daughter was child’s health
scheduled for needs while at
telemedicine work
consultation. Were
having videochat
with her pedia.
Good thing about
work from home, I
can still manage to
prioritize my
daughter’s wellness.
1
Reference number.

these questions include “what does the participant want to tell?” “what ideas
and values are expressed by the participant?” and “what is the significance
of this photo to the participant?” In answering these questions and coding
our photographs, we also consulted the participants’ narratives in the inter-
views and associated descriptions. A sample coding of photographs related
to “more intimate moments with children” is found in Figure 5.2.

[SEE FIGURE 5.2 at eResource: Coding of Selected Photographs].

The third phase of thematic analysis was the construction of themes, wherein
patterns among codes were examined and grouped into categories. Theme
construction began with formulating prototypes or candidate themes (Braun
et al., 2019). We then attempted to assign codes to each theme and reflect
on whether we could identify a meaningful pattern would surface. In our
study, we had to construct and reconstruct many prototypes to describe the
pattern within our datasets regarding the positive accounts of motherhood
described in the interviews and portrayed in the photographs. For instance,
 Remote Photovoice and Filipino Single Mothers during COVID 83

to synthesize the codes that pertained to mother and child interactions, we

initially formulated “bonding with the child.” However, while codes like
“playing with the child” and “talking to the child” seemed applicable under
this category, it failed to include similar codes such as “observing the child
while at work” or “responding to child’s needs.” Moreover, “bonding with
the child” did not evoke the context of the changing social environment due
to pandemic-­induced confinement, which was also a consideration of the
study and the participants themselves. Hence, to involve emotional and care-
giving functions and how these interactions were intensified during the pan-
demic, we decided that a more appropriate theme would be “more intimate
moments with children.” The themes that were developed in this phase were
also placed in the spreadsheet file in the same row beside their respective
codes and quotes.

Second Step: Identifying Promising Threads

At this stage, the analysis products that we had were codes and lower-­level
themes, the first order of themes directly derived from the codes (Braun
et al., 2019) for each dataset (i.e., interview narratives, photographs, and
photo descriptions). We considered these as intermediate products of the
analysis as a result of doing the first three phases of thematic analysis, which
would now be subjected to a deeper and integrative analysis using the fol-
lowing thread strategy. These themes would then be considered as possible
promising threads that can potentially serve as the “containers” where the
other codes and themes from across datasets were then placed. These threads
also served as the main components or descriptors of the phenomenon of
interest, as suggested by Cronin et al. (2008).
At the end of step one, we had identified multiple promising threads,
which included “more intimate moments with children,” “ability to do more
chores,” “sharing parenthood with relatives,” “taking care of other relatives,”
“positive attitudes towards working from home,” “maintaining work-­life bal-
ance,” “negative psychological impacts of WFH,” “engagement in child’s
schooling,” and “coping with distance learning.” Most of these threads were
identified from interview data. Themes identified from photographs and
descriptions depicted concepts like the ones drawn from the interviews.
Generally, we discussed and determined the threads based on two consider-
ations. The first consideration was the thickness/richness of data assigned to
the theme. The second consideration was how we, the researchers, perceived
the theme’s contribution to telling a more cohesive story to represent the phe-
nomenon of interest. For instance, we decided that “more intimate moments
with children” was a promising thread candidate because most of the moth-
ers described the increased positive interactions they had with their children
when they transitioned to WFH. Moreover, many of the photographs and
84 Dennis S. Erasga et al.

emailed descriptions portrayed mother and child bonding and caregiving.

Also, this thread was closely related to the main phenomenon of the study,
which was single mothering amid the pandemic.

Third Step: Creating a Data Repertoire

This integrative step of following a thread involved juxtaposing the interme-
diate products of analysis (i.e., codes and data segments) of individual data-
sets within the promising thread, which served as the conceptual containers
or data repertoire (Cronin et al., 2008, p. 577). This helped to identify nuances
within the data, deepen the insights in each conceptual category, and improve
the representation of these categories by pulling in diverse but related data
sources, as demonstrated in the work of Cronin et al. (2008).
When the promising threads had been identified, we reorganized our files
in our cloud-­based platform to facilitate the creation of the data repertoires.
First, for the interviews, new spreadsheets were created for each promising
thread, and the codes and their accompanying quotes were migrated to their
respective threads. The reference numbers were maintained to ensure easy
retrievability that would aid in writing the final report. For the pictures and
their descriptions, their files were placed in the folder that corresponds to the
thread that they follow. The interview spreadsheets, image files of photo-
graphs, and Microsoft Word files of photo descriptions under a given thread
were placed in a folder.
For the thread “more intimate moments with children,” which is the
exemplar for this chapter, the coded interview narratives that demonstrated
the increased bonding and caregiving moments across participants were
placed in a single spreadsheet. In its containing folder, the spreadsheet was
located with photographs of mothers and children being close to each other
and doing activities together in one frame. Among the threads, this had the
greatest number of representative images in its repertoire. Putting all the
connected data across datasets in a single folder/repertoire helped us charac-
terize and tell a more robust story for each thread. Due to this, we changed
the names of threads from the original themes they were connected to. For
instance, the promising thread based on the theme from step 1 named “posi-
tive attitudes towards working from home” was changed to “increased drive
to perform work in home setting” after curating the data repertoire. The new
name better encapsulated not only attitudes but also the work-­related activi-
ties and emotional investments that were portrayed in the narratives and
photographs in the repertoire.
This stage of following a thread also helped us narrow down the themes
and conceptual categories to be brought further into the analysis pipeline. For
instance, the previously identified promising thread “negative psychological
 Remote Photovoice and Filipino Single Mothers during COVID 85

impacts of WFH” was only represented by a handful of verbalizations in the

interviews of some single mothers, and no photographs signified the theme.
The data repertoire for this thread was not as robust as the others. It was a
signal for us to exclude this theme from further analysis.

Fourth Step: Synthesizing Followed Threads

Cronin et al. (2008) noted that at the end of curating the threads, the
researchers can combine or link together similarly followed threads to
answer the overarching research question. At this stage, we already had a
total of eight folders, each containing the data repertoire and labeled accord-
ing to the thread it represented, and our highest conceptual categories were
represented by the threads themselves. To tell a more cohesive story about
the experiences of single mothers while working from home during the pan-
demic, we had to synthesize related threads to develop even higher levels of
themes. Because the threads were originally derived from first-­level themes
from step 1, the resultant category from the fusing of two threads would be
superordinate themes, which is another term for higher-­level themes devel-
oped from lower-­level/subthemes (Braun et al., 2019).
We noted similarities between the threads “more intimate moments with chil-
dren” and “increased engagement with housework.” The first thread described
the close and active interactions between the mother and the child, whereas the
second described the ability of the mother to do more household chores because
WFH afforded more time at home. Both threads cohered under the superordi-
nate theme, “increased ‘presence’ at home.” In this study, we developed four
superordinate themes, each representing the fusion of two threads.
We synthesized the superordinate themes further by identifying a core cate-
gory in which all the lower-­level abstractions could be subsumed. A core cate-
gory is the “main theme” that efficiently condenses the insights drawn from the
photographs and narratives of the project. Although commonly associated
with grounded theory, core categories can also be used in other qualitative
designs (Madill, 2008). For our study, the core category was “performing over-
lapping roles of motherhood in confinement during a pandemic” (see Andrada-­
Poa et al., 2022). Figure 5.3 illustrates the four steps of “following a thread”
using the exemplar of “more intimate moments with children.”

[SEE FIGURE 5.3 at eResource: Flowchart for “Following a Thread”].

Taken together, these steps helped in managing the integrative analysis of

qualitative data across multiple datasets. We note, however, that although we
presented these procedures as steps, our actual implementation was iterative,
like many qualitative designs. In our analytic process, these steps often
86 Dennis S. Erasga et al.

overlap with one another. For instance, during our initial coding of a specific
data source (i.e., narratives) in step 1, we also allowed ourselves to look at
other datasets (i.e., photographs) to examine similarities that would help
ease the process of constructing and reconstructing themes for promising
threads. Another instance is that during our creation of data repertoires per
thread (step 3), we were already reflecting on possible higher-­level categories
(super-­ordinate themes and core categories).
In writing the results section of the research report, we also made use of
the data repertoire to easily locate the options of the narratives and photo-
graphs that would serve as representations of the themes. We decided to have
one photograph to represent each subtheme during a videoconference meet-
ing. The photos under a theme were shared on a single screen. We voted on
the picture that was closely related to the theme and had the most aesthetic
visual (i.e., the image that “pops out” from the rest).

Ensuring Trustworthiness
To establish the rigor of our methods and the quality of our findings, we
followed the basic guidelines for ensuring trustworthiness forwarded by
Lincoln and Guba (1986). Some features inherent to our photovoice project
provided assurances of credibility. These included the collection and inte-
gration of multiple data sources and engaging with participants multiple
times via emails, interviews, and member checking of results. We also wrote
reflexivity memos to help manage our preconceived notions regarding the
phenomenon. The analytic insights and reflections of each researcher were
discussed during our synchronous classes or through quick private messag-
ing sessions online. The memos of the student researchers were saved as
document files and submitted to Ronaldo as a part of their qualitative port-
folio, which was a major requirement of the course. These memos were also
stored in our shared cloud folder. An excerpt of Mary Rose’s memo is as
follows:

I am a married working mother. While I am not a single mother, my expe-

riences, assumptions, and beliefs as a mother working remotely during
the pandemic inspired me to recommend the topic in class. My own expe-
riences managing work and family care during the prolonged lockdown
and my personal conversations with single-­parent co-­workers sparked my
interest and drive to give voice to a group of workers rarely considered in
shaping the policies around remote work during the pandemic.

As demonstrated in the previous section, we also kept a digital audit trail (i.e.,
analytic products and memos regarding the data) to ensure dependability and
 Remote Photovoice and Filipino Single Mothers during COVID 87

confirmability. For the transferability of findings, we managed the expectations

of our readers by presenting the profile of the respondents and clarifying the
specific sociocultural contexts under which the phenomenon of WFH single
motherhood was embedded (see Andrada-­Poa et al., 2022). In the next subsec-
tions, we highlight other notable strategies we used to ensure trustworthiness.

Working Remotely as a Qualitative Research Team

Constant communication among research team members is important to
achieve consensus on various data collection and analytic points and to
improve the credibility, confirmability, and dependability of the processes
and outputs. This coursework-­based, rapid research project was conducted
amid a global pandemic, specifically during the outbreak of the Delta vari-
ant of the 2019 coronavirus. Hence, all the research team members had to
work together online. Rapid qualitative research during the early COVID-­19
outbreak had to innovate the research team management processes due to
social restrictions (Vindrola-­Padros et al., 2020). We used three digital plat-
forms to work remotely: videoconferencing, instant messaging, and cloud-­
based office applications. These tools were also used in previous qualitative
COVID-­19 research (Biesty et al., 2020).
Because this project was embedded in a PhD research course, the team
met synchronously via Zoom once a week. The first few meetings during the
term were devoted to lectures and discussions regarding the nature and pro-
cess of qualitative research in general, developing the photovoice protocol,
and obtaining ethical clearance. The latter part of the course was the actual
conduct of the photovoice project. From the onset of the research, we spent
one to two hours weekly in a videoconference meeting discussing the status
of the project, challenges encountered, reflections during collection and
analysis of the data, providing feedback on the analytic outputs of individ-
ual members, and building the research report. These meetings were recorded
using the functions available on Zoom and uploaded to our shared cloud
folder. Key points in every discussion were then documented in a Microsoft
Word file and stored as collaborative memos. These files were included in our
shared Google Drive folder.
Outside our formal class meetings, we used Facebook Messenger to com-
municate about urgent matters and reflective musings as we conducted data
collection and analysis and wrote the research report. Although informal,
Facebook Messenger had advantages that introduced ease to our processes.
First, it was available on a need-­by-­need basis as long as we were connected
to the internet. Information and feedback can be sent and received seam-
lessly and instantaneously, facilitating timely decision-­making. Second, this
messaging application also stored all our conversations in multiple forms,
88 Dennis S. Erasga et al.

such as text, audio recordings, and images (i.e., screenshots of our ongoing
analysis). While using an instant messaging application, we avoided discuss-
ing or revealing any of the participants’ identities or personal information to
ensure data security. Figure 5.4 shows screenshots of some conversations we
had on Messenger related to our analyses.

[SEE FIGURE 5.4 at eResource: Sharing Analytical Musings on Instant Messenger].

Finally, we used cloud-­based office applications (i.e., Google Drive, Google

Docs, and Google Sheets) to work collaboratively, check on the analytic out-
puts of other members, and provide feedback via comments asynchronously.
An advantage of these applications is the interoperability of the files for
both online and offline platforms. These applications stored all of our digital
audit trails.
These synchronous and asynchronous modalities of communication and
collaboration among team members were expedient in negotiating our dif-
fering insights and making critical decisions as we transitioned from one
step of the analysis to the next. For example, as noted earlier, it was through
these meetings that we shared our insights regarding the photographs and
interviews going in the direction of single mothering instead of the original
interest, which was working from home, prompting us to decide to shift the
key issue under study. These conversations were helpful in identifying prom-
ising threads, curating the contents of the data repertoire, determining
superordinate themes, constructing the core category, and addressing the
comments of peer reviewers.
However, working remotely had setbacks. First, it lacked the personal
touch normally done, seen, and felt in most qualitative research projects.
The exchange of ideas dotted with banter and personal stories, while also
done online, appeared muted and reserved. Second, we had to bear with and
manage potential technology-­related issues during online interactions, such
as delayed latency and unreliable internet connectivity.

Peer Review Dynamics

Peer review is a strategy that can enhance the rigor of qualitative analysis.
This entails asking third-­ party researchers to examine and critique the
details of the analytic process and outputs to improve confirmability and
dependability (Johnson et al., 2020). Upon finishing our audit trails and the
initial draft of the research report on the photovoice research, we sought the
guidance of a sociologist who was an expert in qualitative methodology
from our department (Dennis) to evaluate our work. He had minor com-
ments on the analytic products, subthemes, themes, and core category.
 Remote Photovoice and Filipino Single Mothers during COVID 89

His major comments focused on reflecting on and explaining our identified

findings in the context of the cultural and gendered nuances of Filipino
families. We included his comments in the discussion section of the single
mothering study (see Andrada-­Poa et al., 2022).
We also received blind peer reviews regarding our study as a part of the
journal publication process. Two reviewers provided sets of comments for
revision purposes. One reviewer’s comments were minor and focused on
editing issues, while another suggested a theoretical consideration to anchor
the entire research framework on WFH. Although it was tempting to con-
sider the prospects of such sound advice, the team took stock of the issues
within the larger pool of photovoice literature to respond. Juxtaposing the
benefits of a theoretical/conceptual direction (as recommended) with the
basic features of photovoice as evoked by a bevy of available critical litera-
ture, the team opted to retain the research’s inductive approach with a recog-
nition of the affordance of the theory-­directed photovoice approach for
future research endeavors.
Consistent with the issue tackled above, we entertained the benefit of
theoretical guidance. However, we adjusted the optics of a theoretical
utility approach (see Anfara, 2008) by focusing on the advantage of theo-
retical musing in view of the available qualitative insights drawn from the
photographic and narrative expressions of the WFH single mothers on
hand. Accordingly, the theory inventory we conducted yielded the appro-
priateness of the socioecological model proposed by Bronfenbrenner
(1979; see also Grzywacz & Marks, 2000). Through the prism of the
socioecological theory, we inferred that the changes and multiplicities
in the roles performed by single mothers were a function of socio-­
environmental factors, mainly the pandemic-­induced confinement and
transition to WFH setup. Other socioecological factors such as Filipino
gender norms, family dynamics, and passage of time were also considered
per consultation with our presubmission peer reviewer (Dennis). We inte-
grated these new details in the introduction and discussion sections of
the article. Additionally, where relevant, we invoked the socioecological
perspective as a major talking point throughout the paper. The revised
draft of our manuscript was then accepted and published (Andrada-­Poa
et al., 2022).

Conclusion
In this chapter, we highlighted the flexible nature of qualitative research, and
how it allows for innovation when traditional methods are impractical or
impossible. We discussed the contexts that prompted the innovations imple-
mented in the study (i.e., pandemic-­induced, course-­based, and rapid, online
90 Dennis S. Erasga et al.

nature of the research project). We also highlighted how digital strategies

can be applied to collecting and analyzing data for visual research, such as
photovoice. We explained how we utilized following a thread to integrate
diverse qualitative data and described our strategies for ensuring trustwor-
thiness using online and digital affordances. We demonstrated that qualita-
tive researchers could thrive despite the restrictions imposed by the
COVID-­19 pandemic and financial limitations. Researchers only need to
determine the resources available (i.e., online platforms and programs) and
maximize their potential while ensuring the fundamental principles of trust-
worthiness are observed. Also, the flexibility of qualitative research allows
for the inclusion of multiple data types for deepened analysis. Qualitative
strategies such as following-­a-­thread open possibilities for an expansive yet
focused analysis.

Acknowledgment
We thank the single mothers who participated in this study.

KEY WORKS GUIDING OUR DATA ANALYSIS

Cronin, A., Alexander, V., Fielding, J., Moran-­Ellis, J., & Thomas, H. (2008). The analytic
integration of qualitative data sources. In P. Alasuutari, L. Bickman, & J. Brannen
(Eds.), The SAGE handbook of social research methods (pp. 572–584). Sage. http://
[Link]/10.4135/9781446212165.n34
This book chapter explains clearly and succinctly the process of “following a
thread” using an exemplar from longitudinal, large-­scale qualitative data. Cronin
and colleagues provide an easy-­to-­understand, step-­by-­step guide to integrating
heterogeneous qualitative datasets.
Wang, C., & Burris, M. A. (1997). Photovoice: Concept, methodology, and use for par-
ticipatory needs assessment. Health Education & Behavior, 24(3), 369–387. https://
[Link]/10.1177/109019819702400309
This is the first methodological article introducing and describing the principles
and methodology of photovoice. Wang and Burris outline the participatory roots of
photovoice, the importance of visuals in elucidating women’s struggles, and the
procedure of classical photovoice. Through photovoice, women gain more con-
sciousness about their conditions and are aided to engage in participatory actions
to address their concerns.
Braun, V., Clarke, V., Hayfield, N., & Terry, G. (2019). Thematic analysis. In P. Liamputtong
(Ed.), Handbook of research methods in health social sciences (pp. 843–860). Springer
Nature. [Link]
This book chapter defines, categorizes, and explains the procedure of thematic
analysis (TA). Braun and colleagues clarify the misconceptions regarding TA, iden-
tify common mistakes done by TA researchers, and exhaustively discuss the differ-
ences and processes of each type of TA. This source also mentions the applicability
of TA to multiple types of qualitative data.
 Remote Photovoice and Filipino Single Mothers during COVID 91

References
Andrada-­Poa, M. R. J., Jabal, R. F., & Cleofas, J. V. (2022). Single mothering during
the COVID-­19 pandemic: A remote photovoice project among Filipino single
mothers working from home. Community, Work & Family, 25(2), 260–278. https://
[Link]/10.1080/13668803.2021.2006608
Anfara, V. A. (2008). Theoretical frameworks. In L. M. Given (Ed.), The SAGE ency-
clopedia of qualitative research methods (pp. 869–873). Sage.
Biesty, L., Meskell, P., Glenton, C., Delaney, H., Smalle, M., Booth, A., Chan, X. H.
S., Devane, D., & Houghton, C. (2020). A QuESt for speed: Rapid qualitative
evidence syntheses as a response to the COVID-­19 pandemic. Systematic Reviews,
9(1), Article 256. [Link]
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by
nature and design. Harvard University Press.
Braun, V., Clarke, V., Hayfield, N., & Terry, G. (2019). Thematic analysis. In P.
Liamputtong (Ed.), Handbook of research methods in health social sciences (pp.
843–860). Springer Nature. [Link]
Chang, R., Varley, K., Tam, F., Munoz, M., & Tan, A. (2021, November 30). The
best and worst places to be as winter meets Omicron. [Link]. https://
[Link]/graphics/covid-­resilience-­ranking/
Cronin, A., Alexander, V. D., Fielding, J., Moran-­Ellis, J., & Thomas, H. (2008). The
analytic integration of qualitative data sources. In P. Alasuutari, L. Bickman, & J.
Brannen (Eds.), The SAGE handbook of social research methods (pp. 572–584).
Sage. [Link]
Grzywacz, J. G., & Marks, N. F. (2000). Reconceptualizing the work-­family inter-
face: An ecological perspective on the correlates of positive and negative spillover
between work and family. Journal of Occupational Health Psychology, 5(1), 111–
126. [Link]
Hoonaard, W. C. V. (2008). Sensitizing concepts. In L. M. Given (Ed.), The SAGE
encyclopedia of qualitative research methods (pp. 812–814). Sage.
Iskander, D. (2015). Re-­imaging malaria in the Philippines: How photovoice can
help to re-­imagine malaria. Malaria Journal, 14(1), Article 257. [Link]
10.1186/s12936-­015-­0770-­8
Johnson, J. L., Adkins, D., & Chauvin, S. (2020). A review of the quality indicators
of rigor in qualitative research. American Journal of Pharmaceutical Education,
84(1), Article 7120. [Link]
Lincoln, Y. S., & Guba, E. G. (1986). But is it rigorous? Trustworthiness and authen-
ticity in naturalistic evaluation. New Directions for Program Evaluation, 1986(30),
73–84. [Link]
Madill, A. (2008). Core category. In L. M. Given (Ed.), The SAGE encyclopedia of
qualitative research methods (p. 131). Sage.
Suprapto, N., Sunarti, T., Suliyanah, S., Wulandari, D., Hidayaatullaah, H. N.,
Adam, A. S., & Mubarok, H. (2020). A systematic review of photovoice as par-
ticipatory action research strategies. International Journal of Evaluation and
Research in Education, 9(3), 675–683. [Link]
Tsang, K. K. (2020). Photovoice data analysis: Critical approach, phenomenological
approach, and beyond. Beijing International Review of Education, 2(1), 136–152.
[Link]
92 Dennis S. Erasga et al.

Valiquette-­Tessier, S.-C., Vandette, M.-P., & Gosselin, J. (2015). In her own eyes:
Photovoice as an innovative methodology to reach disadvantaged single mothers.
Canadian Journal of Community Mental Health, 34(1), 1–16. [Link]
10.7870/cjcmh-­2014-­022
Vindrola-­Padros, C., Chisnall, G., Cooper, S., Dowrick, A., Djellouli, N., Symmons,
S. M., Martin, S., Singleton, G., Vanderslott, S., Vera, N., & Johnson, G. A.
(2020). Carrying out rapid qualitative research during a pandemic: Emerging les-
sons from COVID-­19. Qualitative Health Research, 30(14), 2192–2204. https://
[Link]/10.1177/1049732320951526
Wang, C. C. (1999). Photovoice: A participatory action research strategy applied to
women’s health. Journal of Women’s Health, 8(2), 185–192. [Link]
jwh.1999.8.185
Wang, C., & Burris, M. A. (1997). Photovoice: Concept, methodology, and use for
participatory needs assessment. Health Education & Behavior, 24(3), 369–387.
[Link]
Wang, Q., & Hannes, K. (2020). Toward a more comprehensive type of analysis in
photovoice research: The development and illustration of supportive question
matrices for research teams. International Journal of Qualitative Methods, 19, 1–
15. [Link]
6
PHENOMENOLOGICAL ANALYSIS
AND RACIAL SOCIALIZATION
INTERPRETATION OF INTERVIEWS
WITH AFRICAN AMERICAN PARENTS
OF TODDLER SONS
Sheresa Boone Blanchard, Stephanie Irby Coard, and
Mariana Mereoiu

In this chapter, we describe two analytic processes for a study of self-­

identified African American parents’ culturally relevant beliefs, practices,
and goals for raising their toddler sons. We employed a two-­phase approach
for the study. We first describe how we used a transcendental phenomeno-
logical approach in the study design, data collection, and initial analysis,
including phenomenological design strategies such as bracketing for Boone
Blanchard’s 2014 dissertation. We then describe how we returned to the data
to carry out a directed content analysis of the data, informed by a racial
socialization theoretical lens, which led to the Blanchard et al. (2019) publi-
cation in the Journal of Child and Family Studies. We used an iterative
approach to gain a deeper understanding of the parental perspectives and
experiences. We discuss additional methods to support rigor, such as mem-
ber checks, an audit trail, and thick descriptions.
Little research had focused on African American parents’ experiences
with their toddler boys, and the first author, Sheresa, chose a phenomeno-
logical approach for two reasons. First, phenomenology is used by research-
ers to describe the essence of a phenomenon by exploring it from the
perspective of those who have experienced it (Teherani et al., 2015). Second,
a phenomenological approach allows researchers to probe deeply into areas
where research is limited so that the voices of underrepresented individuals
and their lived experiences are heard (van Manen, 1997). By examining an
experience as it is subjectively lived, researchers can develop new meanings
and appreciation for the phenomenon of study.

DOI: 10.4324/9781003254041-7
94 Sheresa Boone Blanchard et al.

There are different approaches to phenomenology research. Two

approaches considered for phase one of the study (Blanchard, 2014) were
transcendental (descriptive) phenomenology and hermeneutic (interpretive)
phenomenology, each rooted in distinct ways of conceiving the what and how
of human experiences. Choosing a phenomenological research methodology
required Sheresa to reflect on the research philosophy she embraced.
Transcendental phenomenology—the methodology that she used—examines
the fundamental components of a phenomenon and exposes the richness and
complexity of these components, including their implicit dimensions, with
assertions that are supported and validated intuitively by the data collected
(Finlay, 2009). This approach allowed for a deep inquiry into the complexi-
ties and nuances of parenting African American male toddlers.
One important aspect of a transcendental approach, which we discuss in
more detail in the next section, is to understand the “what” and “how” of
human experiences. Because there were not a lot of recent publications on
African American parents of toddlers, it was important to engage in deep
inquiry about what the parents were experiencing, feeling, thinking, and
doing. The interview protocol also included probing questions to understand
how the parents engaged in parenting, including socialization experiences,
parenting decisions, priorities, and hopes. Although questions included some
possible racialized ways parents might make parenting decisions, aspects of
race were probed more directly in phase two.
Directed content analysis is used when “existing theory or prior research
exists about a phenomenon that is incomplete or would benefit from further
description” (Hsieh & Shannon, 2005, p. 1281). This kind of analysis helped
us engage in a deep secondary analysis of the transcripts used for the Blanchard
et al. (2019) publication. Because racial socialization was integrated into the
research questions, it provided a logical theoretical framework for analysis in
phase two to determine the occurrence of racial socialization examples in the
interview content. That is, this approach offered a powerful way to understand
African American parents’ subjective experiences and gain insight into the
content of, prevalence of, and rationale for racial socialization (i.e., parental
messages to their children addressing the significance and meaning of race
and ethnicity), as well as the methods and modes by which parents use, or plan
to use, racial socialization with their toddler sons. Our findings suggested that
although participating parents did not typically engage in promoting mistrust,
their engagement in common racial socialization tenets, such as preparation
for bias, cultural socialization, or egalitarianism, varied. Blanchard et al.
(2019) focused on the importance of expounding the cultural context of par-
enting in the early years because most of the previous published articles
focused on school-­age children (kindergarten through high school).
 Phenomenological Analysis and Racial Socialization Interpretation 95

Phase One: Transcendental Phenomenological Design

Edmund Husserl, considered the founding philosopher of phenomenology,
stressed the importance of learning about the experiences of others in the
most unadulterated way possible so that the essence, or meaning, that partici-
pants ascribe to a phenomenon is identified (Bogdan & Biklen, 1992; Dowling
& Cooney, 2012; Gallagher & Zahavi, 2012; Gearing, 2004; Husserl, 1964;
Moustakas, 1994). This study employed a transcendental phenomenological
approach that helped researchers learn what participants have in common
and that, by nature, “describes the meaning for several individuals of their
lived experiences of a concept or a phenomenon” (Creswell, 2007, p. 57).
Because the approaches, methods, and processes for engaging in phenom-
enological research vary widely, the hallmark of a high-­quality phenomeno-
logical study is not necessarily the chosen approach, method, or process but
rather the explicit description of, justification for, and adherence to such
choices. We agree with other scholars that consideration of this specificity
should occur from the conceptualization of the study through data collec-
tion, analysis, and interpretation (Dowling, & Cooney, 2012; Moustakas,
1994; Pereira, 2012; Wojnar & Swanson, 2007). A transcendental phenome-
nological approach extends basic qualitative research features by adding
more specific attributes such as focusing on the whole human experience,
engaging in a quest for finding the essence of experience, and gathering
information from those who experienced the same phenomenon (Dowling &
Cooney, 2012; Moustakas, 1994).
Transcendental phenomenology is tied to Husserl’s concepts of intention-
ality, intuition, and the epoché process. Intentionality relates to one’s con-
sciousness (Moustakas, 1994) and means to be present with one’s thoughts
and being, as well as the events going on around oneself (Dowling & Cooney,
2012; Moustakas, 1994; Wojnar & Swanson, 2007). Intuition refers to consid-
ering a phenomenon from one’s own mind rather than from how someone
else might think of or perceive it. Also, with intuition, one is aware of the
presence of an essence in an experience (Moustakas, 1994). Discussing the
meaning of essences from Husserl’s point of view, Dahlberg (2006) explains
that essences are structures of meaning that explain what a phenomenon is by
illuminating the essential characteristics of the phenomenon. The epoché pro-
cess, also known as bracketing or phenomenological reduction, involves “set-
ting aside predilections, prejudices, [and] predispositions” about a topic to see
it afresh (Moustakas, 1994, p. 85). Considering these concepts in transcen-
dental phenomenology and engaging in the epoché process were important to
helping us identify the essence of African American child-­rearing practices
for toddler sons directly from African American parents. For example, during
data collection and initial analysis, Sheresa bracketed her own knowledge
96 Sheresa Boone Blanchard et al.

and opinions about the phenomenon of African American parenting, which

we describe in the next section.

Bracketing Method Choice (Epoché Process)

Bracketing elements include (a) the type of information that will be brack-
eted, (b) when bracketing will occur, and (c) how the researcher will system-
atically engage in bracketing. Since the introduction of bracketing in
Husserl’s (1931) early work, many variations in the type and scope of brack-
eting have emerged related to qualitative research (Creswell, 2007; Gearing,
2004; Tufford & Newman, 2012). Our research study included reflexive, or
cultural, bracketing (Gearing, 2004).
The key to reflexive bracketing is the researcher’s engagement in identify-
ing and noting their internal suppositions to guard against their interference
in the study. Internal suppositions include “personal values, judgments, cul-
ture, and history” (Gearing, 2004, p. 1445). For example, Sheresa reflected on
the connections between her current life circumstances and identities and
those of the participants. Being an African American married mother of two
young sons, it was important to be aware of her own values and culture to
ensure that her experiences were not conflated with those of the participants.
Although the researcher can make their internal suppositions apparent,
external suppositions (e.g., phenomenon) are harder to circumvent because
they relate to phenomena such as “its history, definition, and larger environ-
mental factors” (p. 1434) and are more difficult to set aside. Additionally,
inclusion of “world suppositions” (e.g., culture, more broadly, society) rele-
vant to the phenomenon is important in analysis (Gearing, 2004).

Reflexive Bracketing
Bracketing can occur at various points in study development and execution
(Conklin, 2007; Finlay, 2009; Gearing, 2004; Tufford & Newman, 2012). In
this study, bracketing occurred during initial planning and again during
analysis and interpretation. Because Sheresa was the primary study devel-
oper and data collector, she engaged in bracketing. For example, she reflected
upon societal perceptions of African American families as the historical
context as well as the existing body of research that often characterizes
African Americans as single-­parent families with absent fathers. Thus, part
of the purpose of Blanchard (2014) was to develop a counter-­narrative
through the voices of African American two-­parent households.
Through typed reflexive bracketing and reflection, Sheresa described her
internal suppositions (Gearing, 2004; Moustakas, 1994) as they related to
the beliefs and child-­rearing practices of parents of African American
 Phenomenological Analysis and Racial Socialization Interpretation 97

toddler sons and supported emotional and social development. With exter-
nal suppositions, she acknowledged perceptions of the context and culture
of African American parents with a male toddler, while these suppositions
were also set aside to understand participants’ descriptions of their lived
experience more fully. In other words, Sheresa considered societal and his-
torical perceptions of African American parents and children, reflected
upon her own experience as an African American married mother of sons,
and concurrently focused on the individual experiences of the participants
through in-­depth interviews that supported sharing their own experiences,
hopes, dreams, and perceptions. Table 6.1 shows an example of bracketing
written during the data collection for the study, which was not included in
the 2019 publication. Although bracketing was not edited, Sheresa reflected
on and reread the bracketing narrative throughout the study, including dur-
ing data analysis, to minimize the possibility of influencing data collection
and analysis with personal thoughts and allow participants’ lived experi-
ences to remain the focus.
Additionally, at the time of data collection and analysis, Sheresa’s per-
sonal home situation was important to consider and attempt to bracket.
When the interviews began, she had two African American sons—a seven-­
year-­old and a four-­year-­old. Although they were no longer toddlers, one of
them had been so about two years previously, so historical parenting knowl-
edge could impact perceptions about parenting, race, and gender norms.

TABLE 6.1 First Author Bracketing Excerpt

I am the youngest child of married African American parents, both born in

different parts of North Carolina but initially meeting in New Jersey. After
meeting in 1969, they married in 1971 and had their first child, my older brother,
in 1973. Both parents have some college, worked full-­time in several occupations
throughout their careers, and retired from Verizon Telephone Company in
different departments. We lived in a three-­bedroom home in a suburban
neighborhood of New Jersey, not far from New York City, for most of my life. I
attended Catholic school for almost all of my elementary through high school
years. According to my parents, they made the investment in private school
because they felt it would provide a higher-­quality educational experience for my
brother and me, who both participated in local paid and complementary
extracurricular activities, including Little League baseball, Girl Scouts, and piano
and dance lessons. Reading and books were important from an early age. My
parents were dedicated to trying to build a solid life for their family.
My extended family was an integral part of my childhood. We gathered with family
members on the paternal and maternal sides for holidays and other special events
like birthdays and dance or music recitals. My maternal grandmother was the
extended family member we spent the most time with.
98 Sheresa Boone Blanchard et al.

Moreover, there were similarities between Sheresa’s home life and aspects of
the participants’ home lives. During data collection and analysis for Blanchard
(2014), Sheresa would remind herself to focus on the interview questions and
open-­ended questions that would minimize conflating participant experi-
ences with her own. When there were potential misinterpretations, Sheresa
would ask questions to probe and better understand the participants.

Method
To provide more context about the study, we share details here about the par-
ticipants and methods. Participants were six couples (25–43 years old) residing
in a southeastern city in the United States. We employed snowball sampling to
identify and give voice to participants who could best provide relevant informa-
tion so we could report findings authentic to the participants’ lived experiences.
Data collection measures included three in-­depth, 15-­question interviews per
couple (one with each parent separately and one with the couple together); a
demographic form; and an observation with the family in a natural environ-
ment (e.g., home or community setting). We developed the interview based on
reviews of literature on toddler development, emotional and social develop-
ment, the integrative model of child development (i.e., Garcia Coll et al. 1996;
see its focus on children of color), and African American parenting (Table 6.2).

[SEE TABLE 6.2 at eResource: Sources for Interview Questions].

Questions focused on the child, family, and related systems such as regional,
neighborhood, and societal inquiries. Interviews lasted between 45 and 90
minutes. Finally, the observation added a layer of triangulation to the inter-
view content. During the observation (30–60 minutes), Sheresa took descrip-
tive and reflective notes on family interactions, conversations, toddler
activities, and descriptions of what was happening (e.g., eating dinner and
reading a book). For additional details about study participants and data
collection methods, see the full article (Blanchard et al., 2019).

Phase One Analysis

During the final phase of data analysis for the parent study, we used
Moustakas’ (1994) modified van Kaam method, which resulted in individual
textural and structural (Table 6.3) descriptions for each couple, family sum-
maries (Table 6.4; fictitious names were used to protect confidentiality), a
descriptive summary (Table 6.5), and a composite description for all partici-
pants combined (Blanchard, 2014). In the van Kaam method, the researcher
begins with the data (transcripts and field notes in our study) and completes
eight steps as part of the analysis:
 Phenomenological Analysis and Racial Socialization Interpretation 99

(a) listing and preliminary grouping; (b) reduction and elimination; (c)
clustering and thematizing the invariant constituents; (d) final identifica-
tion of the invariant constituents and themes by application–validation;
(e) construction of individual textural description; (f) construction of
individual structural description; (g) construct individual textural–struc-
tural description; and (h) construct a composite description.
(Moustakas, 1994, as cited in Blanchard, 2014, p. 57)

The first step, also known as the horizontalization process, included reviewing
transcripts, grouping content, and extracting significant statements (Conklin,
2007; Moustakas, 1994). Then, statements were combined to eliminate redun-
dancy. The next three steps involved creating more and more integrated sum-
maries or textural descriptions, including individual textural descriptions for
each family and individual structural descriptions for each family that

combines the individual textural description with my individual con-

sciousness of the phenomenon of raising an African American male tod-
dler … this would be the essence of the experience or the parts of the
experience that one cannot change without changing the experience.
(Blanchard, 2014, p. 58; see also Gallagher & Zahavi, 2012;
Moustakas, 1994)

The final summaries were textural–structural descriptions with constituents

and themes, a detailed description that included the parental and familial expe-
riences combined, and a composite description based on all study participants.

[SEE TABLE 6.3 at eResource: Excerpt from Family 1 Textural Description].

[SEE TABLE 6.4 at eResource: Excerpt from Family 1 Structural Description].

[SEE TABLE 6.5 at eResource: Excerpt from Family 1 Descriptive Summary].

During this stage of the analysis, we developed essences because in phenom-

enological studies, one of the most important methods of engaging with the
data and processing the transcript content is to develop essences. Following
the van Kaam method, the essences were developed by grouping the tran-
script contents and then creating two summaries for each family (textural
and structural) and one across families (composite). The textural summary
was developed by reading the transcripts across the three interviews for each
family several times and highlighting significant phrases and statements.
Table 6.3 is a textural summary excerpt for Family 1, which includes signifi-
cant statements and excerpts from the interview transcripts. Table 6.4 shows
a structural summary in which phrases and statements were expanded into
paragraphs, providing more detail about family daily life.
100 Sheresa Boone Blanchard et al.

Then, significant themes (e.g., excerpts or phrases) across all six families
were grouped to create essences. Seven essences described the lived experi-
ences of African American parents with toddler sons based on what the cou-
ples shared: (a) Fatherhood Motivation: Present, Active, and Different; (b)
Child Outcomes and Goals Influence Present Parental Decisions; (c) Lack of
Time = Elevated Stress and Guilt; (d) Extended Family/Friends Provide
Network of Support; (e) Current Perception of African American Male
Cases Shadow on Parenting and Childhood; (f) Importance of Faith and
Religion; and (g) Shared Household Responsibility/Division of Labor
(Blanchard, 2014). Essence excerpts were taken from both single and joint
interviews.

Phase Two: Rationale for Racial Socialization Lens

Phase two (Blanchard et al., 2019) focused on understanding the extent to
which racial socialization was a salient theme within the interviews. Racial
socialization refers to the process through which individuals learn about
race and racism intergenerationally and internalize such messages (Coard
et al., 2004). These teachings include the direct, explicit messages children
receive about the existence of racism and the meaning of race, as well as
related indirect or implicit messages. In addition to being a process whereby
parents and/or extended family members transmit inter-­and intra-­
generationally the significance and meaning of race and ethnicity, it also
involves the teaching of problem-­solving skills that enable children to be
flexible in their approach to race-­related situations and in their negotiation
of racially hostile environments (Coard & Sellers, 2005; Coard et al., 2007).
Racial socialization is a pervasive aspect of parenting in Black/African
American families, as children are exposed to messages about race from a
young age. As such, it is thought to be one of the most critical developmen-
tal processes for Black/African American youth (Lesane-­Brown, 2006;
Neblett et al., 2012). It supports the development of a strong and positive
racial and ethnic identity and encourages children to embrace and appreci-
ate their cultural heritage and identity. In addition to families, communities,
educational institutions, and media all play roles in shaping and transmit-
ting these messages (Jarrett, 1999).
It is important to note that racial socialization is a complex, multidimen-
sional construct (Lesane-­Brown, 2006) that consists of various kinds of
parental messages, activities, behaviors, and contexts (Bowman & Howard,
1985; Boykin & Toms, 1985; Hughes et al., 2006; Priest et al., 2013;
Stevenson, 1994; Umaña-­Taylor & Hill, 2020). Although some youth may
be raised in environments that promote racial tolerance, diversity, and
inclusivity, others may encounter messages that perpetuate stereotypes,
bias, or discrimination.
 Phenomenological Analysis and Racial Socialization Interpretation 101

For example, Coard and Sellers (2005) and Lesane-­Brown (2006) note that
the most common core dimensions of racial socialization are as follows:

Cultural socialization: This emphasizes the importance of cultural heritage,

values, practices, and pride. This dimension of racial socialization encour-
ages children to take pride in their cultural background and heritage and
involves celebrating the traditions, customs, and contributions of one’s
racial group.
Preparation for bias: This focuses on preparing children to cope with and
overcome challenges related to racism and discrimination. The realities
of inequalities between groups are highlighted, as is the importance of
navigating living in a society in which historical and present-­day racism
creates obstacles for them. Resilience is key to this dimension and to
empowering youth to navigate adversity while maintaining a positive
sense of self.

These tenets combined contribute to fostering a positive and informed

approach to racial identity development, promoting understanding and
resilience. In short, racial socialization influences children’s beliefs about the
way the world works, informs children’s beliefs and attitudes regarding “the
self,” helps shape children’s repertoire of strategies and skills for coping with
and navigating racism, and impacts the nature of the child’s inter-­and intra-­
racial relationships and interactions (Coard & Sellers, 2005).
Because we developed research questions with inclusion of racial social-
ization, the results contained themes of racial socialization throughout. Race
emerged as a major tenet in parenting for some participants, but for others, it
was less central to their parental goal and socialization strategy. But we
wanted to look more specifically at how racial socialization was experienced.
We were intrigued by the ways in which racial socialization showed up, par-
ticularly given that existing research posits that racial socialization is a rou-
tine and natural aspect of African American parenting (Coard et al., 2004;
Hughes et al., 2016; Lesane-­Brown, 2006; Priest et al., 2013), and interested
in giving voice to the ways in which the process of engaging in racial social-
ization happened for parents and the extent to which it aligned with the racial
socialization tenets (Hughes et al., 2006). Therefore, given the frequency of
parents describing racial socialization practices, we decided to use the racial
socialization theoretical lens to guide the content analysis in phase two.

Phase Two Analysis

We reanalyzed the interview transcripts from phase one using directed con-
tent analysis (Hsieh & Shannon, 2005) and informed by the concept of racial
socialization (Hughes et al., 2006). The goal was to understand how and to
102 Sheresa Boone Blanchard et al.

what extent parents engaged in the tenets of racial socialization (i.e., cultural
socialization, preparation for bias, promotion of mistrust, and egalitarian-
ism). Directed content analysis occurred around four subthemes to deepen
our understanding of parental racial socialization strategies: (a) prevalence,
(b) rationale for use of racial socialization (i.e., parents’ descriptions of the
reasons for engaging in racial socialization), (c) content of racial socializa-
tion messages (e.g., what parents said or planned to say in racial socializa-
tion messages), and (d) methods or modes of racial socialization (e.g., ways
parents described engaging in racial or cultural socialization). These ele-
ments of transmitting racial socialization were selected in order to under-
stand the degree to which parents were practicing racial socialization, why
they were using racial socialization, what they were actually sharing with
their children, and how they did so.
The directed content analysis was done by Sheresa and Stephanie, the first
and second authors, who are female African American doctorate-­level research-
ers. This team coded transcripts in Microsoft Word by highlighting and docu-
menting statements that relate to the tenets of racial socialization, using
different colors based on the racial socialization tenets. We assigned the codes
to each excerpt by adding a comment and entering which code was applicable.
We confirmed subtheme definitions following conferencing through virtual
meetings because we did not live in the same place, and we re-­coded and dis-
cussed any inconsistencies until we achieved 100 percent interrater agreement
in our decisions about racial socialization subthemes. For example, when there
was disagreement, we discussed our perspectives about which code was appli-
cable and reviewed the racial socialization tenets again. We then continued our
discussion until we reached a decision about which code was most applicable
for the excerpt. During this analysis, we found that parents did engage in racial
socialization with their toddlers, which extended the documented age of racial
socialization in the literature. Parental racial socialization processes were evi-
dent, as we identified the presence of racial socialization beliefs and practices
of African American parents of toddler sons. Parents make decisions daily that
influence and shape their children for adult life. It is one of the most challeng-
ing and important tasks of adulthood (Ryan & Padilla, 2019). For parents of
color, this task involves additional burdens and greater complexities as they
attempt to navigate and negotiate the terrain of racism, prejudice, and discrimi-
nation (see Blanchard et al., 2019, for more details about study results).

Establishing Reliability and Validity

Guba (1981) and Guba and Lincoln (1981) proposed four criteria to estab-
lish trustworthiness in naturalistic inquiry: (a) credibility (internal validity),
(b) transferability (external validity), (c) dependability (reliability), and (d)
 Phenomenological Analysis and Racial Socialization Interpretation 103

confirmability (intersubjective agreement). However, in their reconceptual-

ization of validity in qualitative research, Whittemore et al. (2001) distin-
guished between criteria as quality standards in qualitative research and
techniques or methods to decrease the risk of validity threats. They proposed
refining the criteria into primary and secondary levels. Primary criteria are
those required for all qualitative research to establish trustworthiness—the
credibility, authenticity, criticality, and integrity of the data collected, proce-
dures, analysis, and interpretation. Secondary criteria include explicitness,
vividness, creativity, thoroughness, congruence, and sensitivity as additional
benchmarks of validity. Inquiries employing qualitative research utilize
techniques or methods that demonstrate validity criteria.
We addressed credibility with researcher positionality through reflexive
bracketing in the design of the study, analysis, and interpretation of results.
We also used triangulation in our data analysis through information used in
transcript content, checks to verify the accuracy of transcripts, and written
summaries (Whittemore et al., 2001). Member checking was achieved by
inviting participants to review the summaries and suggest any revisions or
highlight inaccuracies. One participant identified a section of the summary
he felt was inaccurate, and it was revised until contents were agreeable and
accurate.
Table 6.6 lists the approaches to support rigor and the strategies used. To
address dependability and confirmability in our study, we explicitly described
the recorded review of methodological decisions and rationale throughout
the process, as well as observation notes from interactions with the partici-
pants, to ensure researcher influences had been accounted for and eliminated
to the highest extent possible. These notes throughout the process of data
collection constituted our audit trail.
Because transferability includes thick descriptions, the van Kaam method
of phase one data analysis supported this process. As described earlier, a
multiple summary process was used to ensure the process of going from

TABLE 6.6 Strategies to Determine Rigor

Approaches to rigor Strategies

Credibility Triangulation
Peer debriefing
Member checking
Dependability Audit trail
Reflexivity
Confirmability Audit trail
Reflexivity
Transferability Thick descriptions
104 Sheresa Boone Blanchard et al.

transcripts to essences was rigorous and true to what participants intended.

Aspects of the essences can be noted in the summaries for each family
(Tables 6.3–6.5).

Publication Journey
Our 2019 publication had a long and winding road from data collection to
publication. Table 6.7 outlines the many submissions and decision points
from the initial draft manuscript to acceptance and publication. Several
aspects of this journey might be important to qualitative researchers. Going
from Sheresa’s dissertation to publication was the first big hurdle, as there
was so much information that was not appropriate or relevant for a

TABLE 6.7 Manuscript Publication Journey

Timeline Activity

1/2015 Sheresa worked on identifying dissertation content to write a

manuscript focused specifically on African American parents of
male toddlers. There were many drafts of this paper.
8/2015 The submission guidelines from the Child Development journal were
reviewed in preparation for submission.
9/2015 A draft was submitted to the Child Development journal.
11/2015 We received reviewer feedback along with a decision notification to
revise and resubmit the manuscript. The feedback discussed the
merit of the content but pointed out that the manuscript needed to
reflect a developmental framework. The editor suggested that we
consider resubmitting the manuscript if the revisions reflected the
developmental approach required by January 15, 2016.
1/2016 We resubmitted the manuscript along with a written explanation.
The submission included a letter with all revisions to the
manuscript that were incorporated into the resubmission.
2/2016 We received reviewer feedback along with a decision to not accept
from the editor. The editor commented that the paper was too long
and read like two papers—a conceptual one and an empirical
report. It was suggested that we take part of the paper and submit
it as a brief report.
8/2016 During the six months from the 2/2016 submission to this 8/2016
submission, the Child Development journal had an editor change.
We cannot recall why we made this decision (likely due to a new
editor), but we submitted the full (revised) manuscript again, not a
brief report. We received a desk rejection from the editor and a
recommendation to submit to the Parenting journal.
9/2016 We submitted the manuscript to the Parenting journal. The
manuscript was rejected.

(Continued)
 Phenomenological Analysis and Racial Socialization Interpretation 105

TABLE 6.7 (Continued)

Timeline Activity
9/2016 We submitted the manuscript to the Journal of Child and Family
Studies at the end of September. The manuscript was revised
throughout several cycles over a period of two years. The revisions
suggested by reviewers include a more specific connection to racial
socialization through directed content analysis and reviewing
additional literature.
9/2018 The manuscript was accepted for publication.
2018 The manuscript was published online in November.

publication. Another challenge was finding a theoretical framework, like

racial socialization in this case, to better understand the data and organize it
into a publishable manuscript. Finally, the team was challenged by editors
and reviewers who (a) found merit in the manuscript but could not give con-
structive feedback, (b) did not agree on revisions, or (c) appeared not well-­
versed in qualitative research and therefore had unrealistic expectations for
the data (e.g., wanting a larger sample because they assumed that partici-
pant numbers should be more similar to quantitative study sample sizes). It
took several years for us to overcome these challenges and have our manu-
script published.
Besides practicing perseverance, this review process allowed us to learn
several lessons about ourselves and publishing. Although there were times
when we wanted to give up, editor and reviewer comments and locating a
journal that was a better fit led to publication. One of the biggest lessons we
learned was how to clarify and strengthen our qualitative analysis process to
describe it in enough detail and organize our results in a manner that made
them easier to understand. For example, an early submission of the paper
did not include the direct content analysis phase, and racial socialization was
integrated into another theme. After rereading the transcripts and discuss-
ing their content several times, we landed on how racial socialization was
represented in what the parents shared, which was not a surprise because it
was one of the areas of the Garcia-­Coll et al. (1996) model informing the
research questions. Phase two was instrumental in revising our manuscript
for publication and strengthening the organization of the results.

Conclusion
As previously mentioned, transcendental phenomenology is a good fit for
in-­depth qualitative studies in that “annotating and clarifying human expe-
rience can be a challenging task not only because of the complexity of
106 Sheresa Boone Blanchard et al.

human nature, but also because an individual’s experience is a multidimen-

sional phenomenon, that is, psychologically oriented, culturally driven, and
socially structured” (Alhazmi & Kaufmann, 2022, p. 1). This approach
allows for deep inquiry into topics that are relevant for participants based on
their lived experiences. In our study, we were able to understand the partici-
pating couples’ experiences and give voice to families. We also saw similari-
ties across families related to the intersections of race, culture, socioeconomic
status, and gender that illustrated the experiences of the families. However,
because we felt the phenomenological analysis and essences were not ade-
quate to capture parental perspectives, we applied the tenets of racial social-
ization in a directed content analysis to understand families’ lived experiences
and organize families’ preparation for and reaction to the racialized society
in the United States.
We support a phenomenological approach and directed content analysis
as effective approaches to understanding individuals’ lived experiences and
giving them voice, particularly for those in underrepresented groups. These
approaches help to shake up the status quo and not make assumptions based
on the researcher’s agenda. They also give voice to lived experience, and
although the result may highlight societal tensions, it allows the opportunity
for reflection with implications for programs and policies.

KEY WORKS GUIDING OUR DATA ANALYSIS

Gearing, R. E. (2004). Bracketing in research: A typology. Qualitative Health Research,

14(10), 1429–1452. [Link]
The detailed descriptions about bracketing in this article supported the strate-
gies the team used to engage in bracketing. The author defines bracketing while
clarifying its origins and discussing seminal work that informs this practice. In addi-
tion, six different forms of bracketing are outlined to provide readers with a clear
understanding of bracketing and ideas for how to engage in this practice.
Hsieh, H.-F., & Shannon, S. E. (2005). Three approaches to qualitative content analysis.
Qualitative Health Research, 15(9), 1277–1288. [Link]
1049732305276687
This article assisted the team with more clearly describing their qualitative analy-
sis process. The authors provide information about three approaches to content
analysis: conventional, directed, and summative. The article also includes tech-
niques for the qualitative analysis process.

References
Alhazmi, A. A., & Kaufmann, A. (2022). Phenomenological qualitative methods
applied to the analysis of cross-­cultural experience in novel educational social
contexts. Frontiers in Psychology, 13, Article 785134. [Link]
fpsyg.2022.785134
 Phenomenological Analysis and Racial Socialization Interpretation 107

Blanchard, S. B. (2014). Understanding beliefs and practices of African American par-

ents with male toddlers: A focus on emotional and social development [Doctoral
dissertation, University of North Carolina at Greensboro]. [Link]
edu/ir/uncg/f/Blanchard_uncg_0154D_11538.pdf
Blanchard, S. B., Coard, S. I., Hardin, B. J., & Mereoiu, M. (2019). Use of parental
racial socialization with African American toddler boys. Journal of Child and
Family Studies, 28(2), 387–400. [Link]
Bogdan, R. C., & Biklen, S. K. (1992). Qualitative research for education: An intro-
duction to theory and methods (2nd ed.). Allyn and Bacon.
Bowman, P. J., & Howard, C. (1985). Race-­related socialization, motivation, and
academic achievement: A study of Black youths in three-­generation families.
Journal of the American Academy of Child Psychiatry, 24(2), 134–141. [Link]
org/10.1016/S0002-­7138(09)60438-­6
Boykin, A. W., & Toms, F. D. (1985). Black child socialization: A conceptual frame-
work. In H. P. McAdoo & J. L. McAdoo (Eds.), Black children: Social, educa-
tional, and parental environments (pp. 33–51). Sage.
Coard, S. I., Wallace, S. A., Stevenson, H. C., & Brotman, L. M. (2004). Towards
culturally relevant preventive interventions: The consideration of racial socializa-
tion in parent training with African American families. Journal of Child and Family
Studies, 13(3), 277–293. [Link]
Coard, S., & Sellers, R. (2005). African American families as a context for racial
socialization. In V. McLoyd, N. Hill, & K. Dodge (Eds.), Emerging issues in
African-­American family life: Context, adaptation, and policy. Guildford Press.
Coard, S. I., Foy-­Watson, S., Zimmer, C., & Wallace, A. (2007). Considering cultur-
ally relevant parenting practices in intervention development and adaptation: A
randomized control trial of the Black Parenting Strengths and Strategies (BPSS)
Program. The Counseling Psychologist 35(6), 797–820. [Link]
0011000007304592
Conklin, T. A. (2007). Method or madness: Phenomenology as knowledge creator.
Journal of Management Inquiry, 16(3), 275–287. [Link]
1056492607306023
Creswell, J. W. (2007). Qualitative inquiry and research design: Choosing among five
traditions (2nd ed.). Sage.
Dahlberg, K. (2006). The essence of essences – the search for meaning structures in
phenomenological analysis of lifeworld phenomena. International Journal of
Qualitative Studies on Health and Well-­ being, 1(1), 11–19. [Link]
10.1080/17482620500478405
Dowling, M., & Cooney, A. (2012). Research approaches related to phenomenology:
Negotiating a complex landscape. Nurse Researcher, 20(2), 21–27. [Link]
org/10.7748/nr2012.[Link].c9440
Finlay, L. (2009). Debating phenomenological research methods. Phenomenology &
Practice, 3(1), 6–25. [Link]
Gallagher, S., & Zahavi, D. (2012). The phenomenological mind (2nd ed.). Routledge.
García Coll, C. G., Lamberty, G., Jenkins, R., McAdoo, H. P., Crnic, K., Wasik, B.
H., & García, H. V. (1996). An integrative model for the study of developmental
competencies in minority children. Child Development, 67(5), 1891–1914. https://
[Link]/10.2307/1131600
108 Sheresa Boone Blanchard et al.

Gearing, R. E. (2004). Bracketing in research: A typology. Qualitative Health

Research, 14(10), 1429–1452. [Link]
Guba, E. G. (1981). ERIC/ECTJ annual review paper: Criteria for assessing the
trustworthiness of naturalistic inquiries. Educational Communication and
Technology, 29(2), 75–91.
Guba, E. G., & Lincoln, Y. S. (1981). Effective evaluation (1st ed.). Jossey-­Bass.
Hsieh, H.-F., & Shannon, S. E. (2005). Three approaches to qualitative content anal-
ysis. Qualitative Health Research, 15(9), 1277–1288. [Link]
1049732305276687
Hughes, D., Del Toro, J., Harding, J. F., Way, N., & Rarick, J. R. D. (2016). Trajectories
of discrimination across adolescence: Associations with academic, psychological,
and behavioral outcomes. Child Development, 87(5), 1337–1351. [Link]
10.1111/cdev.12591
Hughes, D., Rodriguez, J., Smith, E. P., Johnson, D. J., Stevenson, H. C., & Spicer, P.
(2006). Parents’ ethnic-­racial socialization practices: A review of research and
directions for future study. Developmental Psychology, 42(5), 747–770. [Link]
org/10.1037/0012-­1649.42.5.747
Husserl, E. (1931). Ideas: General introduction to pure phenomenology. Macmillan.
Husserl, E. (1964). The phenomenology of internal time-­ consciousness. Indiana
University Press.
Jarrett, R. L. (1999). Successful parenting in high-­risk neighborhoods. Future of
Children, 9(2), 45–50.
Lesane-­Brown, C. L. (2006). A review of race socialization within Black families.
Developmental Review, 26(4), 400–426. [Link]
Moustakas, C. (1994). Phenomenological research methods. Sage.
Neblett, E. W., Jr., Rivas-­Drake, D., & Umaña-­Taylor, A. J. (2012). The promise of
racial and ethnic protective factors in promoting ethnic minority youth develop-
ment. Child Development Perspectives, 6(3), 295–303. [Link]
j.1750-­8606.2012.00239.x
Pereira, H. R. (2012). Rigour in phenomenological research: reflections of a novice
nurse researcher. Nurse Researcher, 19(3), 16–19. [Link]
[Link].c9054
Priest, N., Paradies, Y., Trenerry, B., Truong, M., Karlsen, S., & Kelly, Y. (2013). A
systematic review of studies examining the relationship between reported racism
and health and wellbeing for children and young people. Social Science &
Medicine, 95, 115–127. [Link]
Ryan, R. M., & Padilla, C. M. (2019). Transition to parenthood. In M. H. Bornstein
(Ed.), Handbook of parenting, volume 3, being and becoming a parent (3rd ed., pp.
513–555). Routledge.
Stevenson, H. C. (1994). Validation of the Scale of Racial Socialization for African
American Adolescents: Steps toward multidimensionality. Journal of Black
Psychology, 20(4), 445–468. [Link]
Teherani, A., Martimianakis, T., Stenfors-­Hayes, T., Wadhwa, A., & Varpio, L.
(2015). Choosing a qualitative research approach. Journal of Graduate Medical
Education, 7(4), 669–670. [Link]
Tufford, L., & Newman, P. (2012). Bracketing in qualitative research. Qualitative
Social Work, 11(1), 80–96. [Link]
 Phenomenological Analysis and Racial Socialization Interpretation 109

Umaña-­Taylor, A. J., & Hill, N. E. (2020). Ethnic–racial socialization in the family:

A decade’s advance on precursors and outcomes. Journal of Marriage and Family,
82(1), 244–271. [Link]
van Manen, M. (1997). Researching lived experience: Human science for an action
sensitive pedagogy (2nd ed.). Althouse Press.
Whittemore, R., Chase, S. K., & Mandle, C. L. (2001). Validity in qualitative research.
Qualitative Health Research, 11(4), 522–537. [Link]
1129119299
Wojnar, D. M., & Swanson, K. M. (2007). Phenomenology: An exploration. Journal
of Holistic Nursing, 25(3), 172–180. [Link]
7
BLACK FEMINIST THEORY
AND THEMATIC ANALYSIS
Analyzing the Motherwork of Black Women
Nursing Professionals during the
COVID-19 Pandemic

Adrienne L. Edwards-­Bianchi

During the COVID-­ 19 pandemic, People of Color, especially Black

Americans, experienced more severe health outcomes than white Americans,
exacerbated by systemic racism within healthcare settings, such as the
decades-­long delivery of inadequate healthcare services to Black communi-
ties (Carroll, 2020). Black Americans were also more likely to experience
severe illness and die from COVID-­ 19 illness than non-­Hispanic white
Americans (Centers for Disease Control and Prevention [CDC], 2020).
Throughout the COVID-­ 19 pandemic, media outlets raised national
­awareness of the interrelatedness of systemic racism in healthcare and the
disproportionate numbers of Black Americans impacted by COVID-­19;
however, few, if any of them, addressed how Black healthcare professionals
experienced the pandemic. Nurses were important in containing COVID-­19
outbreaks in communities and providing medical care (Chen et al., 2020).
Black women nurses, however, had to contend with racism within the work-
place, such as racial microaggressions, while providing care that increased
their levels of emotional distress (Thomas-­Hawkins et al., 2022).
In this chapter, I reflect on and explain how I analyzed interview data
from six Black women nursing professionals who were working during the
first wave of the COVID-­19 pandemic. One major theme, “educating as
motherwork,” explained the women’s experiences. Two subthemes, “educat-
ing our own” and “balancing education with advocacy,” described processes
related to how educating as motherwork happened among the women. More
specifically, the subthemes explained how Black women nursing profession-
als embraced a community mothering role in Black communities and used
that role to educate those communities about making lifestyle changes to

DOI: 10.4324/9781003254041-8
 Black Feminist Theory and Thematic Analysis 111

reduce the impact of COVID-­19. Further, the women advocated for better
healthcare for Black communities while negotiating workplace racism. The
women’s education and advocacy reflected an interrelatedness of Black con-
sciousness and their understanding of their identities as Black women. An
in-­depth explanation of the findings of the study is published in Families,
Systems, & Health (Edwards, 2021).

Positionality Statement
I am a Black woman family science scholar who has devoted her career to
examining and explaining Black women’s mothering. Mostly, my work has
centered on Black women’s mothering in connection to the racial social-
ization of Black children. During the first wave of the COVID-­19 pan-
demic, however, my scholarly interests turned to how the pandemic was
impacting Black mothering. Three of my relatives died due to COVID-­19-­
related complications during the first wave of the pandemic. As painful
and tragic as the losses were, they were not unusual. As media coverage
about the disproportionate numbers of Black Americans seriously ill or
dying from COVID-­19 began to grow, it became clear that Black families
and communities were dealing with a health crisis underscored by decades
of systemic racism. Moreover, as Black America was reeling from the
COVID-­19 pandemic, Mr. George Floyd, an unarmed Black man, was
murdered by Derek Chauvin, a white police officer. Global Black Lives
Matter (BLM) protests followed the murder. As I watched the protests, I
realized that Black women were on the frontlines, organizing and leading
protests while also caring for family members impacted by COVID-­19.
The work of Black women protestors catalyzed me to ruminate about how
Black women nursing professionals might be mobilizing themselves to
help Black communities. Nurses were important to curbing the spread of
COVID-­19 (Chen et al., 2020), and given the importance of Black women
to the survival of Black communities (Collins, 1994; Dill, 1988), I wanted
to learn about what the work of Black women nurses looked like during
the COVID-­19 pandemic.

Sociopolitical Context of the Study

I collected data between May 2020 and July 2020 after the United States was
slowly emerging from a nationwide lockdown to reduce the spread of
COVID-­19. At that time, 97.9 out of every 100,000 Black Americans had
died from COVID-­19 as compared to 64.7 per 100,000 Latino Americans,
46.6 per 100,000 white Americans, and 40.4 per 100,000 Asian Americans
(Vasquez Reyes, 2020). The differences in the numbers of deaths drew
112 Adrienne L. Edwards-Bianchi

attention to widespread racism and systemic inequalities, particularly in

access to affordable, quality healthcare that underscored those deaths
(Williams & Cooper, 2019; Vasquez Reyes, 2020). For example, Black women
were overrepresented among the numbers of “essential workers” who had to
continue working during the nationwide lockdown, which placed them at
greater risk of contracting the virus while simultaneously facing a lack of
access to quality healthcare, including medical professionals who sometimes
disregarded their symptoms (Ubinna, 2021). The overrepresentation of
Black Americans in COVID-­19 illness revealed how generations of struc-
tural and systemic racism had created barriers to adequate healthcare for
them. The United States experienced a period of racialized awakening in
which it realized that it was experiencing “dual pandemics:” the COVID-­19
pandemic and a racism pandemic that merged to contribute to negative out-
comes for Black Americans (Jones, 2021). Both major macrosystemic events
helped contextualize the experiences of Black Americans that year. To better
understand the sociopolitical context of the study, I reviewed the burgeon-
ing literature on COVID-­19, as well as the more robust literature on systemic
racism in healthcare, nursing, and health disparities.

A Black Feminist Theoretical Perspective

I used Black feminist theory because it applies an Afrocentric, cultural lens
to produce knowledge about Black women’s experiences within society,
their families, and communities that captures the simultaneous oppres-
sions (e.g., racism, classism, and sexism) they navigate (Collins, 2000).
Further, the theory explains how those oppressions may look different
based on the sociohistorical contexts and events in which they occur (King,
1995). Black feminist theory functions as an epistemological tool for Black
women scholars to use to resist Eurocentric theories that neglect how
political and economic structures in the United States have created social
echelons that marginalize Black women and create a different reality for
them that Black men, white women, and white men do not experience
(Collins, 2000; Simien, 2004). Unlike traditional feminist approaches,
Black feminist theory accounts for the shared experiences of racism among
Black women and men while highlighting how patriarchy within Black
communities creates a layer of intragroup oppression, further complicat-
ing Black women’s lived experiences (Collins, 2000). The reality of being
relegated to a low social status means that Black women experience mul-
tiple burdens that characterize how they perceive themselves, understand
and fulfill their roles within their families, and survive within their com-
munities and in the larger society. From a healthcare perspective, examples
of burdens that Black women negotiate include systemic racism in
 Black Feminist Theory and Thematic Analysis 113

healthcare (Carroll, 2020) that leads to high rates of maternal and infant
mortality (Chinn et al., 2021), health disparities such as higher incidences
of breast cancer (Eldridge & Berrigan, 2022), and medical professionals
who delay or refuse to diagnose them with COVID-­19 despite their show-
ing symptoms of the illness (Obinna, 2021). Although Black feminist the-
ory accounts for the dual inter-­and intragroup oppressions that Black
women experience, it also provides a strengths-­ based perspective that
allows Black women scholars to investigate Black women’s “everyday acts
of resistance” (Collins, 2000, p. 184) to the burdens posed by living at the
junctions of multiple oppressions.
The theory also allowed me to place Black women nursing professionals’
experiences in a sociohistorical context by helping me exercise Black con-
sciousness throughout data collection and analysis. Black consciousness
refers to centering the sociopolitical experiences of Black women and their
families in research studies (Few, 2007; Few et al., 2003), including how their
identity shapes how they understand those experiences (Okech & Harrington,
2002). This means that I interpreted all data by accounting for the leadership
roles that Black women have historically held in Black communities
(Edwards, 2022), the role that Black women have had in building a sense of
community and belonging among Black people even during times of slavery
(Dill, 1988), and Black women’s efforts to educate communities to promote
the survival of Black children and families (i.e., motherwork; Collins, 1994).
I used Black consciousness to acknowledge the collective action and resis-
tance against systemic racism that Black women nursing professionals dem-
onstrated during the COVID-­19 pandemic while advocating for better care
for Black patients.

Using Black Feminist Theory to Engage Intersectionality

Black feminist theory helped me to explore intersectionality in the study.
Intersectionality, the multilayered and simultaneous oppressions that arise
from interconnected race, class, gender, and sexuality identities that shape
the realities of Black women (Crenshaw, 1993), is at the core of Black femi-
nist theory and praxis. Although Black feminists had long demonstrated an
intersectional understanding of Black women’s experiences, Crenshaw
(1993) articulated the term “intersectionality” that has become widely used
within and across disciplines and with populations other than Black women.
Engaging intersectionality allowed me to examine how Black women
understood their identities within a specific context (Few-­Demo, 2014;
McCall, 2005), such as a healthcare setting. Tapping into the intersectional-
ity component of Black feminist theory allowed me to practice intersec-
tional consciousness in addition to Black consciousness in data collection
114 Adrienne L. Edwards-Bianchi

and analysis. Intersectional consciousness refers to an active awareness of

how social structures interact to create oppressive experiences and how that
oppression manifests itself in inter-­and within-­group interactions (Brown
et al., 2017; Tormos, 2017). I used intersectional consciousness to pay atten-
tion to how Black women nursing professionals interacted with Black
patients and strategized about how to interact with white colleagues to
improve healthcare for them.

Doing Research with Black Women

I reached out to a Black woman nurse practitioner whom I knew through
collaborating with her on a community-­based project a year before the pan-
demic started to assist me with recruiting participants. After recruitment, I
had a snowball sample of six Black women nursing professionals from
California, North Carolina, and South Carolina. I collected data through
phone interviews. See the publication in Families, Systems & Health
(Edwards, 2021) for more method details, including the full interview guide.
During the interviews, I was careful not to assume that being a Black
woman automatically granted me an “insider status” to accurately under-
stand and interpret the participants’ experiences. Instead, I centered them
in the interview process by prioritizing their voices and building rapport
with them through appropriate and timely self-­disclosure (Few et al., 2003).
Appropriate self-­disclosure can reveal the researcher’s vulnerabilities (Allen
& Piercy, 2005) but can also be an effective tool to complement interview
techniques. For example, I shared about my own experiences with systemic
racism in healthcare and how a Black woman nurse had advocated for me
to receive better quality care. I noted how I reacted to each participant and
their experience. Such reactions included my emotional response to the par-
ticipants’ narratives, my own experiences in the healthcare system as a
Black woman, and differences between my lived experience and those of the
participants. I noted my reaction to each participant because I wanted to
hold myself accountable for how my experiences intersected with those of
the participants. Recognizing shared as well as divergent experiences with
participants allowed me to practice self-­reflexivity by acknowledging and
challenging how my lived experiences influence the data analysis process
(Allen, 2000; Few et al., 2003). More specifically, through practicing self-­
reflexivity, I dismantled “blind spots” stemming from my own intersecting
identities that subconsciously impacted how I analyzed the data (Few,
2007). Practicing self-­reflexivity allowed me to be transparent about the
data analysis process which, “in turn,” contributed to the trustworthiness
of the findings.
 Black Feminist Theory and Thematic Analysis 115

Coding and Thematic Analysis

Thematic analysis is a “method for identifying, analyzing, and reporting pat-
terns (themes) within data” (Braun & Clarke, 2006, p. 79). It is a flexible quali-
tative analysis tool that is applicable across theoretical orientations and yields
rich, detailed interpretations of data (Braun & Clarke, 2006; Kiger & Varpio,
2020). Qualitative methods such as thematic analysis also allow researchers to
examine the “cultural distinctions between and within groups” (Few et al.,
2003, pp. 206–207), which is important for doing research with Black women.
Other scholars have used thematic analysis in health-­related research with
Black women (e.g., see Chandler et al., 2021; Manchikanti Gomez & Wapman,
2017; West et al., 2016). Using thematic analysis in conjunction with Black
feminist theory helped me identify detailed patterns that explain how Black
consciousness and collective action operated across a dataset of Black women.
I analyzed data using three cycles of coding methods associated with the-
matic analysis (Saldaña, 2016). First cycle coding was a combination of in
vivo and process coding. Second cycle coding was focused coding. Third cycle
coding involved identifying subthemes and major themes in the data.
Explanations of the types of codes associated with first cycle coding and how
I engaged each cycle of coding are provided in detail later in this section.

Getting Started with Data Analysis

I used notetaking during interviews as the first step in the data analysis pro-
cess. In addition to my reaction to the participants, I noted my initial
thoughts about possible codes and references to sensitizing concepts during
each interview. I also listened to each audio-­recorded interview multiple
times. This allowed me to pay specific attention to voice inflection, nuances
in participants’ word usage, pauses, and moments of silence in the inter-
views—all of which were indications about the emotional and intellectual
ways Black women nursing professionals processed their experiences during
the COVID-­19 pandemic. After listening to the interviews, I personally tran-
scribed them into Word documents. I created individual file folders for each
participant and placed a transcribed interview and participant biosketch
into each folder. I transcribed six interviews and had 44 single-­spaced pages
of text. Then, I read each transcript several times. Audio recording each
interview, listening to each interview multiple times, transcribing each inter-
view myself, and reading each interview multiple times helped me become
very familiar with the data—something that other researchers such as Allen
and Roberto (2019) have noted.
I used sensitizing concepts or “starting points” (Charmaz, 2003, p. 259)
associated with Black feminist theory in my coding. I have used Black
116 Adrienne L. Edwards-Bianchi

feminist theory in most of my research with Black women and their families
and was familiar with sensitizing concepts that would be applicable to the
study at hand. I identified the following sensitizing concepts: motherwork,
Black consciousness, and power. I used motherwork to conceptualize how
Black women nurses used their understanding of the pivotal role of mothers
in Black communities to educate and promote the survival of Black com-
munities during the COVID-­19 pandemic (Collins, 1994). Educating others
is a critical component of motherwork (Johnson, 2013). I used Black con-
sciousness not only to account for the sociohistorical experiences of the
women during data analysis but also to examine how the women accounted
for the sociohistorical experiences of Black Americans in their nursing prac-
tices. Power referred to the control that Black women nurses felt like they
had over the education and health of Black patients and communities. I used
these sensitizing concepts to help me construct codes, identify relationships
between codes, and explain processes (Charmaz, 2003).

Cycles of Coding
After familiarizing myself with the data, I started coding by placing codes
into comment boxes in Word documents that contained the transcribed
interviews. In each cycle of coding, I allowed the participants’ interviews to
tap into my own Black consciousness about how Black communities have
suffered abuses in the US healthcare system, and I noted those abuses in the
coding. For example, each woman in the study discussed how the abuses of
the Tuskegee Syphilis Experiment, which left hundreds of Black men with
untreated syphilis (Jones, 1993), helped to create the resistance that some
Black Americans have toward the US healthcare system. The participants’
interpretations of the impact of the experiment on Black Americans were
captured in a first cycle code and labeled “understanding racial disparities.”
By practicing Black consciousness in my data analysis, I identified partici-
pants’ actions that reflected a collective understanding of the sociohistorical
experiences of Black Americans (Few, 2007). I exercised intersectional con-
sciousness when identifying aspects in women’s interviews that reflected
their understanding of or resistance to the gendered racism they experienced
by white women and men colleagues who underestimated their capabilities
or tried to reduce their access to patients. Examples of those codes were
“proving myself ” and “opening up doors.”
In the first cycle of coding, I went through each line of each transcript, iden-
tifying in vivo and process codes that reflected participants’ actions related to
caring for patients and how they perceived their roles as Black women in nurs-
ing. In vivo codes are direct quotes from participants that represent the authen-
ticity of their responses, and process codes use gerund words that end in “ing”
to capture participant behaviors or behaviors that are related to theoretical
 Black Feminist Theory and Thematic Analysis 117

concepts (Charmaz, 2003; Saldaña, 2016). I focused on those types of codes

because they allowed me to capture participants’ actions that were related to
the sensitizing concepts (i.e., motherwork, Black consciousness, and power).
Further, those codes helped me conduct an intersectional analysis by alerting
me to participant quotes and processes that explained what Black women’s
racialized and gendered identities looked like and how they were expressed dur-
ing the first wave of the COVID-­19 pandemic. In using in vivo and process
codes, I captured how the participants’ views on their identities as Black women
and the racism they experienced while working in healthcare settings influenced
how they interacted with white colleagues. Table 7.1 shows some examples of
in vivo and process codes during this first stage of coding.
I placed first cycle codes in comment boxes in each transcript and high-
lighted the interview text associated with each code. I also added subcodes
(in separate comment boxes) that further defined, explained, or provided
examples of each first cycle code. Further, I noted my initial thoughts about
processes that may explain how women made sense of their experiences and
how codes may be related to each other by adding explanations to excerpts
of text from transcripts in comment boxes (Allen & Roberto, 2019).
Document 7.1 shows an early iteration of a coded transcript. The first cycle
of coding yielded 20 codes. Because thematic analysis is about finding the
richest and minimal number of codes and themes that best represent partici-
pants’ experiences (Braun & Clarke, 2006; Saldaña, 2016), I reused codes
across transcripts that captured the same experiences and actions among the
participants. I also created new codes, such as “combating misinformation,”
when participants introduced a new experience to the overall dataset.

[SEE DOCUMENT 7.1 at eResource: First Draft of Coded Transcript—Phoenix].

I created a separate Word document where I listed each first cycle code and
excerpts from transcripts that represented each code. Additionally, I included
the pseudonym of the participant associated with each excerpt to help paint
a picture of how each woman conveyed her experiences. In doing so, I began
creating a coding tree that helped me visualize what each code meant as well
as emerging relationships between codes. From the coding tree, I began to

TABLE 7.1 Examples of First Cycle Codes

In vivo codes Process codes

Know your place Using my voice

Turf war Proving myself
It’s still all political Explaining
Fill in the gaps Avoiding stereotypes
118 Adrienne L. Edwards-Bianchi

see that caring for Black patients was not a straightforward process for Black
women in nursing. The coding tree revealed a depiction of concurrent pro-
cesses of educating Black patients on lifestyle choices while proving their
ability as nurses before being able to advocate for changes in healthcare
delivery and policy. Figure 7.1 shows the final version of the coding tree that
depicts how the codes coalesced together to explain the motherwork of
Black women nursing professionals.

[SEE FIGURE 7.1 at eResource: Final Coding Tree: Major Theme, Subthemes, and
Focused Codes].

I wrote analytic memos after the third and sixth interviews to help transition
from the first cycle to the second cycle of coding that included identifying and
grouping the most salient first cycle codes together. Analytic memos included
my initial thoughts about processes that were happening across participants
and potential focused codes that may capture those processes (Saldaña, 2014).
The first analytic memo (see Document 7.2) reflected not only my initial ideas
for codes but also how the women’s experiences resonated with me as a Black
woman and catalyzed me to think about the impact of systemic racism on
Black people’s health. The second analytic memo (see Document 7.3) cap-
tured my thoughts about similarities and differences across the interviews. For
example, I noted my thoughts about similarities in participants’ experiences,
thoughts, and activities across interviews as well as added new information
that had been uncovered. I included each participant’s pseudonym to help me
remember which specific interviews were covered in the analytic memo, and I
included quotes from participants that illustrated relationships among codes.
I also used memo writing to make notes about what to include in the Discussion
section of the manuscript, including any connections to the sensitizing con-
cepts. Additionally, I reread excerpts from interview transcripts to double-­
check the relationships between codes that I was identifying in the interviews.
In some instances, I listened to the audio recordings again to make sure the
first cycle codes were correct. The analytic memos coupled with the coding
tree helped me strategize how to decide which first cycle codes were most
salient and could be grouped together into concise yet more comprehensive
second cycle codes (i.e., focused coding). I engaged in an iterative process of
grouping and regrouping the first cycle codes together. I paid particular atten-
tion to first cycle codes that only appeared one time across transcripts and
recoded or eliminated them from the second cycle coding. In the second cycle
of coding, the number of codes decreased, leaving those focused codes that
were most essential to explaining the women’s experiences. After I identified
the focused codes, I updated the coding tree to reflect the reorganized and
condensed codes.
 Black Feminist Theory and Thematic Analysis 119

[SEE DOCUMENT 7.2 at eResource: Analytic Memo: Phoenix, Ella, and Tracy].

[SEE DOCUMENT 7.3 at eResource: Analytic Memo: Thrive, Farrah, and Sasha].

In the third cycle of coding, I began “themeing the data” by identifying sub-
themes and eventually one major theme that explained what the first cycle
codes meant (Saldaña, 2016). To transition from second cycle coding to the-
meing the data in third cycle coding, I moved beyond identifying relation-
ships between codes to explaining the processes that were represented in
those relationships. During this coding transition, I made data analysis
notes that reflected my thinking about how those processes explained the
steps Black women engaged in healthcare settings to support Black commu-
nities. I revisited the comment boxes in my transcripts, which led me to
reread interview excerpts and identify direct quotes representing processes
that connected one code to another. This meant paying close attention to
how participants interpreted their actions and interactions with others. In
doing so, I re-­engaged the sensitizing concepts and ruminated on how they
explained or contextualized those interpretations. Table 7.2 is an example of
the data analysis notes that I wrote about the subthemes. A single major
theme, “educating as motherwork,” provided the best overall explanation for
participants’ behaviors across the dataset, whereas the subthemes “educat-
ing our own” and “balancing education with advocacy” explained the pro-
cesses that the Black women engaged in to educate and protect Black
communities. In themeing the data, I identified the processes that were read-
ily apparent in the data as well as those underlying processes that I had
gleaned from the common experiences among the women across the dataset
(Boyatzis, 1988). Once again, I engaged in Black consciousness to explain
how the sociohistorical experiences of Black Americans provided the mac-
rosystemic context in which Black women nursing professionals developed
educational strategies to use with Black patients.

Navigating the Analysis Process as the Sole Data Analyst

As the only data analyst on this research study, I took steps to reduce how
my biases influenced my interpretations of the data. I used member checking
not only as a way of checking the accuracy of my codes and themes but also
to show the women that I was invested in understanding and accurately rep-
resenting their voices (Few et al., 2003). Due to the limited availability of the
women, I conducted member checks during each interview. During member
checking, I summarized my interpretations of each woman’s experience as
well as patterns that I started recognizing across interviews and asked for
feedback (Harvey, 2015). This method of member checking allowed me to
120 Adrienne L. Edwards-Bianchi

TABLE 7.2 Data Analysis Notes for Subthemes

Subtheme: Educating Our Own

Black women felt an increased responsibility to educate Black patients and
communities on lifestyle changes to make to reduce the impact of COVID-­19.
Teaching Black patients how to treat diabetes with dietary and exercise changes
instead of just relying on medications was a common thread across all
participants.
Giving Black people trusted information about COVID-­19 and reaching them in
their communities are key to reducing the impact of COVID-­19 on Black people.
In order to reach the Black community, you have to be a trusted voice.
Misinformation about COVID-­19 has complicated Black nurses’ and NPs’
(Nurse Practitioners efforts to educate the Black community about the virus.
Having more providers that look like the Black people they serve is important.
Improving health literacy though education from someone Black people can trust
means someone who looks like them. Using social media is a means to reach
Black communities and combat misinformation about COVID-­19. Because of
the history of distrust for the medical community and racial injustices from the
medical community against Black people, some Black people may not readily
engage in contact tracing or answer questions about their behaviors or
interactions that may have led to their infection with COVID-­19.
Subtheme: Balancing Education with Advocacy
Nurses and NPs negotiate racialized healthcare spaces themselves, but they also
help their patients negotiate those spaces by educating them and advocating for
them. For Black women nurses and NPs, educating Black patients starts with
their own negotiation of white, male-­dominated spaces by being the only or one
of a few Black women nurses or NPs in their professional settings.

check the accuracy of my codes in real time while empowering participants

by allowing them to be involved in the analytical process and maintain
power over how their stories were told.
In addition to member checking, I triangulated the data through an inte-
grated literature review. I reviewed literature on systemic racism in healthcare,
Black women’s studies, and family science as secondary sources of data. The
articles and book chapters in the literature review added to the credibility of
my research by explaining Black women’s experiences with racism in health-
care settings and their centralized importance in Black families.
Note taking and memo writing helped me practice “active self-­
reflexivity” (Few, 2007, p. 462) by allowing me to check how my subjectivi-
ties influenced my interpretation of data at each cycle of the data analysis
process. For example, those activities allowed me to identify and account
for how I related to the participants as a Black woman. Further, it allowed
me to address how I connected participants’ perceptions of the substan-
dard care that Black people received during the first wave of the COVID-­19
 Black Feminist Theory and Thematic Analysis 121

pandemic to substandard care they had received in other health-­related

areas such as maternal healthcare.

Prioritizing Black Women’s Voices over Interrater Reliability

In the peer review process of the article, a reviewer argued that I should
add at least one other analyst to code the transcripts and come to agree-
ment with that analyst about codes and themes. Although I recognized
that interrater reliability is useful, I knew it was not the only way to deter-
mine if codes are accurate. I responded to the reviewer that prioritizing
interrater agreement among multiple data analysts who bring their own
subjectivities and blind spots about Black women and their lived experi-
ences to the data analysis process over accurate interpretation of data
may lead to findings that silence the voices of Black women. Instead of
relying on interrater reliability, I followed the suggestions of Morse
(1997), who argued that interrater reliability is a quantitative way of
doing qualitative research that disrupts the interpretive process; there-
fore, qualitative researchers should rely on their familiarity with the data
and trust the coding processes they used. I prioritized Black women’s
voices that may have been compromised if coming to agreement with
other data analysts had been necessary. The steps I took to reduce bias
allowed me to prioritize the voices of Black women while trusting my
coding procedures and knowledge of Black feminist theory and thematic
analysis.

Conclusion
In this chapter, I have chronicled the data analysis process of a research
study on the experiences of Black women nursing professionals during the
first wave of the COVID-­19 pandemic. In writing this chapter, the impor-
tance of exercising Black consciousness and intersectional consciousness in
research with Black women became even more clear to me. Doing research
with Black women, particularly subgroups of Black women, includes inte-
grating appropriate theoretical and analytical approaches that will accu-
rately account for within-­ group differences. Integrating Black feminist
theory with thematic analysis better captured the nuances of the women’s
experiences more accurately than any Eurocentric theory or single method
could have done. Using Black consciousness and intersectional conscious-
ness allowed me to place the women’s experiences in a sociohistorical con-
text while paying attention to how their identities as Black women influenced
their interactions with other Black people and white colleagues. I prioritized
the women’s voices over interrater reliability by trusting my coding and tak-
ing steps to reduce researcher bias. In conclusion, this chapter sheds light on
122 Adrienne L. Edwards-Bianchi

the specific attention that researchers must pay to the choice of appropriate
theory and analytical methods when doing research with Black women.

KEY WORKS GUIDING MY DATA ANALYSIS

Collins, P. H. (1994). Shifting the center: Race, class, and feminist theorizing about
motherhood. In E. N. Glenn, G. Chang, & L. R. Forcey (Eds.), Mothering: Ideology,
experience, and agency (pp. 45–65). Routledge.
In this book chapter, Collins introduced the concept of motherwork. I was capti-
vated by how she explained what mothering looks like for Women of Color includ-
ing mothering children who are not biologically their own. I believe motherwork
captured the attitudes and behaviors of Black women nursing professionals.
Few, A. L. (2007). Integrating Black consciousness and critical race feminism into family
studies research. Journal of Family Issues, 28(4), 452–473. [Link]
0192513X0629733
I used Few’s discussion of Black consciousness to explain how Black women
nurses placed Black communities’ sociohistorical experiences with medicine and
healthcare systems at the center of their patient and community education and
advocacy. Few provided guidelines on how family scholars can use Black conscious-
ness in their writing. I used those guidelines as I was writing the Discussion section
of my article.
Saldaña, J. (2016). The coding manual for qualitative researchers (3rd ed.). Sage.
I relied on and tried to follow Saldaña’s interpretation of how to code interview
data in my data analysis. Particularly, I followed his suggestions for using in vivo and
process codes with interview data and how to move from one cycle of coding to the
next. I found Saldaña’s suggestions on using memo writing and writing explana-
tions for codes while coding very beneficial. Additionally, I used knowledge I gained
from taking two Qualitative Summer Research Intensive courses from Saldaña that
are based on his book to guide my data analysis. The courses are offered through
the Odum Institute for Research in Social Science at the University of North Carolina
at Chapel Hill.

References
Allen, K. R. (2000). A conscious and inclusive family studies. Journal of Marriage
and the Family, 62(1), 4–17. [Link]
Allen, K. R., & Piercy, F. P. (2005). Feminist autoethnography. In. D. H. Sprenkle &
F. P. Piercy (Eds.), Research methods in family therapy (2nd ed., pp. 155–169).
Guilford Press.
Allen, K. R., & Roberto, K. (2019). Sharing expertise in Appalachia: A collaborative
feminist content analysis of in-­depth interviews with older women cancer survi-
vors. In Á. M. Humble, & M. E. Radina (Eds.), How qualitative data analysis
happens: Moving beyond “themes emerged” (pp. 64–79). Taylor & Francis.
Boyatzis, R. E. (1988). Transforming qualitative information: Thematic analysis and
code development. Sage.
Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative
Research in Psychology, 3(2), 77–101. [Link]
 Black Feminist Theory and Thematic Analysis 123

Brown, M., Ray, R., Summers, E., & Fraistat, N. (2017). #sayhername: A case study
of intersectional social media activism. Ethnic and Racial Studies, 40(11), 1831–
1846. [Link]
Carroll, A. E. (2020). Health disparities among Black persons in the United States
and addressing racism in the health care system. Journal of the Medical Association
Health Forum, 1(6), e200769. [Link]
Centers for Disease Control and Prevention (CDC). (2020, November 30). COVID-­19
hospitalization and death by race/ethnicity. Centers for Disease Control and
Prevention. [Link]
discovery/hospitalization-­death-­by-­race-­[Link]
Chandler, R., Guillaume, D., Parker, A., Mack, A., Hamilton, J., Dorsey, J., &
Hernandez, N. D. (2021). The impact of COVID-­ 19 among Black women:
Evaluating perspectives and sources of information. Ethnicity & Health, 26(1),
80–93. [Link]
Charmaz, K. (2003). Grounded theory: Objectivist and constructivist methods. In N.
K. Denzin & Y. S. Lincoln (Eds.), Strategies for qualitative inquiry (2nd ed.), pp.
249–291). Sage.
Chen, S. C., Lai, Y. H., & Tsay, S. L. (2020). Nursing perspectives on the impact of
COVID-­19. The Journal of Nursing Research, 28(3), e85. [Link]
NRJ.0000000000000389
Chinn, J. J., Martin, I. K., Redmond, N. (2021). Health equity among Black women
in the United States. Journal of Women’s Health, 30(2), 212–219. [Link]
10.1089/jwh.2020.8868
Collins, P. H. (1994). Shifting the center: Race, class, and feminist theorizing about
motherhood. In E. N. Glenn, G. Chang, & L. R. Forcey (Eds.), Mothering:
Ideology, experience, and agency (pp. 45–65). Routledge.
Collins, P. H. (2000). Black feminist thought: Knowledge, consciousness, and the poli-
tics of empowerment (2nd ed.). Routledge.
Crenshaw, K. (1993). Demarginalizing the intersection of race and sex: A Black fem-
inist critique of antidiscrimination doctrine, feminist theory, and antiracist poli-
tics. In D. Weisberg (Ed.), Feminist legal theory: Foundations (pp. 383–411).
Temple University Press.
Dill, B. T. (1988). Our mothers’ grief: Racial ethnic women and the maintenance of
families. Journal of Family History, 13(4), 415–431. [Link]
036319908801300404
Edwards, A. L. (2021). Educating during the COVID-­19 pandemic: The motherwork
of Black women nursing professionals. Families, Systems, & Health, 39(4), 599–
608. [Link]
Edwards, A. L. (2022). (Re)Conceptualizing Black motherwork as political activism.
Journal of Family Theory & Review, 14(3), 404–411. [Link]
12469
Eldridge, L., & Berrigan, D. (2022). Structural racism and triple-­negative breast can-
cer among Black and White women in the United States. Health Equity, 6(1),
116–123. [Link]
Few, A. L. (2007). Integrating Black consciousness and critical race feminism into
family studies research. Journal of Family Issues, 28(4), 452–473. [Link]
org/10.1177/0192513X0629733
124 Adrienne L. Edwards-Bianchi

Few, A. L., Stephens, D. P., & Rouse-­ Arnett, M. (2003). Sister-­ to-­
sister talk:
Transcending boundaries and challenges in qualitative research with Black women.
Family Relations, 52(3), 205–215. [Link]
Few-­Demo, A. L. (2014). Intersectionality as the “new” critical approach in feminist
family studies: Evolving racial/ethnic feminisms and critical race theories. Journal
of Family Theory & Review, 6(2), 169–183. [Link]
Harvey, L. (2015). Beyond member-­checking: A dialogic approach to the research
interview. International Journal of Research & Method in Education, 38(1), 23–38.
[Link]
Johnson, L. (2013). Covert wars in the bedroom and nation: Motherwork, transna-
tionalism, and domestic violence in Black widow’s wardrobe and mother tongue.
Meridians, 11(2), 149–171. [Link]
Jones, J. H. (1993). Bad blood: The Tuskegee syphilis experiment (Rev. ed.) Free Press.
Jones, J. M. (2021). The dual pandemics of COVID-­ 19 and systemic racism:
Navigating our path forward. School Psychology, 36(5), 427–431. [Link]
org/10.1037/spq0000472
Kiger, M. E., & Varpio, L. (2020): Thematic analysis of qualitative data: AMEE
Guide No. 131. Medical Teacher, 42(8), 1–9. [Link]
2020.1755030
King, D. K. (1995). Multiple jeopardy, multiple consciousness: The context of Black
feminist ideology. In B. Guy-­Sheftall (Ed.), Words of fire: An anthology of African-­
American feminist thought (pp. 294–314). The New Press.
Manchikanti Gomez, A., & Wapman, M. (2017). (Under) implicit pressure: Young
Black and Latina women’s perceptions of contraceptive care. Contraception,
96(4), 221–226. [Link]
McCall, L. (2005). The complexity of intersectionality. Signs, 30(3), 1771–1800.
Morse, J. (1997). “Perfectly healthy, but dead”: The myth of inter-­rater reliability.
Qualitative Health Research, 7(4), 445–447. [Link]
104973239700700401
Obinna, D. N. (2021). Essential and undervalued: Health disparities of African
American women in the COVID-­19 era. Ethnicity & Health, 26(1), 68–79. https://
[Link]/10.1080/13557858.2020.1843604
Okech, A. P., & Harrington, R. (2002). The relationships among Black conscious-
ness, self-­esteem, and academic self-­efficacy in African American men. Journal of
Psychology, 136(2), 214–224. [Link]
Saldaña, J. (2014). Coding and analysis strategies. In P. Leavy (Ed.), The Oxford
handbook of qualitative research (1st ed., pp. 581–685). Oxford University Press.
Saldaña, J. (2016). The coding manual for qualitative researchers (3rd ed.). Sage.
Simien, E. M. (2004). Black feminist theory: Charting a course for Black women’s
studies in political science. Women & Politics, 26(2), 81–93. [Link]
J014v26n02_04
Thomas-­Hawkins, C., Zha, P., Flynn, L., & Ando, S. (2022). Effects of race, work-
place racism, and COVID worry on the emotional well-­being of hospital-­based
nurses: A dual pandemic. Behavioral Medicine, 48(2), 95–108. [Link]
10.1080/08964289.2021.1977605
Tormos, F. (2017). Intersectional solidarity. Politics, Groups, and Identities, 5(4), 707–
720. [Link]
 Black Feminist Theory and Thematic Analysis 125

Ubinna, D. N. (2021). Essential and undervalued: Health disparities of African

American women in the COVID-­19 era. Ethnicity & Health, 26(1), 68–79. https://
[Link]/10.1080/13557858.2020.1843604
Vasquez Reyes, M. (2020). The disproportional impact of COVID-­19 on Black
Americans. Health and Human Rights Journal, 22(2), 299–307.
West, L. M., Donovan, R. A., & Daniel, A. R. (2016). The price of strength: Black
college women’s perspectives on the strong Black woman stereotype. Women &
Therapy, 39(3–4), 390–412. [Link]
Williams, D. R., & Cooper, L. A. (2019). Reducing racial inequalities in health: Using
what we already know to take action. International Journal of Environmental
Research and Public Health, 16(4), 606–632. [Link]
8
VISUALIZING RELATIONSHIPS TO
EXPLORE OPPORTUNITIES FOR
FAMILY ENGAGEMENT IN DIABETES
AND HYPERTENSION MANAGEMENT
Meredith P. Fort, Cornelia J. Santos, Maria de los
Angeles Villaverde, and Kelly R. Moore

Our study (Fort et al., 2020) was focused on understanding family engage-
ment in diabetes and hypertension management at an Urban Indian Health
Organization (UIHO) in Albuquerque, New Mexico. We employed an
inductive qualitative approach with 23 English-­and Spanish-­ speaking
American Indian and Latino adults with a dual diagnosis of type 2 diabetes
and hypertension and 13 family members or care partners identified by the
patients. Our analysis identified opportunities for and challenges to family
engagement with the potential to inform care strategies. Key challenges that
we identified for family engagement in disease management were that it is
common for family members to not have a family member readily available
to support them, family members who are supportive may live far away, and
there are barriers to sharing medical information between patients and fam-
ily members. Proposed strategies included: (a) increasing family engagement
at home and in the clinic, (b) increasing education about type 2 diabetes and
hypertension for family members, and (c) recognizing that patients and fam-
ily members benefit from opportunities to learn from and help others both
in group settings and individually.

Background of the Study

UIHOs, which primarily serve American Indians residing in urban settings,
offer a variety of culturally responsive health programming and typically
offer ancillary support services. Additionally, many UIHOs are sources of
care for underserved patients who are not American Indian/Alaska Native,
which is the case for First Nations Community HealthSource (FNCH) in

DOI: 10.4324/9781003254041-9
 Visualizing Relationships in Diabetes and Hypertension Management 127

Albuquerque. These patients include immigrants who are primarily from

Mexico and Central America, the underinsured, and homeless individuals.
Founded in 1972, FNCH is the largest UIHO in the United States and was
designated as a Federally Qualified Health Center in 1997 (Moore et al.,
2023). FNCH is committed to providing primary care as well as other ser-
vices such as Medicaid enrollment, job placement, housing support, alterna-
tive medicine, and traditional healing/ceremonies. FNCH’s dedication to
offering a broad range of services and reaching marginalized populations
presents a supportive setting for studies involving community members.
The topic of family engagement and its possibilities for improved pre-
vention and care is key to chronic disease management because individuals
tend to develop, maintain, and share health-­related behaviors in the context
of their family for diseases such as diabetes and hypertension (Campbell,
2005). One reason is that family members may share similar behaviors for
eating, physical activity, and consumption of unhealthy substances. Another
reason is that chronic disease self-­care and management have been shown
to improve when family members are engaged in patient care (McDaniel
et al., 2005; Rosland, 2009). Patients who do not have regular access to care,
such as underserved and underrepresented patients, may gain the most
from family support.
The idea to study family engagement and employ genograms and eco-
maps in this project came from Meredith’s previous experience in Chiapas,
Mexico about a cardiovascular disease prevention project in community
health centers (Fort et al., 2015). For our Albuquerque study, in addition to
genograms and eco-maps, we used timelines to capture key moments in each
patient’s experience with diabetes and hypertension. We decided to include
timelines because we recognized that they added a dynamic element to the
study. Although genograms and eco-maps captured family relationships and
resources at the time of the interview, timelines encouraged patients to reflect
on how their diabetes and hypertension evolved over time and whether fam-
ily support had changed.

Drawing on the Strengths of Our Diverse

and Multidisciplinary Team
A key value that we share is the importance of ensuring diverse perspectives
and representation in research. Our project team includes researchers with a
range of disciplinary and professional backgrounds as well as lived experi-
ences that allowed us to contribute different types of insight and under-
standing to all phases of the project. Our multidisciplinary team included
clinician researchers (internal medicine, pediatrics, and family medicine), a
professor of social work, and researchers trained in qualitative methods in
128 Meredith P. Fort et al.

anthropology, education, and public health. Furthermore, our team included

American Indian-­identifying researchers and clinicians, as well as Latina
researchers, including one who had immigrated to the United States from
Cuba. Additionally, our experienced, Spanish–English bilingual interviewer
with a health education background was crucial in establishing trust with
patients and their family/care partners. In the following paragraphs, we pro-
vide additional information about the positionalities of the four authors of
this chapter.
As an American Indian doctor with years of experience conducting
research and clinical care on reservations and in urban communities, Kelly
was familiar with the diversity of healthcare settings where American Indian
patients with diabetes and hypertension seek care. She lived in Albuquerque
during the time of the project and had worked with the Indian Health
Service on Navajo Nation as well as other reservations, allowing her to
understand patients’ urban experience while also feeling connected to family
and community far away. She also drew on her experiences as a practicing
pediatrician and as a researcher with experience analyzing data from previ-
ous diabetes and cardiovascular health projects.
Maria contributed her lived experience as a Cuban American as well as
her familiarity with the region that many recent immigrants to Albuquerque
come from, having served as a Peace Corps volunteer in Central America.
She also drew on her training in public health with a focus on communica-
tion and her experience living and working in Albuquerque with many
community-­based organizations. As the lead interviewer for the study, being
fully bilingual in Spanish and English allowed her to pick up on nuances and
references during the interviews that others could have missed. Her familiar-
ity with the local community as well as the broader region was important for
understanding the specific references that participants made about organiza-
tions and community resources, neighborhoods, pueblos and reservations,
and cities and towns abroad. During this project, she listened deeply to peo-
ple’s experiences and explanations and wanted to make sure that we were
conveying them accurately in our article.
Corrie (short for Cornelia) brought a background in education, health
disparities, and qualitative methods, as well as the perspective and lived
experience as an urban American Indian from Denver with connections to
the broader Southwest region. She joined the project to provide support
with coding, analyzing transcripts, and interpreting genograms and eco-­
maps. While reading the transcripts and coding, she would mention her own
broad family and community support networks and share insights from
American Indian scholars. She also drew on her previous experience con-
ducting interviews and qualitative analysis for her dissertation that focused
on cancer care networks and Native cancer survivors.
 Visualizing Relationships in Diabetes and Hypertension Management 129

As a public health researcher and practitioner from Seattle, Washington,

with many years of experience living and working in Guatemala, Meredith
had lived in Denver for only a few years when she started this study. She had
previously worked on a study on the topic of family engagement in Chiapas,
Mexico, and was interested in expanding that line of inquiry. Her years of
living and working in Central America and her familiarity with the broader
region allowed her to understand and interpret the interviews and visualiza-
tions in Spanish and English. In addition, she could consider and reflect on
the healthcare contexts that patients from other countries referred to and
how that could influence diabetes and hypertension self-­management.
Together, the four of us recognized the importance of community expertise
and creating space for underrepresented voices. We aimed to build that into
the study design, although we recognize that in the future, it would be ideal to
engage community members in the design and analysis phases of projects. We
also valued emphasizing family and community strengths in our study. To pre-
pare the article in the Journal of Health Care for the Poor and Underserved that
this chapter references (Fort et al., 2020), we worked with other experienced
collaborators and mentors who provided guidance and support.
Our study consisted of many elements: contacting participants, inter-
views, listening to recordings, preparing a codebook, reviewing and coding
transcripts, interpreting the genograms and eco-maps and sharing prelimi-
nary findings with small groups. Our team brought a diversity of disciplin-
ary perspectives and lived experiences that allowed us to collectively
contribute to all the study components with an openness to learn from one
another. During the analysis process, we began to learn more about each
other. For example, Maria shared about her experience as an immigrant,
spurred by comments made by some of the study participants who were
immigrants from the same country.

Reflecting on Interviews during Data Collection:

Initial Analysis
Maria conducted all but one of the interviews (in English and Spanish),
and Meredith conducted one interview (in English). Although Meredith and
Maria were both fluent in English and Spanish, not all the other team mem-
bers spoke both languages. Thus, to make the entire dataset accessible to all
team members, we sent out all the recordings to a transcription service that
had the capacity to translate the Spanish version into English.
Both the index patient (the person identified from clinical records as hav-
ing a dual diagnosis of diabetes and hypertension who was recruited to
participate in the study) and family member/support interviews were con-
ducted whenever a family member/support partner was identified. We then
130 Meredith P. Fort et al.

organized each interview into a physical folder that included the index
patient’s printed transcript, a family member/support partner’s transcript
(if one was identified), and the index patient’s (a) genogram, (b) eco-­map,
and (c) diabetes and hypertension timelines.
We began to immerse ourselves in the data and started the reflection and
analysis process early, recognizing that it is advisable to not wait until the
end of data collection because of the volume of data that we expected. For
example, as soon as the first few interviews had been conducted, Meredith
listened to the interview audio recordings to begin to get familiar with the
data. We found it helpful to listen to these recordings in one sitting—one
after another. We found this approach to be helpful for multiple reasons. As
studies progress, the quantity of data can be quite voluminous and over-
whelming. Starting to analyze data early helps to keep the workload man-
ageable, and it also helps keep participants’ distinctive and unique experiences
separate so that they do not run into each other. By dedicating time to famil-
iarize ourselves with each index patient and family support interview, we
were able to recognize details that we could reflect on when hearing the next
set of interviews.
During the interviewing phase, Meredith and Maria met over the phone
on a weekly basis to discuss progress and challenges, and sometimes Kelly
joined the meetings. Taking time to meet regularly throughout the study
and review interviewing experiences helped us confirm that the interview
questions were being interpreted appropriately. These regular meetings also
provided a way for us to share surprising experiences and common stories.
One example that stood out from this phase was the recognition that two of
the identified index patients turned out to be each other’s care partners.
They had both been listed as contacts to be recruited as index patients
because they both had a dual diagnosis of diabetes and hypertension, but
they also turned out to be each other’s primary support in managing their
conditions. Maria realized this when she contacted them over the phone to
recruit them and recognized the phone number as one that she had seen
before. This experience opened our eyes to the fact that many patients are
on the receiving end of support, and they also provide support to others, in
some cases for the same conditions. We took note of these kinds of reflec-
tions in our notebooks during the study and mentioned this as a finding in
our published study.

Codebook Development and Uploading Transcripts to

Qualitative Software
Prior to starting the interviews, Maria, Meredith, and Kelly developed a
codebook based on the interview guides, and we updated it soon after the
 Visualizing Relationships in Diabetes and Hypertension Management 131

first set of interviews. We prepared the codebook in a Microsoft Word docu-

ment and drew on coding guidance from an earlier edition of a qualitative
data analysis book by Miles et al. (2019). For example, we included codes
from the interview guides that were focused on the topics of interest for our
study. We then added subcodes to codes that were broad descriptions and
included brief definitions for each code. Table 8.1 includes examples of our
initial codes and definitions.
We created an [Link] project file and added our codes to this file.
[Link] is a qualitative data analysis software that Meredith had learned
to use during her doctoral studies in public health and had experience using
in other studies. As the transcription service returned completed transcripts,
Meredith uploaded them into this project file.
After Maria conducted more interviews, we discussed additional codes to
add to the codebook to make sure to capture new ideas that were shared
during the interviews. For example, we added a code that we referred to as
“younger generation” because several participants mentioned the younger
generation as a source of motivation for getting or staying healthy. Another
code that we added was “family tragedy.” We had anticipated death of ­family
members and had a code for that, including a distinction of deaths from
diabetes or hypertension coded as “death_relatedcause” as compared to
deaths from another cause which we coded as “death_unrelatedcause.”
However, some interview participants referred to family tragedies, such as
suicide, that influenced their ability to manage their conditions. We felt it
was important to further distinguish such events from the other two codes
related to a death, and so these were coded as “family tragedy.”

TABLE 8.1 Excerpts from the Initial Codebook

Code Definition

Facilitator Something that facilitates disease management

Family involvement Comments on there being family involvement
in supporting disease management
Family involvement_lack of Comments that there is a lack of family
involvement in supporting disease
management
Homeless Refers to being homeless
Insulin Refers to taking insulin
Meals_alone Eats meals alone
Meals_growingup Talks about typical meals growing up
Meals_shared Shares meals with other
Motivation Source of motivation for disease management
Physical activity Healthy behavior related to physical activity
132 Meredith P. Fort et al.

Visual Diagrams: Genograms, Eco-maps, and Timelines

Our analysis was aided by three types of visual diagrams: (a) genograms, (b)
eco-­maps, and (c) timelines. The genogram is a visual tool that is used to cap-
ture the complete list of family members, their ages and other relevant charac-
teristics, and key relationships between the index family member and others
(McGoldrick et al., 2008). For our project, we included the index patients’ par-
ents, siblings, spouse/partner (and ex-­partners), children, and grandchildren (if
any) on the genogram. We listed diabetes and hypertension for the index patient
and any other family members who had them. Family members residing in the
same household were included within a circle on the genogram.
Eco-­maps are a type of visual diagram used to document the major sys-
tems and resources that are a part of a person’s/family’s life and the nature
of the relationship to the different systems and resources (Hartman, 1995).
On the eco-­maps for this project, we documented key individuals (friends,
extended family, and work colleagues), systems that patients referenced as
enabling disease management (e.g., recreation, healthcare, and social ser-
vices), as well as existing challenges or conflicts in the relationships or chal-
lenges affecting their ability to access resources.
For the timelines, we documented when each patient was diagnosed with
hypertension and diabetes. We also identified other key moments during
their condition, including where they were living, what was happening at
that point in their lives, and if they had support from family members.
After all the interviews were completed, we took out the hand-­drawn ver-
sions of the genograms and eco-­maps and laid them out on desks and tables to
“meet” the people and understand their family and community context, look
for patterns and unique experiences, and consider perspectives that we had not
yet been able to include in our interviews. In some cases, it was almost over-
whelming to take in the experience of index patients who had multiple family
members with chronic conditions—often diabetes and/or hypertension—in
addition to challenging life events. We reviewed the diagrams multiple times
because they had a lot of information. We did not immediately see the key
points that we wanted to highlight or which genograms/eco-­maps that we
wanted to include in the manuscript on our first review. Our review of eco-­
maps and genograms was supported by Deb Ortega, a consultant and social
work professor with experience using genograms and eco-­maps in clinical prac-
tice and teaching. It was helpful to have her experienced eye recognize patterns
and pose questions for us to consider.
We used the software Genogram Analytics to create genograms for data
presentation. Figure 8.1 is an example of a genogram that was included in
Fort et al. (2020). This figure includes pseudonyms, and some information
was modified to protect participants’ and family members’ identities before
presenting the findings. For example, we made changes to the number of
 Visualizing Relationships in Diabetes and Hypertension Management 133

siblings on a genogram and modified people’s ages. We recognize that this is

especially important to do in a smaller community where it is easy for com-
munity members to identify one another. In addition, immigration status
can make people more vulnerable.

[SEE FIGURE 8.1 at eResource: Sample Genogram from Original Article (Fort et
al., 2020)].

Copyright © 2020 Meharry Medical College. This article first appeared in

the Journal of Health Care for the Poor and Underserved, Volume 31, Number
2, May 2020. Published with permission by Johns Hopkins University Press.

The review of visual diagrams helped our team see people who we inter-
viewed in terms of their multiple characteristics. Rather than a table that
captures individuals based on their individual characteristics (e.g., age, gen-
der, and severity of their condition), we could see their relationships to others
and the pattern of how they interacted with family members, especially as it
related to the conditions of interest. For example, we show in Figure 8.1 that
the index patient called Veronica had numerous siblings who had diabetes
and hypertension, as well as her mother and spouse. In completing the geno-
gram as a part of the interview, patients had all of their family members in
their mind, and they would often refer to them. In some cases, they would
reference key memories related to their diabetes or high blood p ­ ressure, such
as when a family member passed away or required medical attention. This
was important because it reinforced the multi-­generational nature of diabetes
and high blood pressure and indicated the importance of developing inter-
vention approaches going beyond providing support to individual patients.
Similarly, eco-­maps allowed us to visually capture the extent of an index
patient’s support network within the household, their extended family, and
in the community, and see the specific kinds of support and resources that
they tapped into. The sample eco-­map in Spanish (Figure 8.2) points to
Alcoholics Anonymous, the VA hospital, First Nations Community
HealthSource, church, a lawyer, food stamps, and family members and
friends as connections and sources of support. The eco-­map in English
(Figure 8.3) shows a larger family in the household as well as connections
and sources of support, including First Nations Community HealthSource,
church, family members, and friends.

[SEE FIGURE 8.2 at eResource: Sample Eco-­Map in Spanish (with Identifying

Information Removed or Covered)].

[SEE FIGURE 8.3 at eResource: Sample Eco-­Map in English (with Identifying

Information Removed or Covered)].
134 Meredith P. Fort et al.

After all the interviews had been completed, we selected topics from the
interviews to summarize in small group meetings with a subset of patients
and family members. We put together summaries from across the interviews
on types of family support, family engagement strategies, and community
and family resources (in the form of a composite eco-­map).

Small Group Meetings with Index Patients

and Family Members
Two small group meetings were conducted in person with participants and
family members onsite at FNCH in meeting rooms: one meeting in Spanish
and the other in English. These small group sessions were an opportunity to
share back with a subset of patients and their identified family support
members and listen to their insights and interpretations of the summaries
that we shared with them. The purpose of these meetings was to increase the
validity of the qualitative data, as explained by Cho and Trent (2006). This
type of analysis and feedback supports and encourages engagement beyond
a single interview and allows researchers to continue to learn from the
insights of study participants.
Drawing on the perspective of community-­based participatory research
(Wallerstein et al., 2019), we recognize that non-­researcher perspectives are
important to capture in analysis. We shared preliminary findings (Figure 8.4
and Document 8.1) and a version of the composite eco-­map (Figure 8.5) on
a whiteboard with patients and family members. Figure 8.4 shows a spec-
trum of types of family support. After we had completed all the interviews,
we recognized that people referred to different types of family support that
could be helpful to view on a spectrum. This was one of the summaries that
we shared in the small group sessions.

FIGURE 8.4 Spectrum of Types of Family Support

 Visualizing Relationships in Diabetes and Hypertension Management 135

[SEE DOCUMENT 8.1 at eResource: Summary Slide of Family Engagement

Strategies].

Document 8.1 shows an email from October 13, 2016, with a suggestion of
how to best present the family engagement strategies to the group discus-
sions. We had initially organized strategies as at home, at the health center,
and in the community, but this email shows a suggestion from one of the
team members to organize strategies into two lists: (a) strategies at home and
(b) strategies at the health center and in the community. Additionally,
although we prepared summaries on PowerPoint slides, we chose to present
them in the small group sessions on large papers that we taped up on the wall
to allow people to stand up and look at them up-­close and discuss them
together, rather than more passively watch a presentation.
After reviewing our composite eco-­map, patients from one of the small
group sessions pointed out that we were focusing our summary of commu-
nity resources on patients’ receipt of support rather than understanding
their active role as supporters of others—and that highlighting that aspect of
their lives was important because they gain strength by contributing to their
communities. This idea resonated with us based on several interviews in
which index patients were also parents, aunts/uncles, spouses, children, and
care partners and had roles in their communities. This was the final part of
our study, and so we did not make any changes to data capture. However,
because it was very informative for our analysis, we highlighted this finding
in our article (Fort et al., 2020).

[SEE FIGURE 8.5 at eResource: Composite Eco-­Map from Original Article (Fort et
al., 2020)].

Copyright © 2020 Meharry Medical College. This article first appeared in

the Journal of Health Care for the Poor and Underserved, Volume 31, Number
2, May 2020. Published with permission by Johns Hopkins University Press.

Although we gained important insights from the small group sessions, not
everyone we had hoped would attend was able to do so. One patient and
family member/care partner team had a dialysis appointment and were not
able to attend. The mobility of participants was a consideration; some peo-
ple were in wheelchairs and/or lived far from the clinic. For many studies,
and especially those in low-­resource settings, it is important for researchers
to remember that they are likely to hear from the people who are not the
sickest or most challenged. It is important to closely code and review tran-
scripts, but also remember the voices of those who are not in the room and
consider why they are not there.
136 Meredith P. Fort et al.

Full Team Meeting after Data Collection and Small Group Sessions
After we had completed all the interviews, small group sessions, and a pre-
liminary analysis of the interview transcripts, genograms, eco-maps and
timelines, we held an in-­person analysis meeting with the full team. This was
a way to bring the whole project team together, consider higher-­level themes
from the study, define an approach for preparing the manuscript, and decide
what to prioritize. At that meeting, we agreed to prepare an analysis table (see
Table 8.2) to guide the selection of quotes, visuals to include in the manu-
script, and other key learnings to emphasize in the manuscript. The table
included assumptions about families and diabetes/hypertension manage-
ment, what we learned during the project, the method that allowed us to learn
what we did, examples of families or experiences discussed during the inter-
views or small group sessions, and sample quotes. For example, going into the
study, we expected that the examples of support for diabetes and hyperten-
sion self-­management described by patients would primarily be of family
members assisting them. Although patients did share experiences of support
from family members, they also referred to broader support networks, includ-
ing neighbors, roommates, providers, and other people who played an impor-
tant role in providing them with diverse types of support. This exercise of
documenting our assumptions, which is an important part of reflexivity,
allowed us to consider the key learnings from the study (we started to become
aware of certain assumptions as our study progressed, and this meeting
spurred us on to put them into a table, making them more visible). Aligning
specific quotes with each of the key points that we learned from the study
also was a way to make sure that what we perceived from the interviews was
indeed captured in the transcripts. This table also was important for making
sure that we presented quotes and experiences from many participants rather
than relying on just a few of the interviews to inform our project.

[SEE TABLE 8.2 at eResource: Analysis Table for Manuscript].

Transcript Coding and Analysis Meetings

When the timing was right to really delve into the analysis and focus on pre-
paring the manuscript, Meredith completed coding all the transcripts within
[Link]. Meredith and Corrie then met on a weekly basis to discuss coded
segments of transcripts, come to consensus on the application of codes, and
consider new concepts. Corrie did not participate in the data collection
phase but did join in the coding and analysis phase. We met about 8–10
times to go over the coded transcripts—usually by reviewing PDFs of coded
transcripts, which is an output option within [Link] (Figure 8.6). We also
viewed the hand-­drawn eco-­maps and genograms to help complement our
 Visualizing Relationships in Diabetes and Hypertension Management 137

review of the transcripts and in some cases to remember or deepen our

understandings of a specific patient or family member.

[SEE FIGURE 8.6 at eResource: Excerpts from Two Transcripts with Coded
Segments in [Link]].

One of the analytic methods that we used during our discussion sessions was
the constant comparison method (Miles et al., 2019), in which we compared
index patients and the index patient–family member combinations for the 23
participants for whom a family member was interviewed. We also made que-
ries within [Link] for codes to review quotes and understand how people
were referring to certain ideas. An excerpt of an [Link] report for the
code “challenge for self-­management” showing resulting quotes is shown in
Document 8.2. We also reflected on the broader literature relevant to our
topic and took notes in notebooks and Word documents during our meet-
ings to follow up and share relevant references to help us interpret and
understand our data.

[SEE DOCUMENT 8.2 at eResource: List of Quotes from an [Link] Code Query].

Defining Key Messages

During the study, we identified some key themes right away. The most evi-
dent theme—identified during the initial recruitment efforts—was that fam-
ily support is not always available. Many of the index patients who agreed to
participate in the study did not have a family member or support partner
who was available to be interviewed. In our final sample, only 13 of the 23
people interviewed in the study had a family member or support partner
who was also willing to be interviewed. We had envisioned that each index
patient in the study would be able to identify somebody, but it was clear early
on that it would not be possible, and that the lack of family support was an
important finding.
During the review of coded transcripts, in viewing the genograms and
eco-­maps, reflecting on who index patients identified as their primary source
of support, and in considering and observing the kinds of support that
patients referred to, we found that opportunities for family engagement typi-
cally had a component of multi-­generational involvement. Undiagnosed
family members who could benefit from supporting their family members
tended to be younger, as the younger generation is a source of motivation in
addition to providing support, and elders are knowledgeable and can share
lessons about how they have managed and prevented diabetes and high
blood pressure.
138 Meredith P. Fort et al.

Reflections and Recommendations

Our project highlighted the advantage of going beyond traditional qualita-
tive research tools such as interviews and coded transcripts to also engage
with participants through interactive group sessions and with visualizations.
Visual diagrams allowed participants to see their family and community
context developing as a picture; as they watched their picture mapped out on
paper by Maria, it presented an opportunity for them to add to and correct
data gathering during the process. This interactive use is described in more
detail by Rempel et al. (2007).
Eco-­maps are an easy-­to-­understand tool that is underutilized in both
research and clinical practice. In the same way that eco-­maps may enhance a
researcher’s understanding of family member experiences and interactions
with community resources, so too can a patient and their support circle incor-
porate that knowledge into their disease management efforts. These efforts
could then be utilized to track patient progress. Family members and care part-
ners could prepare and review eco-­maps iteratively, on an annual basis or every
six months, so that patients can be reminded of the resources that they had/
have available to them. Along with a short set of guided questions, eco-­maps
may be an adaptable tool to improve the quality of care and extend care man-
agement outside of the clinic. A related idea that could be of interest to explore
in the future is whether interviews with family members and care partners may
have allowed them to recommit or enhance their role in providing support.
As well as using these tools for tracking progress, they could be used for
disease prevention. Many participants were accompanied or transported to
our study site by nieces/nephews, grandchildren, or children. An option
would be to have providers proactively involve the younger generation as a
source of motivation and as a means of facilitating disease prevention. The
younger generation could benefit from mapping out resources in their family
and community to begin to access them before they get to the stage of a diag-
nosis with diabetes or high blood pressure. With a guided set of questions,
youth could interview elders to capture their guidance in visualizations.
Some community members were looking to be actively engaged in improv-
ing care and contributing to their community and would like to be involved in
research and practice improvement projects. There is an opportunity to involve
them more in program design and assessment and in qualitative studies such as
our project on family engagement. In just a few short hours, we found the small
group sessions to be very informative. In the future, we would recommend
engaging with small groups multiple times to capture their insights more fully.
The community receiving care at FNCH experiences a high level of need,
and the interviews brought up many challenging topics. Maria shared that
sometimes people’s situations came out after the key topics of the interview
had been covered. This is typical in qualitative studies and interviews; often,
 Visualizing Relationships in Diabetes and Hypertension Management 139

the most pertinent information comes out after the interviewer has asked all
the questions in the interview guide and once the audio recorder is turned off.
In one case, after Maria had completed her interview, a patient participant
shared that she was about to become homeless. This is not uncommon for the
patient population served by FNCH, and we had already identified homeless-
ness in other interviews. A life change of that kind is of course very important
for diabetes and hypertension management, and although the interviews
focused on understanding people’s situations as they related to diabetes and
hypertension management, we knew that it was important to make sure that
patients were referred to support services. We also recognized that it was
important to create space between interviews to consider the weight of the
topics discussed during the interviews.

Conclusion
Our experience exploring opportunities for family engagement in hyperten-
sion and diabetes management has highlighted the benefit of using
­visualizations, starting to reflect on interviews early, and working with a mul-
tidisciplinary team. Increasingly, we look to engage community members
more actively in the analysis process. We also aim to provide them with tools
to analyze and interpret their context and the family and community resources
that they may access in managing and preventing chronic diseases.

KEY WORKS GUIDING OUR DATA ANALYSIS

Genograms and Eco-maps:

McGoldrick, M., Gerson, R., & Petry, S. S. (2008). Genograms: Assessment and interven-
tion. WW Norton & Company.
Rempel, G. R., Neufeld, A., & Kushner, K. E. (2007). Interactive use of genograms and
ecomaps in family caregiving research. Journal of Family Nursing, 13(4), 403–419.
Hartman, A. (1995). Diagrammatic assessment of family relationships. Families in
Society, 76(2), 111–122.
These references explain how to prepare genograms and eco-­maps in practice
and research. The McGoldrick et al. text is an in-­depth overview of genograms. The
Rempel et al. article demonstrates how genograms and eco-­maps may be used
together in nursing research. The Hartman article explains how to prepare an eco-­
map. Drawing on these three references, we prepared a guide to use during the
interviews, and we returned to them when we reviewed the genograms and eco-
maps, individually and collectively.

Participatory Approaches:
Wallerstein, N., Duran, B., Oetzel, J., & Minkler, M. (Eds.). (2019). Community-­based
participatory research for health: Advancing social and health equity.
140 Meredith P. Fort et al.

This book builds on decades of practice and research approaches that promote
a decolonizing perspective and encourage the involvement of individuals and com-
munities that are most affected by a topic of interest. In our work, we aim to con-
tinually increase the extent to which patients and community members participate
in all phases of project development, analysis, and dissemination. The small group
sessions allowed us to learn from patients and family members, putting a check on
our researcher perspectives.

Validity and Coding:

Cho, J., & Trent, A. (2006). Validity in qualitative research revisited. Qualitative Research,
6(3), 319–340.
Miles, M. B., Huberman, A. M., and Saldaña, J. (2019). Qualitative data analysis: A meth-
ods sourcebook (4th ed.). Sage.
The Cho and Trent article was helpful in the design phase as we made the deci-
sion to include small group sessions; it also helped us think through what to focus
on in the small group sessions. We reviewed the text by Miles and colleagues and
other related texts, lectures, and workshop notes from qualitative analysis courses
to provide guidance with codebook preparation and during the coding process that
we carried out in [Link]. We also drew on and referred to articles and texts on
family systems theory and research with Indigenous and Latinx communities during
the analysis and interpretation phase.

References
Cho, J., & Trent, A. (2006). Validity in qualitative research revisited. Qualitative
Research, 6(3), 319–340. [Link]
Fort, M. P., Castro, M., Peña, L., Hernández, S. H. L., Camacho, G. A., Ramírez-­
Zea, M., & Martínez, H. (2015). Opportunities for involving men and families in
chronic disease management: A qualitative study from Chiapas, Mexico. BMC
Public Health, 15(1), Article 1019. [Link]
Fort, M. P., Steiner, J. F., Santos, C., Moore, K. R., Villaverde, M. D. L. A., Nease,
D. E., Ortega, D., & Manson, S. M. (2020). Opportunities, challenges, and strate-
gies for engaging family in diabetes and hypertension management: A qualitative
study. Journal of Health Care for the Poor and Underserved, 31(2), 827–844.
[Link]
Hartman, A. (1995). Diagrammatic assessment of family relationships. Families in
Society: The Journal of Contemporary Social Services, 76(2), 111–122. [Link]
org/10.1177/104438949507600207
McDaniel, S. H., Campbell, T. L., Hepworth, J., & Lorenz, A. (Eds.). (2005). Family-­
oriented primary care (2nd ed.). Springer.
McGoldrick, M., Gerson, R., & Petry, S. (2008) Genograms: assessment and interven-
tion (3rd ed.). WW Norton.
Campbell, T. L. (2005). Improving physical health through family intervention. In P.
C. McKenry & S. J. Price (Eds.), Families & change: Coping with stressful events
and transitions (3rd ed., pp. 103–127). Sage.
 Visualizing Relationships in Diabetes and Hypertension Management 141

Miles, M. B., Huberman, A. M., & Saldaña, J. (2019). Qualitative data analysis: A
methods sourcebook (4th ed.). Sage.
Moore, K. R., Schroeder, E. B., Goodrich, G. K., Manson, S. M., Malone, A. S.,
Pieper, L. E., Son-­Stone, L., Johnson, D., & Steiner, J. F. (2023). Racial and ethnic
equity in care for hypertension and diabetes in an Urban Indian Health
Organization. Journal of Racial and Ethnic Health Disparities, 10(3), 1319–1328.
[Link]
Rempel, G. R., Neufeld, A., & Kushner, K. E. (2007). Interactive use of genograms
and ecomaps in family caregiving research. Journal of Family Nursing, 13(4), 403–
419. [Link]
Rosland, A.-M. (2009). Sharing the care: The role of family in chronic illness.
California Healthcare Foundation. Retrieved December 30, 2021, from: http://
[Link]/publications/2009/08/sharing-­t he-­c are-­t he-­role-­o f-­f amily-­i n-
­chronic-­illness
Wallerstein, N., Durran, B., Oetzel, J., & Minkler, M. (Eds.). (2019). Community-­
based participatory research for health: Advancing social and health equity (3rd
ed.). Jossey-­Bass.
9
VOICES FROM INSIDE PRISON
Centering People through Intentional
Sampling, Coding, and Analysis within Large
Research Teams

Danielle S. Rudes, Shannon Magnuson, and

Sydney N. Ingel

Qualitative research is a challenging endeavor for researchers who work with

vulnerable populations (e.g., those living and working in carceral institutions)
where both researchers and individuals who participate in the research experi-
ence provide emotional labor during the research process. Although it is
important to recognize the emotional labor of researchers in the process, it is
also extremely important to acknowledge the more nuanced and unnoticed
ways research may harm participants. Preventing these harms requires
researchers to center participants. Researchers can do this by being intentional
about data collection procedures and practices and paying rigorous attention
to sampling, coding, and analysis techniques. This work increases in complex-
ity when large research teams work collaboratively with each other and with
the individuals and organizations centered in their study. Using our paper on
incarcerated individuals’ rights in Restricted Housing Units (RHUs) (Rudes
et al., 2021) as an example, this chapter examines the complex navigation
required for large teams to code and analyze data collected from individuals
living and working in US prisons. The chapter includes a focus on centering
using rigorous, intentional, and in-­depth methodological techniques such as
secondary coding and analysis to gain a richer understanding of participants’
narratives and experiences. We also provide access to our teams’ decision-­
making around processing data and developing interpretive findings.

Background and Context for Our Two RHU Studies

Our team of researchers conducts qualitative fieldwork and interview data
collection in carceral settings, including within RHUs in US prisons. A

DOI: 10.4324/9781003254041-10
 Voices from Inside Prison 143

RHU is essentially a “prison within a prison” (Browne et al., 2011, p. 47).

In these often hidden carceral spaces, residents1 live and staff work in iso-
lating and, at times, desolate conditions. In the prisons we studied, up to
400 prison residents may live in a large RHU complex that holds hundreds
of cells where residents most often share a six-­ by-­nine-­
foot cell with
another resident for approximately 23 hours each day. Some residents stay
in the RHU for around 30 days, while others are there for many years.
Cellies (individuals who co-­reside in a cell) share a metal bunk bed, a toilet
and sink combo, and sometimes a writing desk (all of which are secured to
concrete walls). Depending on their custody status (security/risk level),
RHU residents may have little or no personal property except legal materi-
als. Individuals in RHUs for rule-­breaking behavior are there on what is
called Disciplinary Custody (DC) status. These residents have no televi-
sions, radios, or electronic tablets and receive no visits from family/friends
during their RHU stay. Alternatively, residents confined to the RHU for
non-­disciplinary reasons (e.g., personal protection) called Administrative
Custody (AC) status receive some privileges, including keeping personal
property like photos and televisions/radios in their cell, and occasional
visits from family and friends. Residents confined in the RHUs in our
study did not have access to programming (treatment services or classes)
but did have access to telephones for calls with attorneys. Residents only
leave their cell when staff escort them to showers (10 minutes, 5 days a
week) and/or yard/recreation time (1 hour, 5 days a week) and must agree
to a strip search each time they leave their cell. Staff deliver all meals
through a wicket (a small slot in the metal sliding door) in each cell. A
small team of correctional staff supervise the unit, including two middle-­
level supervisors—a sergeant and lieutenant.
A typical day for RHU correction staff consists of many routinized tasks.
These include making rounds to check on residents through the narrow cell
door windows, collecting names of residents wanting showers and yard
access, and handcuffing and transporting residents to shower or yard. Daily
tasks also include inventorying and distributing property to AC residents,
strip searching incoming and outgoing residents, delivering meals to resi-
dents, changing resident cell assignments to conform with institutional and
legal policies, counting residents, and dealing with unexpected situations
with residents such as physical and mental health needs.

Our Studies
Relying on an existing researcher–practitioner relationship with one US
state’s Department of Corrections (DOC), our research team embarked on
two qualitative studies of RHUs across seven different prisons over a
144 Danielle S. Rudes et al.

four-­year period: Together Alone and Changing the (w)Hole Mind. The work
involved a team of 26 individuals, including 14 undergraduate students, nine
graduate students (both masters and doctoral), and three university faculty.
The 14 undergraduate students primarily served as interviewers, whereas the
graduate students and university faculty served as interviewers, coders, and/or
writers. The majority of this 26-­person team was comprised of white women,
and we also had two men, three Black individuals, two LGBTQIA+ (lesbian,
gay, bisexual, transgender, queer, intersex, asexual, and more) individuals, and
3 individuals with previous criminal legal involvement on our team.
The Together Alone Project included ethnographic observation of staff
and interviews with residents and staff in three sex-­segregated men’s and one
sex-­segregated women’s prison2 RHU units. The Changing the (w)Hole
Mind Project included ethnographic observations of staff and interviews
with residents and staff in four men’s prison RHU units. These study sites
represented two maximum and five medium-­level security facilities in geo-
graphically distinct areas of one US state. Researchers conducted site visits
for two consecutive days per prison trip in the first study and four consecu-
tive days per prison trip in the second study.
The research team used a semi-­grounded theory framework for data col-
lection. We use the term “semi-­grounded theory” as a framework for our
analysis style because our semi-­structured interview style does not fit the
boundaries of traditional grounded theory (Glaser & Strauss, 1967), as our
method combines both deductive (preconceived) and inductive data collec-
tion and coding strategies, not just the latter. Charmaz (2006) notes,

Grounded theory methods consist of systematic, yet flexible guidelines

for analyzing qualitative data to construct the theories from the data
themselves. Grounded theory begins with inductive data, invokes iterative
strategies of going back and forth between data and analysis, uses com-
parative methods, and keeps you interacting and involved with your data
and emerging analysis.
(p. 1)

Combined, the RHU project data collection included 1,666 data collection
hours. The data collection produced 351 resident interviews (representing
roughly 25% of all residents confined in these prisons’ RHUs), 95 staff inter-
views (representing roughly 62% of all staff working in these prisons’
RHUs), and 67 sets of ethnographic observational fieldnotes. We also col-
lected various federal, state, and DOC policy documents from within the
institutions and from publicly available sources. Examining the life of both
individuals confined and individuals working in these desolate conditions
and largely invisible environments was the goal of our study (Rudes et al.,
 Voices from Inside Prison 145

2021). We hoped our research would lead to greater knowledge about these
spaces and conditions and possible policy and practice reforms.
The research team coordinated with RHU staff to escort residents who
volunteered for an interview to designated locations to conduct informed
consent and a 45-­to 60-­minute conversational interview focusing on key
interview focal areas. Staff were recruited and sometimes interviewed during
their daily work within the RHUs, and often this occurred via conversations
in the hallways, offices, and control room. When staff agreed to an interview
and/or observation, researchers commonly accompanied them in the course
of their duties, asking questions as time permitted and engaging in conver-
sational, unstructured interviews. Additional information about data collec-
tion is found in Rudes et al. (2021).
During the interviews, researchers took brief notes—to remain engaged
with the participants—and made mental notes, or what Emerson et al.
(1995) called head notes. We used available downtime (in prison, this is often
while awaiting an interview or during resident count times) to make brief
jottings and elaborate on our interview notes while they were fresh in our
minds (see Emerson et al., 1995). Our researchers did not use recording
devices because they were prohibited in the prisons of study. Instead, mem-
bers of the research team received training to take handwritten notes during
interviews and type full renderings of interview and observational notes
with residents/staff after exiting the field each day. Our training varied with
each group but regularly included individuals receiving initial training on
notetaking followed by the group listening to an interview via a YouTube
video while taking notes. Then, the group worked with their trainer to com-
pare and contrast what each trainee captured. Finally, using discussed skills
such as writing direct quotes or capturing vocal inflection or tone, the train-
ees watched additional videos and perfected their skills.
After each fieldwork trip, researchers uploaded interview and observa-
tional notes (and any other gathered documents such as policies or manuals)
to the shared team folder located in a secured cloud server—we used Dropbox.
Upon completion of data collection, the team then created an [Link] file
and imported the interview transcripts into it. [Link] (version 9) is a qual-
itative software program for coding and analysis. We also kept some docu-
ments relevant to the study only within Dropbox, such as case law, prison
policies and manuals, and related literature. All of these documents could
have been assigned to an [Link] file and coded there, but for this project,
we hand-­coded these documents as needed and did not code them within
[Link]. Hand-­coding is a personal preference we often employ, as there is
something about printing the policies and combing through them by hand
that our team finds helpful for deeper understanding. To do this, we print
each policy and code by hand, using various colored highlighters. We then
146 Danielle S. Rudes et al.

discuss this coding and make changes as needed. Although technically cod-
ing is one of the early steps in analysis, our team primarily considers these
two separate functions. In our work, we view coding as a simple process that
includes labeling or categorizing data into groups. Analysis, however, is the
analytical steps involved with interpreting the coded data and making sense
out of seemingly disparate categories to form a narrative or argument.

Major Concerns among Large Teams for Centering Voices

Of key concern to our research team was centering participants’ perspectives
and ensuring confidentiality. The idea of centering the voices and experi-
ences of participants—without diminishing their confidentiality—privileges
their perspectives, affords analytic interpretation and insight, and does not
reduce their power or voice through translation and dissemination. This was
a crucial part of our team’s work and one we took seriously. To do this, we
kept the concept of centering at the pinnacle of our focus during all project
phases, from study design to data collection, analysis-­level sampling and
sub-­sampling, coding, analysis, and dissemination via writing and presenta-
tion. In the following subsections, we discuss the role centering played in
these various stages of research.

Sample or Subsample
To include multiple perspectives we interviewed both staff and residents and
observed staff in all available prisons’ RHUs. The data yield from this
approach was tremendous, but to blend all the data together for every paper
does not center particular groups of residents or staff in ways that fairly
highlight particular experiences and perspectives. To accomplish centering,
we intentionally examined the data as a whole and subsequently made team
decisions about the whole sample or a particular subsample of data to decide
which grouping(s) would be included in each analysis. For example, after
examining all the resident interviews, perceptions of rights and access to
these rights solidified in our semi-­grounded theory analysis.
However, from our time in the RHU, we understood that rights, by pol-
icy, differ between AC and DC status residents. Going beyond themes about
rights and digging into the perceptions about access to these rights required
our team to center a unique perspective. Given that DC residents have the
fewest rights per policy and the hardest time (in theory) accessing them, our
team decided to focus our subsample on DC residents.
We also made a team sampling decision to exclude the data from women
RHU residents because penal literature commonly noted (e.g., Trammell,
2012), and as our fieldwork suggested, both the culture and context of women’s
 Voices from Inside Prison 147

prisons are vastly different from the culture and context of men’s prisons.
Although the experiences of women in RHUs are critically important, we had
more data from the men, and DC status men seemed to have a more difficult
time accessing their rights. Centering the men on DC status in the RHU was
critical to elevating their story and sharing their unique experiences.
During analysis, we also determined that elevating a more exhaustive
story about how DC residents confined in the RHU access their rights
required centering another voice—the voice of RHU staff. While residents
noted their lack of rights (i.e., some residents believed they had no rights
within RHUs) and their distinct difficulty accessing rights (if they thought
they had some), staff also shared both perspectives.

Data Organization
With large team projects, record keeping and organization are essential. As
previously noted, our team used a shared Dropbox folder for all project files,
including data files such as fieldnotes, interview transcripts, and organiza-
tional documents (for triangulation). We also maintained separate Excel
files within Dropbox where we logged all project data collection activities.
Figures 9.1 and 9.2 provide a glimpse into our teams’ file organization in
[Link] and Dropbox.

[SEE FIGURE 9.1 at eResource: [Link] File Organization].

[SEE FIGURE 9.2 at eResource: Dropbox File Organization].

In our “Hours and Activities” spreadsheet, we documented every hour of

fieldwork, including what activities we performed, observed, or participated
in and which individuals and how many we interacted with. This provided
an easy total calculation of hours spent collecting data for paper-­writing
later. It also afforded instant access to current progress on a project that
helped when funders asked for data updates, and it helped our team assess
where to focus attention during subsequent field visits. Additionally, for
some projects, we kept a “People and Places” spreadsheet where we logged
demographic information about project participants and/or key information
about places. To bolster RHU resident and staff confidentiality, we did not
include any real names in this file but rather identified individuals by a
unique identifying number or pseudonym. Giving individuals names rather
than numbers (CO1 or Resident 2) is an important part of centering people
(Thomas & Hirsch, 2016). It humanizes their experiences, particularly for
readers who are socially distant from the conditions and impacts of prison.
This file provided easy access to demographic information, but perhaps
148 Danielle S. Rudes et al.

more importantly, it assisted with our goal of centering people by helping us

focus on which stories we were elevating and which stories we had not care-
fully considered and/or included.

Coding Strategies
When coding a large dataset collected by a large team, there are many pos-
sible options to accomplish this with rigor and integrity. Before thinking
about papers or presentations, we coded the entire dataset first using line-­by-­
line coding. This process goes “deeper into the studied phenomenon and
attempts to explicate it” (Charmaz, 2006, p. 124). This coding process took
considerable time but was worth the investment in later coding iterations.
Figure 9.3 provides an image of one of our coded documents in [Link].
This picture shows not only the [Link] interface but also the work we do
to rigorously code each interview file.

[SEE FIGURE 9.3 at eResource: [Link] Coded Document].

In our RHU study, we had two different projects with different interview
foci in both men’s and women’s carceral facilities over nearly four years. To
code these diverse data and meet our need for consistency and inter-­coder
reliability (ICR), we divided the full dataset into three distinct projects for
coding. “ICR is a numerical measure of the agreement between different
coders regarding how the same data should be coded” (O’Connor & Joffe,
2020, p. 2). The first dataset included the data from both studies within
women’s prison(s) only, the second dataset included the data from men’s
prisons during the first study, and the third dataset included data from the
second study of men’s prisons. We divided coders into teams of two gradu-
ate students, with each team coding one sub-­dataset.
To begin the coding process, the coders worked together to collaboratively
code one interview transcript within [Link]. They opened a transcript
within [Link] on one laptop and sat next to each other to verbally discuss
how to code the data by reading the transcript together and making collab-
orative choices regarding how, when, and why to code each data segment. For
line-­by-­line coding (Charmaz, 2006), some coders may prefer literally coding
per line. Our team found it more productive to read each line of data but
assign codes per section (by hard returns within typed notes). The coders col-
laboratively added codes to the code list and continued coding the entire
document. Depending on the size of the dataset, the coders, and the expected
variability between interviews, fieldnotes, and other documents, some coders
coded an additional sample of documents (maybe 5%). Using ICR standards
(MacPhail et al., 2016), the coders, after reaching at least 90% coding consen-
sus, then divided up the remaining project documents for coding.
 Voices from Inside Prison 149

Coders remained in continual conversation with each other throughout

this process, discussing new codes and anything that was different from the
initial documents they coded. They used the comment feature in [Link] to
define all initial and new codes. These coding notes might include slight defi-
nitional changes or accommodations when coding data that mostly fits within
a current code (as defined) but adds some potentially new facets. An example
of this might include a code for Right: Shower (meaning a person discussed
their right to take a shower). Then, as one coder is reading data, they notice
that an interviewee mentioned their belief that they have a right to hygiene and
discussed their need to not smell like sweat, but the word “shower” is not men-
tioned. In this case, a coder might not code this data segment as “Right:
Shower” or they might choose to code it that way even though specific men-
tion of a shower is not made. To alert other coders of this definitional change,
they add to the existing definition of Right: Shower (using [Link]’s com-
ment feature), which presently noted this code applies to any time an individ-
ual mentions their perceived right to take a shower. The addition might note it
is not only about showers but also about personal cleanliness and hygiene.
Then, our team practice was for the two coders to email each other, talk to
each other, or sometimes share a Google document with updated code defini-
tions, immediately after coding so the next coder recognized the change,
included the new definition in all future coding, and remembered to go back
to previous coding to make sure they coded all data correctly for this code.
Using the Google document in addition to the code comment features in
[Link] was necessary because [Link] does not have a mechanism to
alert you when something has been updated or changed. The coders met peri-
odically throughout the process as a form of ICR checking to ensure no drift
had occurred from the original coding standards. We used this process whether
we were employing a constant comparative analysis (Glaser, 1965) or waiting
to code data after data collection was complete. Figure 9.4 provides an image
of our teams’ use of the code comment feature in [Link]. It depicts the
data (on the right side of the screen), the codes on the left side of the screen,
and the comments for each code in the bottom left corner of the screen.

[SEE FIGURE 9.4 at eResource: [Link] Code Comment for Right: Shower].

We would like to point out that although [Link] has a feature to auto
calculate ICR within a project, our team preferred a more organic approach
(as outlined above) because we felt the [Link] calculations were too
strictly formulated for our interview data. For example, one coder may code
the same bit of text as “mad” and another as “frustrated.” In some analysis,
the distinction between mad and frustrated may matter greatly, but for our
purposes, it typically did not. We would likely consider those as coded reli-
ably. We could have deleted one of those two codes, boxing coders into one
150 Danielle S. Rudes et al.

or the other labels, but again, we preferred more, rather than fewer, options
for coders so their independent views of the data could shine. In future itera-
tions of coding and/or analysis, we may similarly merge codes, but we also
may not. Thus, hand-­calculating ICR is the best choice for our team.
As noted earlier, coders worked in pairs to code the data to provide both
coding consistency and reliability. To determine ICR, we agreed on a few
decision points, such as data size (e.g., hard returns) and a code list. Then,
ICR was a simple calculation of the number of times two coders coded the
same chunk of data with the same (or similar) codes. Inter-­coder reliability
was achieved above 90% within all three coding teams. Although this pro-
cess may seem far askew from centering, it is absolutely not. Coding sub-­
datasets improved ICR, with fewer coders coding each dataset. It also
afforded the mental space and intense focus for a smaller number of coders
to focus on a smaller number of documents, and it improved understand-
ing of the collected data for each data subset. Moreover, it allowed teams
to center subsamples of people to tell clear and unique stories and elevate
specific experiences. It centered the realness by not holding people in the
aggregate, which occurs when we group people simply by their status as
“prison residents,” where we lose the nuance of individuals’ unique
stories.
Taking an intersectionality approach (Crenshaw, 1989) to the prison set-
ting meant considering gender identity by prison housing residency (RHU)
by status (DC/staff) by perceptions/experiences by other unique identities or
experiences. The subsample process, by design, drilled down to understand
the complexity and differential experiences of penal life as experienced by
complex and unique people. [Link] has a code view option called
“groups” that allows researchers to only see certain codes while hiding oth-
ers. We used this option sometimes when the code list became excessively
long. Our team also used the query function with [Link] to pull data
from only one or more subsamples. For example, we queried our data to see
if there were any racial/ethnic differences regarding which residents had dif-
ficulty accessing yard, shower, and meals. To do this, we searched our
[Link] file using the query function to look for data we coded both as
“Black/African American” and “Burned for Shower,” and completed subse-
quent queries for other racial/ethnic categories and rights. We then com-
pared the results. This allowed researchers to use operators like “and,” “or,”
and “within” to search coded data. Both of these methods improved our
ability to center the voices and experiences of our participants by paying
more, and more careful, attention to each sub-­group.
With two coders coding each dataset, they divided the interview tran-
scripts into equal numbers and concurrently coded them using the pre-­
established coding criteria. Although [Link] has a “merge” function so
 Voices from Inside Prison 151

two coders can code at the exact same time on different computers, until the
most recent versions of the software, this process was clunky and somewhat
unreliable. Thus, our team established one project file in our Dropbox folder,
and everyone coded off the same file. The coders arranged coding times with
each other so that everyone knew who was coding when. At the end of each
coding session, the coder saved their [Link] project file and uploaded it
to the shared Dropbox folder with the date of upload in the naming conven-
tion. When the next coder was ready to start, they downloaded the most
recent [Link] project file and began coding. With this process, all coding
is done within one document, and coders can see each other’s work.

Analysis Strategies
Once coding was complete and before we began analysis, we met as a team
to discuss the coders’ initial thoughts. In this brainstorming session, each
team of coders came to the meeting with initial thoughts about research
questions that solidified from the data during coding. For example, we
asked interview questions about perceptions of rights in our data collec-
tion, but we did not specifically ask about how staff might transform rights
into privileges within the RHU. We discovered this facet of perceptions of
rights during coding, and we discussed it at our brainstorming session. This
insight led to the paper we reference in this chapter (Rudes et al., 2021).
Although these brainstorming sessions occurred preanalysis, we were not
sure yet if these questions were answered by the data in ways that were use-
ful and publishable. However, the coding teams came to the meeting pre-
pared to discuss their ideas, get feedback from the groups, and see if there
were ways that the dataset they coded could speak to the data in other coded
datasets for the project(s).
At brainstorming sessions, we discussed ideas for research questions
that we believed might lead to a better understanding of living and working
in RHUs and ultimately to papers and presentations. Each team presented
their ideas, and coders from other teams expressed interest in ideas they felt
connected to or to which they could contribute. At the end of the day, our
team had an assemblage of 17 analysis ideas from our three datasets that
we tracked in a Word document. Some of these ideas combined datasets,
some took a subsample from a particular dataset, and some just used a
singular sub-­dataset. We created a list of all the potential papers, added the
names of possible authors to each paper (according to their interest), and
then assigned a lead to each paper idea. If an individual had more than one
lead, we also prioritized the papers, privileging the ones with the most
strength and potential contribution. Team members then spent months on
their paper teams conducting grounded theory-­based qualitative analysis
152 Danielle S. Rudes et al.

of their particular dataset. This involved recoding the data with an eye
toward focused or axial coding (also called secondary coding) (Tracy,
2019). In secondary coding, “the researcher critically examines the codes
already identified in primary cycles and begins to organize, synthesize, and
categorize them into interpretive concepts” (Tracy, 2019, p. 225). The teams
frequently met to discuss ideas for theoretically framing the papers, how
the concepts in the data could be explained, and how they hung together to
form an argument that would contribute to the literature and/or to policy/
practice reform.
To describe our analysis process, our team often uses an onion metaphor.
An onion has deep layers that a person can peel back to reveal the inner core.
This method fits with two imperative questions all qualitative researchers
should ask of their data: (a) Under what conditions did something occur?
and (b) What are the mechanisms by which that something occurred? The
onion imagery evokes a slow and intentional pulling to identify each deeper
layer and fits well with a “beyond themes emerged” analysis strategy as the
surface level (primary coding) does not look at all like the layers on the
inside (secondary and deep coding) once identified.
In our first layer, we conducted line-­by-­line or primary (Tracy, 2019) cod-
ing to identify major themes and patterns, but that was not nearly enough to
understand and then justly center the voices and experiences of the individu-
als in our studies. The second layer of analysis, post brainstorming session,
represented our secondary coding cycle (Tracy, 2019). This layer was more
nuanced and analytical as coders organized key findings into interpretive
concepts. Finally, as qualitative researchers know, there is never an “end” to
coding, so we worked through what we call axial coding (some also refer
to this as hierarchical or deep coding, see Tracy, 2019). Here, the story began
to take shape when we used [Link]’s query function to answer the condi-
tions and mechanisms questions from above. In essence, this was the sense-
making part of the onion peeling. The core or center of the onion, for us, is
the narrative that participants told, but with our interpretive and analytic
expertise added in a way that centered the participants while providing sen-
semaking for reader(s). Only then, once we had identified what we thought
was the core of the analysis, and we could verbally explain the argument to
our team, did we begin to write.
We return to the earlier conversation about the paper focused on inacces-
sibility of rights (Rudes et al., 2021). During the second round of coding, the
coders assigned new codes identifying the specific rights (e.g., meals, show-
ers, yard time, and medical care) mentioned in each example, whether the
resident/staff member believed this right was guaranteed, and the circum-
stances or mechanisms that would lead to a right being denied. Together,
staff and DC residents identified 17 rights afforded to individuals confined
 Voices from Inside Prison 153

in the RHU, and they discussed in their interviews several events or circum-
stances where residents could not access these rights. Interestingly, in these
stories, it appeared as though some rights were more inaccessible than oth-
ers. To confirm this concentration and understand more about which rights
were the most inaccessible, we used Pareto’s principle, or the 80:20 rule or
law of vital few. This principle asserts that roughly 80% of the effect comes
from 20% of the causes (Clarke & Eck, 2007; Kock, 1999). In the RHU con-
text, this means a small number of rights account for a large percentage of
instances of rights inaccessibility. Using the query function in [Link], we
identified how many individuals listed something as a right and how many
individuals reported that each of those rights was denied. Queries in ATLAS.
ti are exportable to Microsoft Excel, so we did this. Then we used this data
to create Table 9.1 (in Microsoft Excel), which detailed the 17 rights and resi-
dent perceptions regarding inaccessibility.
Our analysis suggested that across 78% of the events participants described
where a right was inaccessible for some reason and, four specific rights
accounted for these events: (a) yard (25.44%), (b) showers, (23.68%), (c) meals
(21.98%), and (d) medical services (7.02%). This suggests potentially contex-
tual circumstances at play, resulting in missed opportunities for residents’

TABLE 9.1 Concentration of Reported Instances of Inaccessible Rights

Right Reported Instances % of Cumulative % of Cumulative %

Inaccessible Rights (n) Reported of Rights in
Instances RHU

Yard 25.44% (29) 25.44% 5.88%

Showers 23.68% (28) 49.12% 11.76%
Meals 21.93% (25) 71.05% 17.65%
Medical services 7.02% (8) 78.07% 23.53%
Sanitation 6.14% (7) 84.21% 29.41%
Hygiene products 2.63% (3) 86.84% 35.29%
Psychological care 2.63% (3) 89.47% 41.18%
Law library 2.63% (3) 92.11% 47.06%
Phone 1.75% (2) 93.86% 52.94%
Clothes 1.75% (2) 95.61% 58.82%
Heat 1.75% (2) 97.37% 64.71%
Mail 0.88% (1) 98.25% 70.59%
Visits 0.88% (1) 99.12% 76.47%
Winter clothes 0.88% (1) 100.00% 82.35%
Linens 0.00% (0) 100.00% 88.24%
Water 0.00% (0) 100.00% 94.12%
Working toilet 0.00% (0) 100.00% 100.00%
Total 100.00% (115) - -
154 Danielle S. Rudes et al.

access. We return to this point a bit later in this chapter. As part of our analy-
sis, we also had a team meeting where we made the decision to focus on the
first three rights (yard, showers, and meals) because they were the most
prominent.
With a firm understanding about which rights were more important (as
measured by the frequency participants discussed them) and most often inac-
cessible (as measured through the Pareto Principle’s concentration analysis),
we deepened our analysis and identified several findings. First, our analysis
revealed that although staff and residents discussed a litany of rights (n = 17),
not all of them were rights afforded to residents by policy (unit, agency, or case
law). That is, some of the “rights” described were merely norms or privileges.
Second, our team developed conceptualizations based upon narratives about
why the rights were inaccessible, including expressive (personal reasons) and
instrumental (functional reasons) rationales. Combined, these layers of analysis
revealed that staff and resident data together created a deeper analysis around
resident right accessibility that suggested three major findings: (a) neither resi-
dents nor staff knew or largely agreed regarding what rights were afforded to
RHU residents; (b) residents and staff largely agreed that interpersonal dynam-
ics (such as favoritism and racism) often prevented residents from accessing
their rights, and (c) even though residents thought formal, written policy pre-
vented their rights access, staff overwhelming suggested that it was informal
policy that blocked resident access to rights. Additionally, a deeper theoreti-
cally based analysis moved our writing in Microsoft Word documents toward
trying to understand why and how this occurs. We determined that rights in
prison operate at the intersect between the written policy and the policy-­in-­
action, and our paper—“rights-­ in-­
between”—focuses our argument on
expanding existing theory while uncovering important perceptual and behav-
ioral action of both individuals living and staff working in RHUs (Rudes et al.,
2021). Figure 9.5 shows a sample from our queried code report in [Link] for
just one of our codes: “Rights—Perceptions.” Table 9.2 shows how we sum-
marized our interpretive findings of both instrumental and expressive
rationales.

[SEE FIGURE 9.5 at eResource: [Link] Queried Code Report].

We could have stopped at these three key findings. However, we did not fully
comprehend how the staff operated within their organizational and systems
context. Prisons are individual institutions with their own internal policies
and practices, and they are also part of a state DOC and part of the carceral
apparatus in the US. Therefore, to understand the context of rights at the
unit, agency, and national levels, we looked back through our notes and doc-
uments for individual prison or DOC policies. Surprisingly, in a carceral
 Voices from Inside Prison 155

TABLE 9.2 Expressive and Instrumental Rationales

Residents Staff

Expressive Rationales
CO lazy 16% 0%
CO dislikes you 16% 20%
Instrumental Rationales
Resource constraints 2% 6%
Treat as privilege 8% 6%
Discipline individual/unit 6% 53%
Not following policy 50% 22%
Note: “CO = correctional officer”.

space that is bound by seemingly endless rules and layers of bureaucracy,

there were very few formal rules or guidelines in these documents that per-
tained specifically to rights afforded to RHU residents. So, we turned to fed-
eral case law involving rights in the RHU and found almost none. After
another round of coding for policies and practices within our interview and
observational notes, we used [Link]’s report function to create a list of all
coded data for a particular policy (like showers) and compared it against the
data in our Dropbox files on the case law and the DOC and individual pris-
on’s RHU policies.
Although the comparison between interview and observational notes and
policies is fruitful for deeper understanding, the data were complex, and we
needed a way to focus on what was important. One way our team works
through analysis was by hand-­drawing or creating figures in Microsoft Word or
PowerPoint. Sometimes this includes flow charts to depict processes, and other
times drawn diagrams help us understand what we saw in the data. Our first
drawing for the rights paper, represented in Figure 9.6, is an example of one of
our hand-­drawn diagrams. In brainstorming sessions about data we coded as
“rights,” we discussed which residents and staff we might include to make this
argument. To center the people we interviewed and observed, we needed to
understand how individuals got into the restricted housing units in the first
place. Thus, Figure 9.6 displays the pathways that individuals take to get into
restricted housing. We initially thought this content would become part of the
rights paper, but later decided that this analysis needed to be fleshed out more
and deserved its own paper. Currently, a paper describing pathways individuals
take to enter restricted housing units is in progress. Although we did not use
this diagram to formulate the rights paper, it was nonetheless a valuable and
productive activity. It focused our rights argument by helping us determine
who did not belong in the rights paper (DC status residents and staff), and it
led us to writing a second paper.
156 Danielle S. Rudes et al.

[SEE FIGURE 9.6 at eResource: Model of Deep Analytical Processing for Qualitative
Data].

Next, we created a diagram of what we were learning from the comparison

of case law, DOC’ policy, and individual prison policy regarding rights such
as showers. Figure 9.7 shows one of the three diagrams like this that we cre-
ated to graphically depict residents’ rights to shower. In this example, and as
is typical for our team, we first hand drew flow charts, diagrams, and figures
and then recreated them in PowerPoint for presentations and papers.
Visualizations such as this are one way our team makes sense of complex
processes and/or interactions. A diagram like this allows us to manipulate a
pictorial image of our findings as we process how to structure an argument
in a written paper. In Figure 9.7, the image suggests how the shower policy
that stems from case law vastly differs from both the organization’s and a
particular prison’s policy regarding showers. For our paper on rights, this
meant that an individual’s right to a shower was not guaranteed by carceral
institutions even if existing case law or policy suggests that it was.
This deep analysis step identified yet another onion layer, one that was
excessively helpful in formatting our argument. The finding here was that there
was an absence of policies/rules regarding rights and that the federal law, the
DOC policy, and the individual RHU policies were vague and open to wide
interpretation by correctional staff. According to federal case law, showers shall
“be accorded” to prison residents (not specifically to RHU residents). Then,
the DOC policy notes, showers “shall be permitted.” And the unit’s policies
varied across prisons if a shower was “a privilege,” “an opportunity,” and
“offered.” There is wide and important variance in these terms. Although
“accorded” means “to grant to someone” and permitted means “allowed,”
other words like “privilege” mean something that is earned, but how RHU resi-
dents earn the right to shower is unclear. Additionally, if something is “offered,”
can it also be not given? For example, “I offered it to him, but I didn’t actually
give it to him.” We represented these different perspectives in Figure 9.7. We
also used these findings to form the introduction and discussion sections of our
paper and to contextualize how the voices and experiences we centered must
live within the context of layered systems—unit policies, agency policies, and
case law. Continuing to peel back the onion for this paper allowed us to not
only elevate an unbelievably raw story about people not getting basic rights like
showers, meals, and recreational time but also understand why this might hap-
pen systematically.

[SEE FIGURE 9.7 at eResource: Diagram of Contextualizing Analysis Within Law

and Policy].
 Voices from Inside Prison 157

Conclusion
The point of this chapter, and this entire book, is to ask (or perhaps
demand) qualitative researchers to dig deeper and do more with the data
they collect. That is, to go well beyond basic analysis such as “themes
emerged” to the richer and contextualized core of the narratives and obser-
vations yielded from fieldwork. There are several ways to accomplish this
via team (or even individual) research, including sampling and sub-­sampling
from existing datasets; coding and recoding with thought and intention;
and analysis that identifies layer after layer of a story or narrative to answer
what, why, how, and in what context. This deeply intentional level of analy-
sis provides researchers with the ability to extend or expand existing theory
or develop new theoretical building blocks toward a new way of under-
standing a problem or situation. It also affords researchers with the oppor-
tunity to contribute to scholarly knowledge through conceptual and
analytic contributions via dissemination methods such as publishing. And
it offers qualitative researchers a way to hone their research pursuits for
greater good by way of policy or practice recommendations for change.
What it does not automatically do, however, is center the voices of the par-
ticipants who gracefully and selflessly partner with researchers so that we
may collect data for our studies.
Centering individual perceptions, thoughts, beliefs, narratives, stories,
and actions is our most important job as qualitative researchers. It is no
longer an “insider/outsider” debate (Corbin Dwyer & Buckle, 2018).
Instead, it is about the humanity of the people we study and how we may
set aside our pride as subject matter experts to let their knowledge and
experience shine so that we may elevate and amplify their expertise to an
audience they might not otherwise reach. That is centering. In the field of
design craft, a person-­centered approach is one in which the people come
first in all design elements—thinking about what people need, what they use
a space for, and how they want or need to use a space. In qualitative research,
centering involves treating research participants as a person first and then
supporting their ability to lead the narrative, the actions, and to find their
own pathway. As researchers, we learn the craft of methods and focus
intently on rigor and ethics. However, we do not often discuss or even con-
sider the people we include as participants beyond how they contribute data
or raise our sample size. But they live in the world, bodies, communities,
institutions, groups, and units we are studying. To center them throughout
every step of the process, and especially within sampling, coding, and anal-
ysis, is to truly support them and to authentically create and disseminate
research of consequence.
158 Danielle S. Rudes et al.

KEY WORKS GUIDING OUR DATA ANALYSIS

Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative

analysis. Sage.
Although we acknowledge that we did not use a strict and traditional grounded
theory analysis for our data collection or analysis, we relied heavily on Charmaz’s
adaptation of grounded theory. Charmaz’s attention to ways that assist researchers
in identifying themes is imperative to how we center people in our projects.
Glaser, B. G. (1965). The constant comparative method of qualitative analysis. Social
Problems, 12(4), 436–445.
This work was published a long time ago, but still today, we find it invaluable to
our methods processes. In this work, Glaser painstakingly details the importance of
an open-­mind and an ever-­changing way of looking at and analyzing data. We use
Glaser’s constant comparative method to challenge ourselves to continually work
between methods, findings, and coding/analysis to dig deeper and ask ourselves
intentionally robust questions in a quest for better interpretive answers.
Tracy, S. J. (2019). Qualitative research methods: Collecting evidence, crafting analysis,
communicating impact. John Wiley & Sons.
In this accessible textbook, Tracy walks readers through each part of a qualitative
project with intense focus and thoughtful guidance. It may seem odd to recom-
mend a textbook as a key work, but when working with large teams, the PI on our
projects often worked with students at every level (undergraduate, master’s, and
doctoral) and even other faculty and post-­doctoral researchers. This textbook pro-
vides an important guide for emerging qualitative researchers as they learn and
navigate the qualitative processes described in this chapter.

Notes
1 In keeping with the growing and important movement toward relying on de-­
stigmatizing language, we opt to use the term “residents” to refer to individuals
confined to RHUs. We use this term because it denotes a particular locale rather
than relying on a stigmatizing a label or judgement on an individual the way
words like “inmates,” “prisoners,” and “convicts” convey.
2 We refer to these prisons as sex-­segregated men’s and women’s prisons but
acknowledge transgender and non-­binary residents who live in the institutions
that do not meet their gender identity. The state currently uses a bifurcated sys-
tem to house individuals by sex assigned at birth and refers to these prisons and
“men’s” and “women’s” prisons without regard for gender identity. We rely on
their classification language in our writing.

References
Browne, A., Cambier, A., & Agha, S. (2011). Prisons within prisons: The use of seg-
regation in the United States. Federal Sentencing Reporter, 24(1), 46–49. https://
[Link]/10.1525/fsr.2011.24.1.46
Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualita-
tive analysis. Sage.
 Voices from Inside Prison 159

Clarke, R. V. G., & Eck, J. E. (2007). Understanding risky facilities. US Department of

Justice, Office of Community Oriented Policing Services. Retrieved July 4, 2023,
from [Link]
ved=2ahUKEwj5nJTQtfX_AhVJj4kEHbDOAR8QFnoECA0QAQ&url=https
%3A%2F%[Link]%2Fresourcecenter%2Fric%2FPublications
%2Fcops-­p 118-­p [Link]&usg=AOvVaw2zL7VPiFhZ8QiiAxpNmVkV&opi=
89978449
Corbin Dwyer, S., & Buckle, J. L. (2018). The space between: On being an insider-­
outsider in qualitative research. International Journal of Qualitative Methods, 17,
1–2. [Link]
Crenshaw, K. (1989). Demarginalizing the intersection of race and sex: A Black fem-
inist critique of antidiscrimination doctrine, feminist theory and antiracist poli-
tics. The University of Chicago Legal Forum, 1, 139–167.
Emerson, R. M., Fretz, R. I., & Shaw, L. L. (1995). Writing ethnographic fieldnotes.
University of Chicago Press.
Glaser, B. G. (1965). The constant comparative method of qualitative analysis. Social
Problems, 12(4), 436–445. [Link]
Glaser, B., & Strauss, A. (1967). Grounded theory: The discovery of grounded the-
ory. Sociology: The Journal of the British Sociological Association, 12(1), 27–49.
Kock, N. F. (1999). Process improvement and organizational learning: The role of col-
laboration technologies. IGI Global. [Link]
MacPhail, C., Khoza, N., Abler, L., & Ranganathan, M. (2016). Process guidelines
for establishing intercoder reliability in qualitative studies. Qualitative Research,
16(2), 198–212. [Link]
O’Connor, C. & Joffe, H. (2020). Intercoder reliability in qualitative research:
Debates and practical guidelines. International Journal of Qualitative Methods,
19, 160940691989922. [Link]
Rudes, D. S., Magnuson, S., Ingel, S. N., & Hartwell, T. N. (2021). Rights-­in-­between:
Resident perceptions of and accessibility to rights within restricted housing units.
Law & Society Review, 55(2), 296–319. [Link]
Thomas, H., & Hirsch, A. (2016). A progressive’s style guide. [Link]. Retrieved
July 3, 2023, from [Link]
Sum-­Of-­Us-­Progressive-­Style-­[Link]
Trammell, R. (2012). Enforcing the convict code: Violence and prison culture. Lynne
Rienner.
Tracy, S. J. (2019). Qualitative research methods: Collecting evidence, crafting analysis,
communicating impact (2nd ed.). John Wiley & Sons.
10
TRIANGULATING PARTNERS’
VIEWS OVER TIME
Analyzing Multiple Perspective Qualitative
Longitudinal Interviews on Non-normative
Work-Care Arrangements in the Transition to
Parenthood in Practice

Susanne Vogl, Eva-­Maria Schmidt, and Ulrike Zartler

Qualitative methods are most often based on the ontological assumption of

constructed social realities. As a consequence, qualitative researchers aim to
reconstruct individuals’ meaning making, which is embedded in social con-
text and interactions. Acknowledging the intersubjectively constructed and
thereby dynamic character of social reality suggests that it can be insightful
to research more than one (individual) perspective over time—at least when
social groups are in focus of the research interest. In multiple perspective
interview (MPI) approaches, members of a social group (e.g., families) are
interviewed separately. The aim is to gain insights into the working of such
groups, the members’ relations, and their different perspectives (Deković &
Buist, 2005; Harden et al., 2010; Ribbens McCarthy et al., 2003; Vogl et al.,
2018, 2019; Zartler, 2010). In some studies, the data might also be collected
over time, and qualitative longitudinal approaches lend themselves to recon-
struct dynamics and negotiation processes within social groups over time
(Bernardi, 2021; Calman et al., 2013; Lewis, 2007; Neale, 2019; Saldaña,
2003). A multiple perspective qualitative longitudinal interview (MPQLI)
design thus takes account of the constructed and dynamic character of
social realities over time.
In this contribution, we describe the data analysis for the book chapter
“Swimming against the tide? Austrian couples’ non-­normative work-­care
arrangements in a traditional environment” (Schmidt et al., 2019), in
which we linked institutional contexts with subjective experiences of
the transition to parenthood and with work-­ care arrangements at an
individual and couple level over time. We also build on methodological

DOI: 10.4324/9781003254041-11
 Triangulating Partners’ Views over Time 161

publications derived from this study (Vogl et al., 2018, 2019). In this chap-
ter, we illustrate how triangulating perspectives over time by employing an
MPQLI approach was accomplished during the analytic process. We pro-
vide detailed accounts of analytic steps and reflections on challenges, deci-
sions, and merits. After briefly introducing the project, we focus on our
analytic practice of triangulating partners’ perspectives combined with the
longitudinal time dimension. We also explain how we developed a typol-
ogy based on the previous analytic steps. The chapter ends with our reflec-
tions about promises of MPQLIs for an in-­ depth understanding of
processes and meanings in intimate relationships, particularly, when look-
ing beyond manifest content and considering latent meanings in the
analysis.

MPQLI Research
In MPIs, the respondents’ accounts are triangulated in the analysis.
Triangulation is a term that originates in geometry, navigation, and mea-
surement (Blaikie, 1991; Flick, 2018; Kelle, 2004). The basic idea for ensur-
ing validity is that at least two sources of different types are compared to
reduce the chances of reaching “false conclusions” (Hammersley, 2008, pp.
22–23). However, this assumes an “objective” reality, and only very few
qualitative researchers would position their work within such a positivist
framework. Thus, another more recent argument for triangulation is that
relating different perspectives can lead to a more nuanced and comprehen-
sive understanding of the research issue (Fielding, 2012; Kelle, 2001). The
resulting “sometimes similar and sometimes-­ contradictory tellings”
(Santoro, 2014, p. 127) offer a variety of readings of an incident and poten-
tially facilitate greater insights into the relationships and interactions
between participants to which individual, non-­ triangulated accounts
would be blind.
Adding the longitudinal aspect of research, MPQLIs can yield insights
into the functioning of a social group from the perspective of different mem-
bers of this group over time and thus into change or stability of these per-
spectives in relation to each other. The research interest thus lies in a social
group and respective changes (or stability) across time (e.g., regarding doing
family)—yet the fundamental research units are individual members of
these groups (Vogl et al., 2018).
Although MPQLI seems to be a promising strategy, there has been little
engagement with the specific practicalities of MPQLI research, particularly
the analysis. The integration of multiple perspectives and potential respec-
tive changes over time can be a challenging endeavor. First, researchers have
to resolve the challenge of triangulating separate but related accounts that
162 Susanne Vogl et al.

can be complementary, consonant, or dissonant to one another (Vogl et al.,

2019). Second, the time dimension adds further complexity because narra-
tives can change over time and participants and researchers may reinterpret
experiences or findings with a changed frame of reference or context
(Calman et al., 2013; Lewis, 2007). The complexity and contradictions can
be frustrating at times but also make research exciting because it is full of
surprises.
Qualitative research in general evolves as an open and cyclical process. It
includes comparing and contrasting and finding similarities and differences
as principal strategies of analysis. In MPQLI research, the levels of com-
parisons are manifold: individuals within a couple, individuals across cou-
ples, couples, and (patterns of) changes across time points, as data collected
in subsequent waves can challenge or alter interpretations made in previous
waves (Vogl et al., 2018). On the one hand, triangulating different perspec-
tives can lead to a more nuanced and comprehensive understanding than
analyzing individual, non-­ triangulated accounts (Fielding, 2012; Kelle,
2001). Triangulation offers a variety of interpretations of the same matter
and potentially facilitates greater insights into the relationships and interac-
tions between participants. On the other hand, analyzing qualitative longitu-
dinal data requires the researcher to consider both a cross-­sectional and a
longitudinal dimension. In a cross-­sectional analysis, we aim to identify per-
spectives on a certain phenomenon. Longitudinally, we look at the develop-
ment of specific perspectives over time (Kirkman et al., 2001; Thomson &
Holland, 2003). Thus, accounts can turn out to be similar or convergent,
complementary as well as contradictory, and silences and gaps can also
become apparent. Overall, in our study, the MPQLI approach allowed for
richer data and deeper insights into joint and potentially changing construc-
tions of meaning within parental couples.

The Families and Societies Study

This contribution is based on an MPQLI study in the Austrian project “The
Gendered Transition to Parenthood,” which was part of the larger European
study “Families and Societies.” Between 2015 und 2018, we conducted 66
face-­to-­face problem-­centered interviews (Witzel, 2000) from 11 Austrian
couples at three time points: (a) during the last trimester of pregnancy, (b) 6
months after birth, and (c) 2 years after birth. An in-­depth consideration of
the study’s design and methods appears elsewhere (see Schmidt et al., 2019;
Zartler et al., 2021). The interviews were conducted with both partners sepa-
rately and covered a wide range of topics, such as pregnancy and prepara-
tions for birth, daily routine and time management, parental leave and
childcare arrangements, work–family–balance, expectations and perceptions
 Triangulating Partners’ Views over Time 163

of parenthood, partnership quality, parent–child relations, and characteris-

tics and development of the child.
Overall, many different analytic approaches were employed in this project,
such as interpretive/reconstructive approaches; elements of a grounded theory
approach (Corbin & Strauss, 1990); sequential analysis (Froschauer & Lueger,
2003), which is based on a hermeneutic approach (Oevermann, 2013); and
new materialist ethnography (Schadler et al., 2017). We did not apply these
different analytic strategies in a linear or step-­wise process but in an often
messy process. Nevertheless, all methods were necessary for the multiple final
products (i.e., publications). In this chapter, we only refer to papers based on
interpretive analyses with elements of sequential hermeneutics (Schmidt et al.,
2019; Schmidt, 2018) and grounded theory (Zartler et al., 2021).
In sequential hermeneutics, manifest (what is said) and latent content (how
something is said) is distinguished. First, summarizing the manifest informa-
tion helped us obtain an initial overview of the data and identify significant
text passages in the light of the research question. Second, we analyzed and
compared these passages and their context in depth by asking the following
questions: How might the situation of text generation have influenced the
sequence? In which context is such a text normally produced and how can this
context be described? Which latent meaning and frames of orientation under-
lie the sequence (i.e., norms, ideals, values, and taboos)? Which possibilities for
action are predetermined by these frames, and which are restricted?
As a team of female researchers, partly with children of our own, we had
our own normative ideas, implementation of work-­care arrangements, and
thus different attitudes and sympathies toward the interviewed couples. This
positionality formed a crucial framework for research and had to be reflected.
In joint analytic sessions, we purposefully developed multiple readings of
the interviews and tried to eliminate potential personal “biases.” We specifi-
cally questioned our reading of transcripts by one team member taking a
particularly critical role in questioning interpretations and developing alter-
native views, asking questions like “What if the opposite was true?” “What
would be a completely different reading of this passage?” or “Which other
reasons can we find for a particular statement or behavior?” We experienced
this as fruitful, but it also required discipline (and time) because no consen-
sus or agreement could be accepted outright but needed to be challenged
and defended.

General Analysis Strategies

Our approach comprised two levels of analysis: (a) the individual level vs. the
level of the couple and (b) the time level: cross-­sectional vs. longitudinal. This
approach resulted in four levels of comparisons with different (generic) aims in
164 Susanne Vogl et al.

each step (see Table 10.1): (a) individual cross-­sectional, (b) couple cross-­
sectional, (c) individual longitudinal, and (d) couple longitudinal. Given the
core interest in parents’ plans and actual arrangements of dividing work, we
focused on comparing multiple perspectives within a parental couple longitudi-
nally, and so levels of comparison (b) and (d) were core to our research interest.
However, the analyses were not as neat and straight-­lined as the table suggests;
we often struggled with constant shifts between levels of comparisons.
As a basic principle, our analytic activities were developed, refined, and
carried out as part of the interplay of data collection and analysis in qualita-
tive research. Our analytical work started with writing and reviewing field
notes and transcripts and assigning thematic codes to this data. Initially, we
used Microsoft Excel files with the interview transcript segmented in cells of
a column and assigned paraphrases and codes in further columns (see Table
10.2 for an example).

[SEE TABLE 10.2 at eResource: Microsoft Excel Sheet Comparing Individuals on a

Couple Level].

TABLE 10.1 Comparison Levels in MPQLI Data and Related Aims

Unit of analysis

Individual Couple

Time Cross-­sectional a • Respondent profile b • Profile of couple

• Typology • Comparison of
perspectives
• Insights into
interaction between
partners
• Common themes
and individual
perspectives
• Typology of couple
Longitudinal c • Individual history d • Case history of
• Change and process couple
of change • Dynamics within
• Comparison the couple
between individuals • Development of
• Typology of change perspectives in
comparison
• Typology of change
Source: Vogl, S., Zartler, U., Schmidt, E.-M., & Rieder, I. (2018). Developing an analytical
framework for multiple perspective, qualitative longitudinal interviews (MPQLI). International
Journal of Social Research Methodology, 21(2), 181. Taylor & Francis Ltd, [Link]
[Link] reprinted by permission of the publisher.
 Triangulating Partners’ Views over Time 165

Each couple had its own Microsoft Excel file, subdivided in tabs per part-
ner and per wave of data collection. This led to 11 files with six tabs each.
However, we realized that splitting the data like this turned out to hamper
cross-­case and longitudinal comparisons (Lewis, 2007). Thus, we moved to
using a qualitative data analysis software program. We first started with
[Link] and eventually switched to MAXQDA to assist with data man-
agement (see Figure 10.1 for a screenshot of our MAXQDA project with
documents, transcripts, and codings).

[SEE FIGURE 10.1 at eResource: Screenshot of MAXQDA File with Documents,

Transcripts, Coding, and a Code Memo].

In our experience, the software of choice should facilitate first and foremost
an easy overview and handling of the entire material—individual respon-
dents, couples, and waves. It also should allow parallel teamwork on differ-
ent devices and facilitate easy and safe electronic storage. For in-­depth
analysis, it was crucial to have easy access to text passages that were assigned
to a specific code. We felt that MAXQDA proved best suited for these
requirements because it allowed us to create free memos that we used to
document intentions and concerns in the coding process and thus make the
process more traceable and consistent. This was particularly important at
times when we worked separately on the dataset.
We started by assigning codes to the transcribed interviews and field
notes. Afterward, it was crucial to have easy access to text passages that we
had assigned to a specific code. We also prepared code memos (see Figure
10.1 for a memo example in MAXQDA) for each code with information on
the subsumed text passages as well as coding instructions and definitions for
further text segments. However, we did not assign memos to documents (i.e.,
transcripts or field notes) or attach them to selected text passages. For in-­
depth analysis, we retrieved text segments of one code and exported them to
conduct this analysis in an external Microsoft Word file. We also used Word’s
text search function of searching for specific texts or words during in-­depth
analysis, such as “aufteilen” (divide).

Multiple-Perspective Analysis: Comparing Partners’ Accounts

The couple analysis (also called dyadic analysis) always requires an individ-
ual analysis of the couple members’ accounts. As a first step, we analyzed
individual interviews on a cross-­sectional level. We carefully read and the-
matically coded each interview transcript using a mixture of inductive and
deductive coding. Inductive codes were identified from the material, deduc-
tive codes were already included as specific questions in the interview guide-
line. The codes were then hierarchically sorted into themes and subthemes in
166 Susanne Vogl et al.

the “code system” window of MAXQDA. Overall, we coded 4,175 text pas-
sages. The memo for the code “father’s role” or “mother’s role,” for example,
contained the following information:

Narratives about everyday life, tasks, identity, and feelings as a mother/

father, expectations of oneself as a mother/father, expectations of a
mother/father in the social surroundings, and dealing with these expecta-
tions, positive and negative aspects of being and becoming a mother/
father, comparisons with other mothers/fathers.
(see Figure 10.1 for this code memo example)

The coding allowed for and was a result of constant comparisons. In the pro-
cess of comparing and contrasting individuals, cases, and themes, we altered,
differentiated, and restructured our code structure in MAXQDA in an itera-
tive process—we deleted codes, merged codes at times, created a ­hierarchical
structure, renamed codes, and so on. These codes served as ­organization prin-
ciples to easily retrieve and compare themes across cases and times.
After close inspection and coding of the transcript of one spouse’s inter-
view, we created a profile of each respondent in separate Word files, containing
socio-­demographic information, job characteristics, parental leave plans, and
other specificities. Then, we read and coded the respective partner’s interview.
After summarizing key information for the second partner, we triangulated
both narrative accounts into case profiles for the couple. These case profiles
contained information on the couple’s story, background data, and key narra-
tives that were structured along all relevant thematic codes. Table 10.3

TABLE 10.3 Couple Profile along Thematic Codes

Patrick and Jennifer are 29 and 30 years old when their wish for a child comes true.
They got to know each other eight years before, just when they both were students
of law. They hold off on starting a family till they both have a stable and
permanent employment and income situation after having finished their studies.
During pregnancy, they start to plan the time of parental leave after the baby’s
birth and try to find an individual solution which is adjusted to his and her
employment situation as well as to the child’s well-­being. For this reason, they have
to gain expertise in employment law (for their rights to parental leave) as well as in
the regulations concerning the Austrian childcare benefit system (which is
independent of the legal regulations concerning employment). After the joint
decision to divide parental leave times and to divide the parts equally, the parents
face several challenges and troubles. First, they have to discern that the existing
system of childcare benefit does not seem to support their wishes for equal
opportunities and an equal division of parental leave. They being to fight against
the—in their view—inflexible and normative system which forces the father into
two fixed months of parental level at a given point of time which he cannot choose.
 Triangulating Partners’ Views over Time 167

illustrates the initial version of a couple profile. The structuring codes were
“wish for a child,” “employment,” “child wellbeing,” and “social policy.”
The profiles were revised and extended several times following insights from
further analyses. With constant comparison the profiles became more struc-
tured. Case profiles thus evolved in an ongoing analytic process with growing
analytic depth, based on our different interpretations that were discussed dur-
ing joint analytic sessions that took place in person in our offices once a week
on average and at least once for each couple. The different lines of comparisons
between cases, couples, and across time fostered abductive thinking—thus, new
ideas and readings then had to be integrated into the couple profiles. In this
process, we identified, tested, and sharpened assumptions, thoughts, hypothe-
ses, and interpretations on patterns, shared or not shared rationales, and con-
structions of meanings, and also noted open questions. The developed
hypotheses fed into evolving theory memos and were saved as separate Word
documents. Compared to case profiles, these memos contained general and
abstract thoughts far beyond manifest content of specific interviewees or cou-
ples. Document 10.1 gives an impression of an evolving theory memo.

[SEE DOCUMENT 10.1 at eResource: Evolving Theory Memo Case 11].

Besides convergent and complementary accounts, triangulating perspectives

often led to discrepancies and divergence (Vogl et al., 2019) that required
further interpretation, reinterpretation, and, consequently, led to a more
nuanced understanding. It was important to go beyond manifest statements
and consider latent meaning. On the descriptive level of analysis, we ana-
lyzed the manifest content, i.e., carefully and repeatedly read the transcript,
connected interview content to codes, paraphrased the superficial and every-
day sense of the content, and aimed at enriching themes. On the interpretive
level, we went beyond manifest content in order to reconstruct respondents’
subjective system of relevance. We accomplished this by posing questions to
short segments of text (a few words or a part of a sentence), like which con-
textual and latent preconditions led to this statement, and which implicit
assumptions, latent meanings, and guiding notions led to it?
For example, Jim made the following statement referring to the timing of
pregnancy, “The main point for us was that I don’t have to study anymore,
but that I can work” (Jim, part-­time freelancer at a bank, doing a bachelor’s
degree at the time of the first interview). We asked questions like the follow-
ing: Why is it important for the man (but not for the woman) to be employed
when a woman becomes pregnant? What might have led to the view that the
man is responsible for earning (a particular amount of) money? How is the
adequate amount of money determined and negotiated? How is the respon-
dents’ statement related to his perception of others’ expectations of his role
168 Susanne Vogl et al.

as a (becoming) father? Which consequences were perceived as ‘bad’ timing

(i.e., pregnancy before graduation)? How was the right timing negotiated
within the couple? For this interviewee, it seemed unquestioned that his part-
ner would stop working with a child and he would be the sole breadwinner.
The statement shows a very clear idea of gender-­specific division of parental
tasks that both partners share (as the “for us” suggests). A solid professional
position was a condition for having a child.
On the descriptive level of analysis, we sometimes found dissonance
between partners that could be (partly) resolved by further interpretive anal-
yses. We regarded dissonance as a starting point for further interpretive anal-
ysis rather than a threat to validity or trustworthiness. Table 10.4 contains
parts of a couple profile on the triangulation of accounts and shows how the

TABLE 10.4 
Dissonance in Triangulating Perspectives: Julia and Jim Talk about
Their Plans for Taking Parental Leave in Wave 1

When talking about their parental leave plans, each respondent presented a story
that seemed to be very different from the other’s. After being asked if Jim also
planned to go on parental leave, Julia did not dismiss the possibility. She assumed
that he would like to do so but noted that a potential parental leave had to be
compatible with his career plans and the reconciliation of his employment and
studies.
Julia:I would be very glad for him and I think he would do it with pleasure.
Jim, in contrast, precluded the possibility of going on leave in his account, even
without being specifically asked by the interviewer.
Jim:But for her (boss) this has no importance whether my wife is pregnant or
not because I don’t take parental leave anyway.
Thus, on a descriptive level, the accounts were dissonant. However, the interpretive
analysis revealed the interrelation of the partners’ perspectives and was able to
explain the dissonance. First, the data contained further indications that Julia
wished for a shared parental leave and hoped to convince her partner, not with a
strong conviction but rather out of a sense of social desirability. Furthermore, we
also found indications for Jim’s strong sense of responsibility for being the
breadwinner and investing in his career. More detailed analysis revealed that Julia
had sensed very early that Jim would probably not take parental leave. Second,
the dissonance might have been instigated by the interview situation itself. Julia
might have been reluctant to admit to a female interviewer that her partner was
not willing to share parental leave. Jim might have been more open about his
reluctance to take parental leave in the interview situation than toward Julia in
order not to disappoint her.
Source: Vogl, S., Schmidt, E-­M., & Zartler, U. (2019). Triangulating perspectives: Ontology and
epistemology in the analysis of qualitative multiple perspective interviews, International Journal
of Social Research Methodology, 22(6), 616–617. Taylor & Francis Ltd, [Link]
com reprinted by permission of the publisher.
 Triangulating Partners’ Views over Time 169

dyadic analysis changed the individual-­level results and fostered deeper

understanding of processes and negotiations within the relational unit. The
explanation for dissonance was in the interview situation, and the respon-
dents’ “impression management” related to social desirability and probably
wishful thinking.
However, we could not always explain or resolve dissonance occurring at
the descriptive level. Table 10.4 shows, for example, how even after intense
interpretive analysis and additional external research, divergence could not
be resolved. The triangulation was sometimes a frustrating process. We
struggled with accepting dissonance on a manifest level and tried hard to
find explanations in one example for why the partners both credibly reported
different dates for re-­entering the job market. The (small) time difference of
one month between both partner’s accounts is important because Austrian
fathers usually take very short leaves of two months. After additional
research and repeated reading of the transcripts, we eventually chose to
accept the dissonance (see Table 10.5).
Furthermore, consonance, agreement, or convergence at the descriptive
level instigated a second look and entailed a deeper understanding (Ribbens
McCarthy et al., 2003; Sands & Roer-­Strier, 2006; Zartler, 2010). In our
practice, one team member took the role of questioning convergence on the
descriptive level. On the interpretive level, we then sometimes detected
divergence. Table 10.6 reflects how accounts appeared to be convergent
could be dissonant upon further interpretation taking both perspectives
into account.
The interpretive analysis of MPQLI produced added value because it
sometimes challenged descriptive outcomes of perspective triangulation,
regardless of whether dissonant data occurred on a descriptive level but
“disappeared” on an interpretive level or vice versa. Intense interpretive
analysis often offered deeper and sometimes unexpected insights and a
greater sensitivity for the complexity of relationships and social
systems.
As a result of the couple level analysis, we thus had an inventory of
central themes represented by codes with code memos and thematic memos,
attached to codes and as free memos in MAXQDA. For the couple level
analyses, we stored thematic memos as Word files but referred to the cod-
ing list in MAXQDA. Furthermore, we structured theory memos and cou-
ple profiles according to theoretically relevant dimensions (e.g., parental
responsibilities and gender roles), which were developed through constant
comparison and through analyzing at both descriptive and interpretive lev-
els. Working with memos, tables, and text in Microsoft Word files, we could
then group cases (couples) into clusters (e.g., all couples who realized
parental leave in a non-­normative way, see Table 10.7) and eventually
170 Susanne Vogl et al.

TABLE 10.5 Dissonance in Triangulating Perspectives: Chris’ and Claudia’s Story

about Chris’ Re-entry to Work
Contradictions on a descriptive level occurred regarding Chris’ re-­entry to work
after 18 months of leave. This issue arose in the third wave of interviews that took
place around the second birthday of their daughter and some months after Chris’
parental leave ended.
After having analyzed Chris’ interview, we concluded that his re-­entry to a new
full-­time job was in September. He said that he and Claudia had planned for a
summer together with their daughter from the beginning, and his return to
full-­time employment in September. He concluded that luckily this plan had
worked out because he was able to find a new job starting in September.
Chris:Luckily it so happened that I returned to fulltime employment in
September.
Claudia’s narration put this view into question. In contrast to Chris, she repeatedly
mentioned August as the time Chris started to work full-­time in his new job and
elaborated on how this changed the whole family routine. She provided very
detailed descriptions of her feelings about that time with a constant reference to
August as the starting date.
Claudia:My husband was at home until the end of July, and in August he
started his new job.
The dissonance in Chris’ and Claudia’s accounts could not be explained or
resolved. Rather, intensive analytical processes on an individual level supported
each account, even though their accounts were clearly contradictory on a
descriptive level. Claudia remembered the summer very vividly and emotionally
because the heat—which could be verified by weather reports for August in the
relevant year—made it difficult for her to handle daily life with her daughter on
her own, which was a new situation for both the daughter and the mother.
Chris also provided very credible accounts, emphasizing September as his month of
re-­entry. Analyzing the data on an individual level, it became clear that Chris had
been very much looking forward to re-­entering work. For him, it meant the end
of a non-­normative and temporary phase of his life, returning to what he viewed
as the ‘natural’ responsibility of earning money through full-­time employment.
There was no indication that he could have made a mistake in the time reference.
Source: Vogl, S., Schmidt, E-­M., & Zartler, U. (2019). Triangulating perspectives: Ontology and
epistemology in the analysis of qualitative multiple perspective interviews, International Journal
of Social Research Methodology, 22(6), 617. Taylor & Francis Ltd, [Link]
reprinted by permission of the publisher.

started to construct a typology. This procedure could also have been

accomplished with the “document variables” tool in MAXQDA and linked
memos, functions that we were not familiar with at the time of our analy-
sis. In our note displayed in Table 10.7, the words after the arrow indicate
a reference to a code.
 Triangulating Partners’ Views over Time 171

TABLE 10.6 Complementary Accounts in Triangulation: Maria and Max Talk about
Re-traditionalization
After childbirth, Maria did not indicate the intention to resume her self-­
employment, while Max returned to his decently paid full-­time job after two
months of leave. Maria was satisfied with the situation and did not feel “forced”
to earn or contribute financially. However, she was aware that her income would
be appreciated because she and Max had to repay a loan for their house.
Maria:It’s ideal because right now I was in the fortunate situation of not
necessarily being forced to earn a living. I mean, it was the plan and it
would also be better and house, loan, and, yes. But it, it’s not that I
absolutely have to.
Max explained that Maria could not start employment yet because she wanted to
complete additional training first. He stated that they were fortunately able to
afford the wait and that Maria’s income was not immediately necessary; however,
he had a more restrictive view than Maria. Although they used almost identical
terms, the context was different and legitimized different interpretations. It
became apparent that he was not particularly happy about Maria’s further
training. He did not value her time studying for admission tests and actually
belittled it. Nevertheless, he thought she could and should at least obtain
part-­time employment like other mothers do.
Max:But Maria has to, well, somehow prepare for another admissions exam.
But ultimately, it doesn’t really matter, because, because others work
part-­time. So, it is, somehow—either you have a job or you don’t.
Effectively, Max was dissatisfied experiencing sole responsibility for the household
income—particularly because he had originally planned to reduce his working
hours after his leave. Thus, he implicitly wished Maria would support him with
family income instead of preparing for one more admission exam. On a
descriptive level, both partners convergently saw no strict necessity for Maria’s
contribution to the household income. On an interpretive level, it became
apparent that Max was reluctant to be the sole breadwinner. Maria, in contrast,
was very happy about this situation.
Source: Vogl, S., Schmidt, E-­M., & Zartler, U. (2019). Triangulating perspectives: Ontology and
epistemology in the analysis of qualitative multiple perspective interviews, International Journal
of Social Research Methodology, 22(6), 618. Taylor & Francis Ltd, [Link]
reprinted by permission of the publisher.

Longitudinal Analysis: Comparison across Time

The steps for the cross-­sectional analysis were then repeated for further
waves and extended for comparisons across all three waves to determine sta-
bility and changes. In the longitudinal analysis, we again started with the
individual analysis comparing interviews across waves. We extended the cod-
ing system for codes for stability and change as well as new themes. In a
172 Susanne Vogl et al.

TABLE 10.7 Excerpt from a Note on Clusters of Couples: Example of a Typical

Couple, Their Difficulties, and Feelings
Couples that manage to realize their non-­normative work-­care arrangement
Couple 1: Tina and Tim
Comparison within individuals and couples: Before realizing their wish for a child,
they both try to stabilize their career and to look for stable ➔ jobs. They also
delay the realization of their wish, as Tina starts a new job and thus wants to
avoid being pregnant within the first months. Although Tom enters a new,
time-­consuming and somehow precarious job as well, they then decide to pursue
their wish for a child again. After about six months, Tina becomes pregnant.
They both start to gain a lot of information about being pregnant, about
upcoming necessities, about legal structures, and both start to think about
upcoming decisions. Both attend gynecological examinations and midwife’s visits.
Tom tries to adjust his time schedule at work in order to be at home earlier, to do
the shopping, and to prepare breakfast and dinner. Tom experiences Tina’s entry
into maternity protection as a turning point: the amount of time they can invest
into the pregnancy clearly drifts apart at that moment.
Their ➔ plans of arranging work and care responsibilities during pregnancy: Both
aim at creating and providing egalitarian opportunities for their career, and at
sharing parental leave and responsibilities in an equal way (each parent 12 months
leave, afterward both in full-­time employment), but during pregnancy, both do not
feel able to realize their visions due to structural constraints of the Austrian
childcare allowance system (which they do not understand entirely correct even
though they are both lawyers) and of availability of childcare facilities. Even
though this couple has egalitarian and non-­normative plans regarding their
work-­care arrangements, their narratives implicitly—and increasingly over the
examined period—also reflect normative attitudes of ultimately ascribing main
responsibility for caregiving to the mother and ascribing ultimate responsibility for
the family income to the father ➔ gender ideologies.

similar vein, code memos and thematic and theory memos were extended. In
the next step, we extended case profiles to case histories. Case histories pro-
vide accounts of how and why events unfolded and how the individuals and
cases have been transformed over time (Thomson, 2007). Again, case histo-
ries were structured by themes relevant for the research question at hand.
Table 10.8 showcases the section of couples’ history of negotiations about
parental care arrangements.
As the example illustrates, cross-­sectional couple profiles were the basis for
longitudinal analysis. Similar to cross-­sectional couple analysis, we analyzed
the couples across waves considering similarities and differences, using con-
stant comparison, and identified collective themes and respective changes
within and between couples. In addition, we followed themes within couples
across time and developed new theoretically relevant themes. Thus, levels of
comparisons extended from cross-­sectional between partners to a longitudinal
 Triangulating Partners’ Views over Time 173

TABLE 10.8 Maria and Max’ Parental Care Plans over Time

In Wave 1, Maria and Max considered egalitarian distribution of care and

household tasks essential and arranged their joint daily lives accordingly. They
planned to continue equally sharing tasks after the birth. Their motivations for
this cooperative, shared responsibility were on financial, political, practical,
ideological, and professional levels. Both planned to take parental leave for at
least one year each, with each of them working part time afterward. However,
Max subliminally questioned the feasibility of this arrangement from a financial
perspective while Maria strongly argued in favor of shared responsibilities, saying
“Why should I do all that on my own?” (Maria, Wave 1). Comparing individual
narratives over time showed that, in Wave 2, Maria felt “much more exhausted
than [she] would have thought before” (Maria, Wave 2). She was still on parental
leave in Wave 3, as her attempt to re-­enter the labor market failed because she felt
too stressed. Also, her expectations regarding egalitarian task sharing had
changed: “I don’t need to have a fifty-­fifty allocation of housework (…) but I still
insist on a little bit of equality” (Maria, Wave 2). In Wave 3, Maria considered it
“natural” not being employed with a small child. She now understood this
arrangement to be “ideal” and beneficial, and considered herself “fortunate not to
be forced to earn money” (Maria, Wave 3).
Although Max took a “daddy month” immediately after delivery, he continued his
professional activity. In Wave 2, he worked full-­time and still considered taking
parental leave, however, only for two months over the summer. Part-­time work
after a longer parental leave—as envisaged in Wave 1—was no longer planned.
Max felt “massively restricted” (Max, Wave 2) because he had reduced his leisure
time activities due to a lack of energy. Overall, he increasingly experienced a
strong constraint in his quality of life. Max was happy to be able to provide for
his family, but still expressed his view that both partners should work 30 hours
(although he was in full-­time employment also in Wave 3). Maria’s accounts
illustrated a reinterpretation of the importance of the egalitarian distribution of
tasks and exemplified re-­traditionalization. For Max, we also found a subjective
reinterpretation of his contribution and parental leave.
Relating to perspectives over time, we identified considerable inconsistencies. For
example, Maria (Wave 2) estimated that Max reduced his work hours to 40%—
mentioning that he would claim it was 10%. Max himself said he had reduced his
workload to 20% or 30%. Furthermore, Maria complained about Max’ multiple
time-­consuming activities beyond his job and his family, while Max complained
about a drastic reduction in his time for hobbies and other interests. Almost
inevitably, conflicts emerged.
Source: Vogl, S., Zartler, U., Schmidt, E.-M., & Rieder, I. (2018). Developing an analytical
framework for multiple perspective, qualitative longitudinal interviews (MPQLI). International
Journal of Social Research Methodology, 21(2), 186. Taylor & Francis Ltd, [Link]
[Link] reprinted by permission of the publisher.
174 Susanne Vogl et al.

dimension. We also deepened our understanding in the longitudinal couple

analysis, which allowed for insights into the dynamics of this couple and the
development of related perspectives over time. For example, Tom and Tina
(see Table 10.7) were described as “implementing ideals,” whereas, in contrast,
Maria and Max (see Table 10.8) were described as “adjusting to re-­
traditionalisation” (see Vogl et al. 2018).
Triangulating partners’ perspectives over time enabled us to explore pat-
terns and dynamics over time and to establish a typology of change. For the
couples, change could be identified for both partners or only in one person,
and partners could diverge or converge over time or converge, diverge, and
converge again. Convergence and divergence, dissonance and consonance,
thus, occurred at the cross-­sectional couple level (see above) but also longitu-
dinally and at both the individual and the couple level. Table 10.9 illustrates
this analytical step that referred to the conceptual question about whether
breadwinning should be regarded as a care activity (Schmidt, 2018). It dis-
plays a draft of a typology-­table grasping the dynamics in couples’ conver-
gence or divergence regarding the shared meaning of paid work both
cross-­sectionally between partners and longitudinally (i.e., between waves).
Looking at this, we were astonished by the number and variety of types of
normative work-­care arrangements. Only Type 1 represents non-­normative
arrangements. As a research team, we shared special interest in this type—not
only due to theoretical reasons and to our own life contexts but also because
these couples put a lot of effort in justifying their arrangements toward the
interviewers. Thus, type 1 couples became the basis for deeper analysis for our
“swimming against the tide” publication (Schmidt et al., 2019).

[SEE TABLE 10.9 at eResource: Breadwinning as Care? Draft of a Typology Table].

Concluding Reflections
In this chapter, we have illustrated the analytical practice of MPQLI research
with examples from a study on the transition to parenthood and provided
examples of triangulating perspectives over time. Our examples show the
many promises MPQLI holds for an in-­depth understanding of processes
and dynamics in intimate relationships.
Triangulating perspectives provides evidence for research to make sense of
a social phenomenon, but a triangulation strategy does not, in and of itself,
provide the sense (Mathison, 1988). Methodologies are not tools we pick from
a toolbox. Instead, they are forms of performative, interpretive practice
(Denzin, 2010), and researchers shape research with their own ontological and
epistemological views. Thus, we advocate in favor of a critical reflection of our
perspectives and interpretations as researchers that lead to analytic results.
 Triangulating Partners’ Views over Time 175

With MPQLI, potential levels of comparisons multiply compared to

individual cross-­sectional approaches. However, it is not necessary to com-
pare everything with everything else to adequately utilize the multilayered
data. Rather, a systematic plan can be used (Boeije, 2002). Developing such
a systematic plan with a clear outline of analytical steps taken, levels of
comparisons, and clearly stated aims (and reminders of them) is crucial in
every research phase. Otherwise, the volume of the data may easily become
unwieldy, and any analysis can lack in traceability.
In practice, the possibility of research results that change from disso-
nance to convergence on different levels of analyses indicates that we should
not accept convergence too fast but continue to query and interpret further
because we could gain additional insights and a deeper understanding. We
therefore strongly recommend going beyond manifest statements in the
descriptive analysis of themes to consider latent meaning in the interpretive
analysis of MPQLI. With a reconstructive approach, we can distinguish
between explicit knowledge (i.e., subjective representations) and tacit knowl-
edge or (genesis of) action orientations (Hollstein & Kumkar, 2021). Intense
interpretive analysis often offers deeper and sometimes unexpected insights
and a greater sensitivity for the complexity of relationships and social sys-
tems. Triangulating perspectives also reveals gaps, silences, and contradic-
tions that can instigate further analysis and abductive reasoning. Overall,
this approach provides numerous merits for reaching an in-­depth under-
standing of meanings, anchored in different perspectives, and processes that
develop over time.

Acknowledgment
The research leading to these results received funding from the European
Research Council under the European Union’s Seventh Framework
Programme (FP7/2007–2013), Grant Agreement Number 320116, and from
the University Jubilee Foundation of the City of Vienna under Grant num-
ber H-­284605/2015.

KEY WORKS GUIDING OUR DATA ANALYSIS

Lewis, J. (2007). Analysing qualitative longitudinal research in evaluations. Social Policy

and Society, 6(4), 545–556. [Link]
In this article, Lewis exemplifies the analytic process of a qualitative longitudinal
evaluation study. She conceptualizes different types of change: narrative change,
reinterpretation by participants or researchers, and absence of change. She shows
how data can be read differently in cross-­sectional and longitudinal analysis, along
themes, cases, and in group analysis.
176 Susanne Vogl et al.

Ribbens McCarthy, J., Holland, J., & Gillies, V. (2003). Multiple perspectives on the ‘fam-
ily’ lives of young people: Methodological and theoretical issues in case study
research. International Journal of Social Research Methodology, 6(1), 1–23. https://
[Link]/10.1080/13645570305052
The authors engage in an important conceptualization of analysis of multiple
perspective interviews regarding similarity and divergence in the light of objectivist
and interpretationist ontologies. They show that it is important to consider the
researcher’s “bird’s eye view” in the interpretation of disagreements and agree-
ments between related interviews. They show how choices and viewpoints of
researchers in the analysis and presentation of results influence findings and argue
for openness and reflexivity concerning the difficult analytic choices and representa-
tion of complexity and tension between accounts.
Flick, U. (2018). Triangulation. In N. K. Denzin & Y. S. Lincoln (Eds.), The SAGE handbook
of qualitative research (pp. 444–461). Sage.
Flick contextualizes the history and different meanings of the triangulation meta-
phor as well as foundations, types, outcomes, and interpretation of triangulation.
Although triangulation is generally associated with combining several methodolog-
ical (or theoretical) approaches and rose in popularity with the mixed methods
movement, it offers a useful framework for multiple perspective interviews.

References
Bernardi, L. (2021). Qualitative longitudinal research in family sociology. In N. F.
Schneider & M. Kreyenfeld (Eds.), Research handbook on the sociology of the fam-
ily (pp. 107–124). Edward Elgar. [Link]
Blaikie, N. W. H. (1991). A critique of the use of triangulation in social research.
Quality & Quantity, 25(2), 115–136. [Link]
[Link]
Boeije, H. (2002). A purposeful approach to the constant comparative method in the
analysis of qualitative Interviews. Quality & Quantity, 36(4), 391–409. [Link]
org/10.1023/A:1020909529486
Calman, L., Brunton, L., & Molassiotis, A. (2013). Developing longitudinal qualita-
tive designs: Lessons learned and recommendations from health services research.
BMC Medical Research Methodology, 13, Article 14. [Link]
com/1471-­2288/13/14
Corbin, J., & Strauss, A. (1990). Grounded theory research: Procedures, canons and
evaluative criteria. Zeitschrift für Soziologie, 19(6), 418–427.
Deković, M., & Buist, K. L. (2005). Multiple perspectives within the family: Family
relationship patterns. Journal of Family Issues, 26(4), 467–490. [Link]
10.1177/0192513X04272617
Denzin, N. K. (2010). Moments, mixed methods, and paradigm dialogs. Qualitative
Inquiry, 16(6), 419–427. [Link]
Fielding, N. G. (2012). Triangulation and mixed methods designs: Data integration
with new research technologies. Journal of Mixed Methods Research, 6(2), 124–
136. [Link]
Flick, U. (2018). Triangulation. In N. K. Denzin & Y. S. Lincoln (Eds.), The SAGE
handbook of qualitative research (pp. 444–461). Sage.
 Triangulating Partners’ Views over Time 177

Froschauer, U., & Lueger, M. (2003). Das qualitative interview. Facultas wuv.
Hammersley, M. (2008). Troubles with triangulation. In M. M. Bergman (Ed.),
Advances in mixed methods research (pp. 22–36). Sage.
Harden, J., Backett-­Milburn, K., Hill, M., & MacLean, A. (2010). Oh, what a tan-
gled web we weave: Experiences of doing ‘multiple perspectives’ research in fami-
lies. International Journal of Social Research Methodology, 13(5), 441–452. https://
[Link]/10.1080/13645571003650979
Hollstein, B., & Kumkar, N. C. (2021). Qualitative methods. In B. Hollstein, R.
Greshoff, U. Schimank, & A. Weiß (Eds.), Soziologie: Sociology in the German-­
speaking world: Soziologische Revue 2020 (pp. 301–314). De Gruyter Oldenbourg.
[Link]
Kelle, U. (2001). Sociological explanations between micro and macro and the inte-
gration of qualitative and quantitative methods. Forum Qualitative Sozialfor­
schung/Forum: Qualitative Social Research, 2(1). [Link]
fqs-­2.1.966
Kelle, U. (2004). Integration qualitativer und quantitativer Methoden. In U.
Kuckartz, H. Grunenberg, & A. Lauterbach (Eds.), Qualitative datenanalyse com-
putergestützt (pp. 27–41). VS Verlag.
Kirkman, M., Harrison, L., Hillier, L., & Pyett, P. (2001). ‘I know I’m doing a good
job’: Canonical and autobiographical narratives of teenage mothers. Culture,
Health & Sexuality, 3(3), 279–294. [Link]
Lewis, J. (2007). Analysing qualitative longitudinal research in evaluations. Social
Policy and Society, 6(4), 545–556. [Link]
Neale, B. (2019). What is qualitative longitudinal research? Bloomsbury Academic.
Mathison, S. (1988). Why triangulate? Educational Researcher, 17(2), 13–17. https://
[Link]/10.3102/0013189X017002013
Oevermann, U. (2013). Objektive hermeneutik als methodologie der erfahrungswis-
senschaften von der sinnstrukturierten welt. In P. C. Langer, A. Kühner, & P.
Schweder (Eds.), Reflexive wissensproduktion (pp. 69–98). Springer Fachmedien
Wiesbaden.
Ribbens McCarthy, J., Holland, J., & Gillies, V. (2003). Multiple perspectives on the
‘family’ lives of young people: Methodological and theoretical issues in case study
research. International Journal of Social Research Methodology, 6(1), 1–23.
[Link]
Saldaña, J. (2003). Longitudinal qualitative research: Analyzing change through time.
Altamira Press.
Sands, R. G. & Roer-­Strier, D. (2006). Using data triangulation of mother and
daughter interviews to enhance research about families. Qualitative Social Work,
5(2), 237–260. [Link]
Santoro, N. (2014). Using a multiple perspectives framework: A methodological
approach to analyse complex and contradictory interview data. Ethnography and
Education, 9(2), 127–139. [Link]
Schadler, C., Rieder, I., Schmidt, E.-M., Zartler, U., & Richter, R. (2017). Key prac-
tices of equality within long parental leaves. Journal of European Social Policy,
27(3), 247–259. [Link]
Schmidt, E.-M. (2018). Breadwinning as care? The meaning of paid work in moth-
ers’ and fathers’ constructions of parenting. Community, Work and Family, 21(4),
445–462. [Link]
178 Susanne Vogl et al.

Schmidt, E.-M., Zartler, U., & Vogl, S. (2019). Swimming against the tide? Austrian
couples’ non-­normative work-­care arrangements in a traditional environment. In
D. Grunow & M. Evertson (Eds.), New parents in Europe: Couples in between
norms and work-­care practices (pp. 110–129). Edgar Elgar.
Thomson, R. (2007). The qualitative longitudinal case history: Practical, method-
ological and ethical reflections. Social Policy and Society, 6(4), 571–582. https://
[Link]/10.1017/S1474746407003909
Thomson, R., & Holland, J. (2003). Hindsight, foresight and insight: The challenges
of longitudinal qualitative research. International Journal of Social Research
Methodology, 6(3), 233–244. [Link]
Vogl, S., Schmidt, E.-M., & Zartler, U. (2019). Triangulating perspectives: Ontology
and epistemology in the analysis of qualitative multiple perspective interviews.
International Journal of Social Research Methodology, 22(6), 611–624. [Link]
org/10.1080/13645579.2019.1630901
Vogl, S., Zartler, U., Schmidt, E.-M., & Rieder, I. (2018). Developing an analytical
framework for multiple perspective, qualitative longitudinal interviews (MPQLI).
International Journal of Social Research Methodology, 21(2), 177–190. [Link]
org/10.1080/13645579.2017.1345149
Witzel, A. (2000). Das problemzentrierte interview. Forum Qualitative Sozialforschung/
Forum: Qualitative Social Research, 1(1), Article 22. [Link]
[Link]/[Link]/fqs/article/view/1132/2520
Zartler, U. (2010). Multiple perspectives in qualitative family research: Crafting and
conducting research projects. Family Science, 1(3–4), 173–182. [Link]
1080/19424620.2010.569371
Zartler, U., Schmidt, E.-M., Schadler, C., Rieder, I., & Richter, R. (2021). “A blessing
and a curse”: Couples dealing with ambivalence concerning grandparental
involvement during the transition to parenthood—A longitudinal study. Journal
of Family Issues, 42(5), 958–983. [Link]
11
MEDIA PRIMING AND RACIALIZED
PRODUCTION DECISIONS IN
COLLEGE FOOTBALL BROADCASTS*
Extrapolating Strategies for Analyzing
Video Data

Sara E. Grummert and Siduri J. Haslerig

In this chapter, we reflect on the process of analyzing 118 hours of video

broadcast footage that was presented in our 2020 article “Invincible Bodies:
American Sport Media’s Racialization of Black and White College Football
Players” in the International Review for the Sociology of Sport (Haslerig
et al., 2020). By analyzing the commentary, visuals and graphics, and pro-
duction decisions of ESPN’s College GameDay, our findings captured the
insidious ways in which sport media reinforces the dehumanization of
Black people.
Throughout this chapter, we detail the processes we engaged in to cohe-
sively analyze such a large, complex dataset. Several methodological deci-
sions had to be made given the nuances within and complexity of the video
broadcasts. We reflect on these coding decisions as well as the memoing
practices that scaffolded our analysis. Additionally, we describe the ways we
used Dedoose, a qualitative and mixed methods data analysis application, to
assist in the management and analysis of the data. Dedoose was essential for
employing the constant comparative method on video data, facilitating the
data’s categorization into themes that we could seamlessly review and sub-
code clip-­by-­clip. In combination with Dedoose, we also utilized Microsoft
Excel to further unpack our coding related to injury. This enabled us to show
the significance of the racial disparities related to injury on a greater, and
more intellectually honest, scale.

* Author order was determined alphabetically and represents equal distribution of effort.

DOI: 10.4324/9781003254041-12
180 Sara E. Grummert and Siduri J. Haslerig

Study Origins and Evolution

The study was initially intended to document and examine discourse, includ-
ing racialized discourse, surrounding academics for graduate student and
postbaccaleaureate athletes within college football (see Haslerig, 2013, 2017,
for more information on this subpopulation of college athletes). The study
was initially an organic extension of Siduri’s research exploring the experi-
ences of graduate and postbaccalaureate football athletes and the portrayal
of this subpopulation to the public in media broadcasts. Siduri argued that
graduate student athletes and those doing postbaccalaureate coursework
were frequently misrepresented in the media by the terms “redshirt senior”
or “fifth-­year senior,” despite having already earned a degree. After seeing
Siduri present on this phenomenon in March 2016, a colleague, Rican Vue,
suggested we collaborate to document how often and in what ways academ-
ics were discussed during game broadcasts. Initially, this was supposed to be
a small study, laser focused on documenting when, how, and for whom aca-
demics were discussed by in-­game commentators. In hindsight, that is not
quite what we did; instead, the study grew exponentially.
In summer 2016, Sara started working on the research project. In the
years leading up to the project, Sara worked in professional sport marketing
with the National Basketball Association (NBA) and Women’s National
Basketball Association (WNBA). Over the course of her work experience,
she became increasingly interested in how sport functioned, particularly in
relation to sport’s role in—and impact on—society, and she left her market-
ing role to begin researching college sport and sport media. Critical theory
(e.g., intersectionality and critical race theory) spurred her to reflect on her
whiteness and positionality, and her research shifted to attend to how media
used sport to reinforce and propagate racist and sexist ideologies and, para-
doxically, how sport could be used to combat them. Our full research team
included Siduri, a Black biracial woman; Sara, a white woman; Rican, a
Hmong American woman; and Emma Allen, an Indigenous woman, who
assisted in the final waves of Microsoft Excel analysis. We each brought our
social locations and life experiences to the research, and the process of
understanding each other’s thinking and coming to consensus often made
our analysis richer and more trustworthy. We brought our whole selves to
the research and it impacted every aspect of this study—some of which we
discussed explicitly and some of which was naturally embedded in our inter-
pretations and discussions.
Siduri and Sara collaborated closely to decide which broadcasts would be
most significant to analyze, which we discuss in more depth in the Data
Collection section. As the scope of the study rapidly expanded during early
data collection in fall 2016, Siduri and Rican renegotiated the level of
 Priming and Racialized Production in College Football Broadcasts 181

involvement that Rican was able to contribute. As a result, throughout the

project, Siduri and Sara worked closely on data collection and analysis,
while Rican was more involved in writing up parts of the literature review
and discussion, particularly in relating the findings to racialization and
dehumanization. In the following sections, we provide context about the
guiding theory and methods, how we accomplished collecting and managing
such a large dataset, and how we coded and analyzed the many layers of
data. Prior to the 2020 article (Haslerig et al., 2020), we published a chapter
in an edited book, Rationalizing Black Death: Sport Media’s Dehumanizing
Coverage of Black College Football Players (Haslerig et al., 2019), which laid
out part of our argument about the societal consequences of how Black
players were framed. This enabled us to focus more explicitly on the com-
parative data between the coverage of white and Black players in the 2020
article (Haslerig et al., 2020).

Theoretical and Methodological Alignment

Our study was guided by constructivist epistemology, emphasizing that
knowledge is co-­constructed and informed by our personal experiences,
ideas, and contexts (Bhattacharya, 2017; Piaget & Inhelder, 1967). Agenda-­
Setting Theory also informed our methodology and analysis. According to
Agenda-­Setting, media and broadcasters direct the public’s attention to spe-
cific subjects, setting the topics and foci of topics for public political dis-
course, and thus influencing the importance and salience of topics (McCombs
& Shaw, 1972; McCombs et al., 1997). Stemming from the field of psychol-
ogy, Agenda-­Setting is often used to analyze news texts to examine the psy-
chological effects of the agenda-­setting function of mass media on audience’s
political orientations and attitudes (McCombs & Shaw, 1972; McCombs
et al., 1997; McCombs et al., 2014). Others have built on Agenda-­Setting to
also consider how priming and framing operate within Agenda-­Setting
(Carragee & Roefs, 2004). Priming has been explained as the second level of
Agenda-­Setting in which media associates or “primes” the topic or subject
put on the agenda with specific concepts or attributes (Moy et al., 2016). As
explained by Moy et al. (2016, p. 5):

A media priming effect occurs if, in the second step, the receiver applies
the primed, now more accessible concept to a target stimulus when s/he
would not otherwise have done this. The first step thus consists of the
priming process, and the second speaks to its consequences.

The impact of priming (i.e., the strength of the priming effect) depends on (a)
how recently it is presented/the relative distance between the presentation of
182 Sara E. Grummert and Siduri J. Haslerig

it and a related topic/subject; (b) repetition, or the frequency with which the
concept is presented; and (c) the applicability and subject relevance the prim-
ing concept has with the intended topic/subject (Moy et al., 2016). In the case
of our 2020 article (Haslerig et al., 2020), for example, we documented how
College GameDay broadcasts primed the audience to believe that Black play-
ers were indestructible and/or impervious to pain (and, more importantly,
that they were therefore an unstoppable threat, even when injured). In turn,
we argued that the subsequent priming effect may impact a pressing social
issue—the use of excessive force against Black people—by biasing police and
others to inaccurately view excessive (indeed, often deadly) force as reason-
able and by rationalizing such force to bystanders and the public.
Framing refers to how media members choose to present or “frame” a
topic with characteristics or perspectives, thus cultivating salience (Moy
et al., 2016). In our study, the pregame show functioned to frame narratives
for the audience each week. In most instances, the framing would then rever-
berate in subsequent commentary and production decisions during the game
of the week. For example, a pregame showing framing Black athletes as
animalistic set the stage for the game, wherein commentators described
Black players as “violent,” “instinctual,” and “galloping;” the audience was
in turn primed to associate athletes with animals.
Some scholars of framing have critiqued Agenda-­Setting for not incor-
porating how power mediates frames and their meaning (e.g., Carragee &
Roefs, 2004; Moy et al., 2016). Research on media framing, originating
from the field of sociology, attends more closely to the ideological implica-
tions of media frames and the sources of power from which the frames are
derived (Carragee & Roefs, 2004). In the previous example, the long his-
tory of dehumanizing Black people through animal comparisons is one
reason the framing is so powerful and enduring (see Haslerig et al., 2020,
for further discussion). Our study put analytical emphasis on examining
the sociological significance in relation to power, while relying on the psy-
chological underpinnings of Agenda-­Setting to communicate more gener-
alized implications of how Agenda-­Setting functions in sport. In the case
of College GameDay, producers and commentators choose which aspects
of a game or season to highlight and discuss, essentially “setting the
agenda” for public discourse of college athletics, as well as providing spe-
cific characteristics and attributes to associate with topics or subjects
through framing and priming.
Our methodology, based on interpretive content analysis, was informed by
a constructivist lens and allowed for analysis of both manifest and latent
content (Drisko & Maschi, 2016; Krippendorff, 2018). In other words, an
interpretative content analysis calls for not only an examination of what
information is presented and how, but also an examination of meanings
 Priming and Racialized Production in College Football Broadcasts 183

within the data that require more theoretical inferences (Drisko & Maschi,
2016). In our study, we accomplished this by analyzing the content in a way
that accounted for context and impact and sought to understand why, not
simply how, sport media sets the agenda of college sport discourse. By con-
necting the psychological roots of Agenda-­Setting with a sociological exam-
ination of how power was informing the agenda-­setting function of sport
media, we were able to draw more powerful conclusions explaining the influ-
ence sport media has on reinforcing racism and white supremacy.

Data Collection
We selected ESPN’s College GameDay as the data source due to its cultural
impact as the most-­watched college sport broadcast of all time (Volner,
2016). The show has become a cultural touchstone in the sport media land-
scape, with their multifaceted approach to covering the college football sea-
son. The show is composed of a three-­hour live pregame show airing on
Saturday morning and a selected “game of the week” that one of their pan-
elists calls. The pregame show is a spectacle in and of itself due to the mobile
production format in which the commentators, production crew, and their
iconic set relocate to a new location—generally a college campus that is
hosting the featured game—each week. They pick the location in the pro-
ceeding weeks, based on (a) which games are emerging as impactful match-
ups for playoff rankings or (b) the perceived storyline potential of the game,
such as a rivalry or recent sociopolitical event. Fans and college students
surround the set of the pregame show, holding handmade signs that often
reference football narratives, current events, and political commentary.
Throughout the 2016–2017 college football season, we recorded 17 live
game-­of-­the-­week broadcasts and 16 pregame shows that totaled 118 hours
of data to capture the full scope of the season and accompanying narratives.
We recorded broadcasts using a screen capture program, Apowersoft, and
saved each as mp4 files in a shared Dropbox folder as well as an external
hard drive.
Additional data we collected included downloaded rosters from each fea-
tured team and game statistics. Rosters were collected to cross-­reference
what we saw and heard in the broadcast. They were useful for confirming
player names and positions, and the accompanying pictures were additional
data in ascribing race. Although imperfect, racial ascription based on
­phenotype was an appropriate measure for this study, given the import of
phenotype in how commentators and fans likely assessed players’ race
(Feliciano, 2016; Maddox, 2004). Game statistics (i.e., box score) were col-
lected as an additional layer of data triangulation. For example, in our cod-
ing for athleticism and talent, there were clear racialized differences in how
184 Sara E. Grummert and Siduri J. Haslerig

players were seen and discussed. In one specific game, the head-­to-­head
matchup between a Black quarterback and a white quarterback shed light
on these differences. When the white quarterback threw an incomplete pass,
it was described as “setting [the team] up” for future plays; when the Black
quarterback had a pass completion that gained modest yardage, it was nega-
tively discussed, with the commentator only begrudgingly acknowledging,
“a reception, nonetheless.” Although those instances together seem to indi-
cate clear-­cut bias in the commentary, we used the game statistics to recon-
firm that the two quarterbacks had very similar game statistics, with
comparable passing yards and completion rates and the same number of
interceptions. Yet, the tenor of the commentary was worlds apart. Although
we did not use this data in our analysis for the 2020 article, it was an impor-
tant part of our larger data collection and helped reaffirm the validity of our
interpretations.

Layers of Data
Video data is multimodal and layered (see Table 11.1 for the layers of embed-
ded data). The existence of so many components nested within the videos
meant that directly analyzing the video was the best way to capture this
nuance. As such, we strongly felt that video data should be coded directly
rather than working with transcripts. In addition to considering each of
these layers of data independently, they are often happening simultaneously,
and it was therefore important to analyze them in relation to each other,
rather than each aspect in isolation. Lastly, the impact of video as a whole is
important and a source of data in itself; transcription, even a thoughtful and
comprehensive transcription process (Fields, 1988), may be unable to main-
tain/document the overall impact/effect of the video. Thus, video data can
enrich analysis and lead to qualitatively different findings than would be
possible if the data was transcribed and subsequently coded.
The layered complexity of this video data in turn impacted our approach
to data management, the digital tools we used, and the entirety of our cod-
ing and analysis process.

Data Management
Video files were imported into Dedoose and labeled with the week number
and whether the file was a pregame show or game broadcast (i.e., Week 1_pre-
game). We used Dedoose, a cloud-­based qualitative and mixed methods data
analysis application, to assist in data management, organization, and analysis.
In Dedoose, we could directly code video data and create discreet video clips
(referred to as excerpts in Dedoose) from the original source file. In other
 Priming and Racialized Production in College Football Broadcasts 185

TABLE 11.1 Components of Video Data

Audio Data
Commentary The verbatim transcript of what commentators said,
what would traditionally show up in transcript
Tone and intonation Exclamations, excitement, surprise, and “ooh and
ahs”
Discursive/conversation Interruptions, pauses, and cooperative overlap
norms
Music Music and scores used in preproduced segments.
These are often used to establish or reinforce a
particular tone or emotional response and can
prime politically or racially
Voiceover Preproduced script that is read and overlaid on
preproduced segments or transitions between
segments
Background noise Producer chatter, crowd noise or chants, and sounds
of play such as a collision
Visual Data
Camera scope and Camera focus during a live game broadcast, such as
movement when they zoom in vs. zoom out and what they
choose to spotlight
Replay Producers replaying certain clips from a live game
broadcast
Graphics Graphics and visuals created by production, generally
preproduced prior to the live broadcasts
Signs Signage in the crowd
Explicit use of spectacle Visual spectacles such as military displays, team
and ritual traditions, and other rituals
Convergence of Audio and Visual
Preproduced segment Preproduced segment aired during a game broadcast
or a pregame show. It combines multiple types of
data and thus a different analysis is necessary due
to the layered production choices (i.e., music,
graphics, voice-­overs, genre, and location)
Mismatches, dichotomies, Mismatch(es) or other relationships between the
and/or parallels various data sources within a given clip (e.g., visual
and commentary); both what co-­occurs and is
paired through sequencing
Overlay/priming Pairing images with specific commentary or
voice-­over
Format/genre The interplay between live, preproduced, and replay
formats; context of the set, the field, and the crowd
or campus; and wherever the preproduced segments
were filmed
186 Sara E. Grummert and Siduri J. Haslerig

words, Dedoose created a database of all video clips we coded that helped
facilitate more complex analysis that we discuss later in the chapter. Our three
other primary reasons for selecting Dedoose included the following:

1. The cloud-­based design enabled collaborative coding in real time. This

enabled us to independently code the exact same file and compare our
code applications and associated memos to reach a high level of agree-
ment. We illustrate this process in the next section. Additionally, we were
not bound to a specific computer and could access our data anywhere via
an internet connection.
2. Dedoose offers user-­friendly options for pulling all data related to one
code for further analysis. This feature was essential in allowing us to use
the constant comparative method in a seamless way. This efficiency
impacted both the ease of using the constant comparative method and
our ability to meaningfully do so with video data.
3. The dynamic data analysis charts and visualizations that link directly
back to the qualitative data (i.e., code co-­occurrence chart and code
application chart). These charts enabled us to pull and view snippets of
our coded video data without having to return to the full broadcast video.

We have used multiple versions of Dedoose over the seven years we have
worked on this project (including three versions leading up to the original
article’s publication). One of the primary benefits of using a web-­based (and
now cloud-­based) application like Dedoose, rather than software like many
other qualitative data analysis software (QDAS) programs, is the seamless-
ness with which feature updates and new versions are available without
added cost or effort from the user.

Direct-Coding Video Data

One of the most important data management decisions we made was to
direct-­code our video data. Paradoxically, a frequent response to us having
coded video directly has been “That’s so smart! You didn’t have to tran-
scribe!” as though direct-­coding video is simply a shortcut to the same ana-
lytic destination. In response to this, we note that although it resulted in a
more thorough analysis, it also complicated the analysis by maintaining the
layered complexity of the data and, consequently, demanding that we ana-
lyze the layers in relation to each other.
Moreover, by coding for mismatches between the audio commentary and
the accompanying visual(s), we were able to analyze what was happening more
accurately and with greater clarity. We found that the discordance between the
visual(s) and commentary served to normalize and underestimate Black
 Priming and Racialized Production in College Football Broadcasts 187

players’ injuries (and potential injuries) and we discussed the priming function
this served. This nuance would have been rendered invisible had we conducted
analysis using transcriptions instead of directly coding the video data.
Lastly, because we coded the video itself, the decisions that had to be
made during transcription, such as what level of analysis the researcher is
interested in (e.g., conversation analysis, topical analysis, and verbatim com-
mentary), did not have to be finalized before our coding process. This
enabled flexibility and allowed us to avoid predetermining what we would be
able to analyze (and, therefore, circumscribing what we could find) based on
how we transcribed. Because we had the verbal commentary/audio data in
the larger context of the video clip, we could determine what level of analy-
sis made the most sense for that specific instance, even where we later tran-
scribed specific excerpts for publication.
For example, when a Black quarterback suffered a concussion and was
lying immobile on the field for several minutes, the commentators were dis-
cussing his “durability” and “resilience.” Additionally, replays of the graphic
injury were repeated several times without mention from the commentators.
In a later analysis of this incident, we did transcribe with times noted, etc. In
that case, the length of time that he lay on the field unresponsive and docu-
menting the layers of data happening concurrently were central to our argu-
ment. For most of our analysis, however, timepoint data was extraneous and
unnecessary. Further, had we only analyzed a transcript, we would have been
unable to examine that the function commentary served in undermining the
player’s injury and instead reifying dehumanizing notions of Black resilience
and invincibility, and the analysis stemming from the replays would have been
completely absent. In the next section, we reflect on our analysis in more depth
and how Dedoose maximized our use of the constant comparative method.

Data Analysis
In this section, we focus on aspects of data analysis such as becoming famil-
iar with the data, coding development and consistency, and trustworthiness.
The process of data analysis was iterative, introspective, and interactive. As
such, though we must write about it linearly, it was anything but.

Familiarization
For the first three weeks of data collection, we both watched the broadcast in
real time and audio-­recorded memos and notes to ourselves while the data was
being recorded, which we stored locally on our respective devices. We met in
person after each of these games to discuss our thoughts about the game
including specific instances that interested us, unexpected topics that had come
188 Sara E. Grummert and Siduri J. Haslerig

up (whether they were explicitly raised in the commentary or came up for us as

viewers interpreting the broadcast), notable norms in the data, common lan-
guage and framing, and themes we identified. We used a whiteboard to track
our expanding focus as we discussed. These weeks were revealing in that (a) we
saw the dynamic layers of data embedded in each type of broadcast and (b) we
were interested in topics beyond the discourse about academics. This prompted
the project’s focus to grow exponentially to capture racialized intersections
with political events and social justice issues within the broadcasts.

Code Development
We used iterative coding, debriefing, and memoing processes associated with
grounded theory and the constant comparative method (Glaser, 1965).
Throughout this section, we detail the processes to collaboratively finalize
the codebook, code the data, and further subcode and refine our interpreta-
tions. The constant comparative method fit our study purpose because it
combines theory generation often associated with grounded theory and a
systematic categorization of the data using codes (Glaser, 1965).
Memoing and debriefing in early weeks of data collection helped us
establish our preliminary codebook and what descriptive information asso-
ciated with the video we were interested in, such as if the broadcast was a
pregame or game, the week number, and the location (Miles et al., 2018). We
initially stored memos in a shared Dropbox folder. As the project progressed,
our insights from the memos were frequently incorporated and collapsed in
the following two ways: Either (a) resulting in specific changes to our descrip-
tors, codes, or future memos; or (b) being uploaded to Dedoose as memos,
some of which were linked to specific excerpts or codes, thus serving as an
extension of analysis (we discuss this further in the following section).
Additionally, we knew early on that we needed to analyze production deci-
sions that were taking place rather than solely the commentary or visuals
presented, given the power of the agenda-­setting function taking place in
those decisions. In other words, we needed to consider not just what was said
or presented, but how it was being framed—both literally and figuratively.
Our preliminary code list incorporated initial observations from the data
(e.g., national anthem, activism, preproduced segment, and injury) and themes
from extant research literature (e.g., athleticism, intelligence, and animalistic
comparison). As mentioned earlier, we chose Dedoose, in part, due to the pow-
erful collaboration features. While establishing our codebook, we were able to
code the exact video file without viewing each other’s work, by using an in-­app
filter that excluded each other’s work from view. This allowed us to compare
our code applications and how much context we were including in each excerpt/
clip directly on the video. Figure 11.1 shows an example of video data in
Dedoose that has two coders’ excerpts, represented by green lines.
 Priming and Racialized Production in College Football Broadcasts 189

[SEE FIGURE 11.1 at eResource: Example of Video Data with Multiple Coders].

Further, we utilized the memo feature in Dedoose to link our memos directly to
coded excerpts to document analysis as well as log our questions about codes
and code applications. Memos were accessible via the linked excerpted clip as
well as in the memos workspace. Distinct memo categories helped us organize
and separate analytic memos from code application memos and general ques-
tions; we did so by creating memo groups titled “analytical” and “code develop-
ment.” Figure 11.2 is an example of a code development memo in which Sara
flagged a question for Siduri about when to apply a code related to athletic
ability.

[SEE FIGURE 11.2 at eResource: Example of “Code Development” Memo].

Lastly, we kept a coding decision log (see Table 11.2) to document our think-
ing and used it to build our final codebook with accompanying definitions
(see Table 11.3 for the final codebook).

[SEE TABLE 11.3 at eResource: Final Codebook].

Our decision log differed from our codebook in that it was created in real
time, specifically to ensure that we knew when the change was made, why,
and how it was implemented. In contrast, our codebook reflected the final or
most current understanding of the codes but did not contain as much pro-
cess data. For example, the final coding may end up subcoding “throwing
ability and talent” separately, in which case that would show up in the code-
book, but a decision log let us know where “throwing ability” was collapsed
into during the initial coding process—making it seamless to subsequently
subcode from that initial broad code and indicating from where we should
start subcoding. Similarly, yet more consequentially, the decision log

TABLE 11.2 Excerpt from Coding Decision Log

Date Person Coding Decision

5/16/2017 SG Coding throwing ability as “talent”

5/16/2017 SG Using the combination of “body” and “athleticism”
to code for strength
5/16/2017 SG Using the combination of “body” and “athleticism”
to code for instinct. Though it is related to an animal
comparison, I’m not sure how to handle it yet
5/19/2017 SG Removed weighting from “adversity”
190 Sara E. Grummert and Siduri J. Haslerig

documented how we intentionally used a code co-­occurrence tool in Dedoose

to capture certain complex concepts rather than creating a separate code.
This process was in keeping with our belief that too many codes are detri-
mental to the coding process being comprehensive (it is difficult to consider
too many codes at the same time), the complexity of an analysis (if data is
subdivided—i.e., coded—into too many categories, it can be difficult for
researchers to assess how concepts and categories are interrelated, obscuring
the larger picture), and, ultimately, one’s understanding of the data.
The decision log also documented indecision at times—we noticed the
relationship between animal comparison and instinct, for example, but
needed to continue analyzing more of the data to decide how to handle it.
So, we parked the decision in the log to be made later. With multiple coders,
it also ensured that decisions were documented as intentional choices, so
everyone was updated on the analytical choices. In this way, it helps serve as
a flag for when further team discussion might be needed.

Trustworthiness and Coder Agreement

After we independently coded Week 1 data, we held an in-­person meeting to
assess and discuss our interpretations, note any differences, clarify code defi-
nitions, and select what codes needed to be included on one finalized excerpt.
This process was made more meaningful and efficient via the memos work-
space in Dedoose because we could filter by memo category (e.g., code devel-
opment vs. analytical), choose the memo we wanted to discuss, and hover
over the “memo links” icon to view the video clip and memo content simul-
taneously. We repeated this process for Week 2 and Week 3 data, after which
we concluded we had high agreement between our respective coding and had
collaboratively reached consensus about how to handle potentially ambigu-
ous or contentious coding decisions. This enabled us to code data indepen-
dently with confidence.
Although Dedoose and other QDAS programs offer interrater reliability
tests, we used the previous procedure rather than the built-­in function for
several reasons. Fundamentally, these reasons reflect our approach to cod-
ing as qualitative researchers and collaborators, rather than as a rote or
workman-­like process that might be meaningfully assessed quantitatively.
Fundamentally, the term “interrater reliability” is reflective of the quantita-
tive underpinnings of the concept and its origin in developing assessment or
survey items that are reliable across different raters, rather than in qualitative
coding. An interrater reliability test via Cohen’s kappa is inherently limiting
because it is a quantitative statistic attempting to measure a qualitative pro-
cess. Many QDAS programs offer this statistical test, yet by extracting cod-
ing from its original dataset, these tests remove the data from the context in
 Priming and Racialized Production in College Football Broadcasts 191

which it was coded. Further, because the tests are often run with only a
limited selection of codes, this process obscures meaningful discussions
related to code co-­occurrence and how much context should be included in
excerpts. Many of our most generative conversations during the process of
establishing reliability were about code co-­occurrences: Whether those co-­
occurrences accurately captured all relevant aspects of the excerpt or if an
additional, more specified, code was needed. In the absence of the full code-
book, the benefit of undertaking a collaborative, qualitative approach to
intercoder agreement is potentially missed.
Second, the test demands precise coding agreement in terms of the size of
the excerpt coded. From experience, instances where the same code was
selected but the excerpt length differed led to thoughtful, productive conver-
sations about how excerpt length should be decided and how that decision
would in turn impact our analysis and approach consensus about how we
would proceed in determining excerpt size. Third, our process did far more
than establish interrater reliability or ensure coding agreement and consis-
tency. During the discussion process, we also assessed whether there were
additional codes needed, how to nuance our code definitions for clarity
(which, in turn, supported consistent application), and what the substantive
consequences of these coding decisions might be and reviewed linked
memos. More generally, having these deep discussions surrounding our cod-
ing fidelity encouraged us to understand each other’s thinking, imagine
implications of coding decisions, and ultimately reach well-­reasoned con-
sensus in finalizing our codebook and operationalizing the coding process.

Coding
Once the codebook was finalized, Sara completed most of the coding for the
video dataset.
We used the constant comparative method to accomplish this analysis
and hone our codes and conceptual categories (Glaser, 1965). We went clip-­
by-­clip comparing each coded clip with other instances of that code and
subcoding for differences and/or nuances. The constant comparative method
bridges the gap between deductive and inductive approaches to data analy-
sis. As outlined by Glaser (1965), the method calls for comparing similar
occurrences within the data, in ever-­more specific and similar groupings, to
solidify categories and analyze them for meaning. Digital tools for qualita-
tive analysis enable the use of the constant comparative method to an extent
and level of granularity that was unimaginable when Glaser wrote about the
method in 1965. As we mentioned earlier in our reasons for choosing
Dedoose, researchers can subcode within a specific set of coded excerpts
rather than returning to the original piece of data as a whole (exemplified in
192 Sara E. Grummert and Siduri J. Haslerig

the discussion of the “injury” code below). This bounds the researcher to the
specific concept of interest in a beneficial—rather than limiting—way.
For video data, this creation of discrete clips from the coded “excerpts” (in
the language of Dedoose) is invaluable. Video data is incredibly time consuming
to analyze and often difficult to navigate; analysis methods frequently require
noting the start time within the data to enable the researcher to return to it
through a cumbersome process for going back to those flagged times within the
larger video. In contrast, the creation of discrete excerpt clips by Dedoose means
that the researcher can seamlessly watch each clip in succession for a given code
without any additional navigation of the data. This enables comparison within
a category that Glaser (1965) describes to happen in real time and without fric-
tion, even for video data. Throughout coding, we continued to hold in-­person
meetings to review each week of data and discuss notable or interesting events,
discuss concepts and larger narrative arcs that were taking place across pregame
shows and game broadcasts, and raise any challenges or questions.

Moving from Analysis to Findings

For the purposes of this paper, we present examples of how we coded data
associated with injury. As we reviewed the video data, it became clear that
the analysis needed much more nuance than simply “which players are
hurt?” or even “how are injured or potentially injured players discussed (by
commentators) and/or treated (by team staff, teammates, and officials)?”
The first nuance we had to account for was that we needed to analyze
instances that were not confirmed as injuries, not only because of the discus-
sions that unfolded during the coverage of a potential injury and but also
due to our argument—supported by the research literature—that injuries
were likely underdiagnosed and those biases may also be racialized (Goff
et al., 2014; Goyal et al., 2015). After our first round of coding for “injury”
and the athlete’s race, we pulled every clip coded with “injury” to carry out
further analysis, comparison, and subcoding. We subcoded for injury type
including “head/neck/spine,” “ankle/leg,” “concussion,” and “replay” to
analyze key distinctions more accurately such as

1. The significance—or potential significance—of the injury.

2. How it was framed and discussed by commentators.
3. If replay was used to present the injury (either the cause of it or its after-
math) over and over again to the viewer.

The coded data were then analyzed separately and compared between Black
and white players to systematically identify differences. We used Dedoose to
disaggregate the coded data by race, prior to further analyzing it, by pulling
excerpts for a given code related to injury for white players and analyzing
 Priming and Racialized Production in College Football Broadcasts 193

them collectively, separate from the data on athletes of color. Both the code
co-­occurrence chart and the filtering tool within the Excerpts workspace
were helpful to isolate domains of interest at this point in our analysis. We
then repeated this procedure for athletes of color and further drilled down
to look at Black athletes specifically. Later in our analysis, we added sub-
codes to account for how an injury or pain was handled in terms of care and
caution within the game (e.g., “taken out,” “stay in,” and “return to play”).
While we were subcoding, we created an Excel table that logged each
instance, which we then used to calculate the “injury-­causing impacts” that
we discuss in more depth in the next section. There were 295 excerpts/clips
that had been coded with “injury” (see Figure 11.3).

[SEE FIGURE 11.3 at eResource: All Excerpts Coded for “Injury”].

Dedoose helped us to operationalize the constant comparative method by

facilitating cycling through each clip and comparing them directly, while simul-
taneously subcoding for differences, similarities, or nuances between clips
within the category (see Figure 11.3 displaying a coded excerpt for “injury”).
Figure 11.4 illustrates the process of pulling excerpts coded as “injury” and
further subcoding with more nuance, such as “affect,” “pain,” and “body.”

[SEE FIGURE 11.4 at eResource: Comparing and Subcoding an Excerpt Coded for
“Injury”].

Injury-Causing Impacts
The data on injury was complex in that there were confirmed injuries that
occurred and impacts that could possibly (and often presumably) cause an
injury. As we explained in the original article (Haslerig et al., 2020, p. 280):

Within the injury data, we make the distinction between injury and
impact: “injury” encompasses possible/likely injuries as well as those that
were confirmed, whereas “injury-­causing impacts” (ICIs) correspond to
the actual play that may have led to injury. The distinction is important
because CGD (college Game Day) aired violent impacts and visuals of
players in pain or ostensibly injured, who were cleared for play with no
diagnosed injury.

It was imperative that we capture this distinction in our analysis, particularly

given the extant research on the harmful impact of cumulative impacts on
the brain and the tendency to underestimate and/or deny Black pain (Goff
et al., 2014; Goyal et al., 2015). Using Dedoose’s Code Co-­occurrence Chart
(see Table 11.4), we could visualize and disaggregate the injury categories by
194 Sara E. Grummert and Siduri J. Haslerig

race. The chart visualized and quantified how many times two codes were
applied to the same excerpt or overlapping excerpts. Essentially, the chart
provided a heat map that illustrated larger themes or relationships in the
data. For example, Table 11.4 shows that the code “athlete of color”
co-­occurred with “head/neck/spine injury” 29 times compared to 1 co-­
occurrence of “head/neck/spine injury” and “white athletes.”

[SEE TABLE 11.4 at eResource: Code Co-­Occurrences Chart].

We had to wrestle with several issues to isolate the information in a way that
was intellectually honest, given the disparity between athletes of color and
white athletes represented in the injury data. Black players may have been
injured more than white players in our data for a variety of reasons such as the
overrepresentation of Black players on Division I football teams (Lapchick et
al., 2021), the playing time of each player, and positional stacking that places
Black players in positions where they are more likely to get injured (Woodward,
2004). To display these findings in a way that was both compelling and accu-
rate, we exported the data from the Code Co-­occurrence Chart into an Excel
spreadsheet and created a table to illustrate and quantify “injury-­causing
impacts” using intra-­racial ratios rather than pure code counts in order to be
accurate about our findings and claims (see Table 11.5 for the Excel analysis).

[SEE TABLE 11.5 at eResource: Injury Data Being Further Analyzed in Microsoft
Excel].

Our disaggregated analyses made clear that the framing of injury between
Black and white players was drastically different in terms of the severity,
frequency, and, in some cases, existence of injuries. The granularity of this
analysis helped tell the story of patterns within our data that consistently
minimized the seriousness of injuries of Black players (thereby undermining
empathy and implying an invincibility), while also amplifying the spectacle
of their injury-­causing impacts and further desensitizing the viewer to Black
pain and risk. We described the analytical significance through the use of an
evocative metaphor: that of a half-­life (Haslerig et al., 2020). In other words,
each injury of a white player was treated as relevant for longer, as measured
by multiple mentions and especially mention of healed past injuries—despite
tending to be less dangerous—than those injuries suffered by Black players.
The time it took for commentators to stop treating an injury as relevant was
variable by race and comparable to the varying speed at which different iso-
topes decay—commonly measured by their half-­life.
Another reason we used Microsoft Excel was to better track the multiple
variables of interest within “injury” and more easily get counts for specific
occurrences, such as the number of head or spine injuries/potential injuries, by
 Priming and Racialized Production in College Football Broadcasts 195

race, or the discrete number of injuries discussed during the season, rather than
the count of how many times injuries were discussed—for which we could have
pulled the number directly from Dedoose. The Microsoft Excel file was useful
because many injuries were discussed multiple times, resulting in multiple coded
excerpts that were irrelevant to the total number of injuries. This Microsoft
Excel “coding” (or, more accurately, reframing of how we visualized the coded
data) still relied on Dedoose, in that we revisited the coded excerpts to drill down
even more deeply and to enable us to ensure our final counts were accurate.
The coding processes, as well as the subsequent table, exemplify the pur-
pose of the constant comparative method—to combine systematic coding
processes and theory generation (Glaser, 1965). Presenting code occurrences
and percentages in a quantifiable table and contextualizing them with com-
mentary and other data from the broadcasts enabled us to illustrate how and
in what ways white supremacy and antiblackness were embedded within
injury discourse and coverage.

Conclusion
Our study has both theoretical and methodological implications, particu-
larly for those conducting content analyses and/or those analyzing multi-­
media data. In particular, it underscores the need for a more comprehensive
approach toward analyzing video data. We caution against treating tran-
scripts as equivalent to multi-­ media data, particularly video. The data
derived from a video are much more expansive than usually conceived and
contain multitudes of different options for analysis (see Table 11.1) and
should prompt researchers to consider how analyzing video data instead of
transcriptions could strengthen their analysis. Decisions about what type of
data and analyses to employ should align with the epistemological and theo-
retical frameworks underpinning the study. For example, our study’s use of
the constant comparative method and accounting for different layers and
genres of data within the video was in direct alignment with the goals of
interpretive content analysis and Agenda-­Setting Theory. The analysis of
video data opens new avenues to study the agenda-­setting function of mass
media and the dynamic, perhaps taken-­for-­granted, forms in which it occurs.
Video data may have been transcribed and analyzed as transcripts in the
past, in part, due to a lack of adequate tools to do more complex and nuanced
analyses. Subsequent researchers have continued to transcribe video data
(albeit in ever-­more comprehensive and sophisticated ways) because it is a
research norm, and the painstaking transcription process is often perceived
as a measure of rigor. The difficulty of a research task does not, however,
equate to its validity or utility. Keeping video data in its original form does
“skip” a stultifying step in traditional data management procedures. However,
196 Sara E. Grummert and Siduri J. Haslerig

more importantly, it ensures that the rich data is fully accessible for analysis.
Fields (1988) explained, “The qualitative analysis of a news videotape begins
with choosing analytical units that meet one’s theoretical goals but which do
not do violence to the nature of the content at hand” (p. 183). In this case, the
extreme difficulty of meaningfully transcribing video data is akin to a symp-
tom of the ways transcription inevitably circumscribes and deadens—i.e.,
“does violence to”—parts of video data, rather than an indication of an
inherently rigorous research process. The increased functionality of qualita-
tive data analysis tools over the last decade(s) enables much more expansive
ways to analyze video data and other multi-­media. Lastly, the ease of opera-
tionalizing the constant comparative method with digital qualitative data
analysis tools amplifies the effectiveness and practicality of the method.
When using Dedoose, the constant comparative method becomes more tan-
gible in terms of comparing within a category (which Glaser refers to as “the
defining rule for the constant comparative method,” 1965, p. 439) because the
application creates distinct clips or excerpts that allow the researcher to com-
pare directly and continue deeper analysis.

KEY WORKS GUIDING OUR ANALYSIS

Glaser, B. G. (1965). The constant comparative method of qualitative analysis. Social

Problems, 12(4), 436–445. [Link]
Glaser’s foundational article asserted the need to avoid framing grounded the-
ory and a systematic coding of data as incompatible. He suggested the constant
comparative method as an iterative approach to blend traditional coding tech-
niques and grounded theory. Using a qualitative data analysis application (Dedoose)
brought this method to life in our data by literally allowing us to compare “like”
excerpts with other “like” excerpts to examine differences and similarities.
Drisko, J. W., & Maschi, T. (2016). Content analysis. Oxford University Press.
Traditional content analysis is often framed as examining and quantifying what
is directly presented to the audience and/or reader. In this book, the authors call for
the use of interpretative content analysis that goes beyond an analysis of explicit
content and more toward latent content that considers context and implications.
This complemented the theory we used (Agenda-­Setting) and provided a much
more powerful analysis in terms of sociological implications of our findings that
otherwise may have gone unexamined or unexplained.

References
Bhattacharya, K. (2017). Fundamentals of qualitative research: A practical guide.
Routledge.
Carragee, K. M., & Roefs, W. (2004). The neglect of power in recent framing research.
Journal of Communication, 54(2), 214–233. [Link]
j.1460-­2466.2004.tb02625.x
Drisko, J. W., & Maschi, T. (2016). Content analysis. Oxford University Press.
 Priming and Racialized Production in College Football Broadcasts 197

Feliciano, C. (2016). Shades of race: How phenotype and observer characteristics

shape racial classification. American Behavioral Scientist, 60(4), 390–419. https://
[Link]/10.1177/0002764215613401
Fields, E. E. (1988). Qualitative content analysis of television news: Systematic tech-
niques. Qualitative Sociology, 11(3), 183–193. [Link]
Glaser, B. G. (1965). The constant comparative method of qualitative analysis. Social
Problems, 12(4), 436–445. [Link]
Goff, P. A., Jackson, M. C., Di Leone, B. A. L., Culotta, C. M., & DiTomasso, N. A.
(2014). The essence of innocence: Consequences of dehumanizing Black chil-
dren. Journal of Personality and Social Psychology, 106(4), 526–545. [Link]
org/10.1037/a0035663
Goyal, M. K., Kuppermann, N., Cleary, S. D., Teach, S. J., & Chamberlain, J. M.
(2015). Racial disparities in pain management of children with appendicitis in
emergency departments. JAMA Pediatrics, 169(11), 996–1002. [Link]
10.1001/jamapediatrics.2015.1915
Haslerig, S. (2013). Super seniors: The educational trajectories and experiences of
graduate(d) student athletes in Division I football. [Unpublished doctoral disserta-
tion]. University of California, Los Angeles. [Link]
7b62g40n
Haslerig, S. (2017). Who or what are graduate(d) student-­athletes? Defining a misun-
derstood sub-­population. Journal of Higher Education Athletics & Innovation,
1(2), 110–122. [Link]
Haslerig, S. J., Grummert, S. E., & Vue, R. (2019). Rationalizing Black death: Sport
media’s dehumanizing coverage of Black college football players. In P. Brug, Z.
Ritter, & K. Roth (Eds.), Marginality in the urban center: Neighborhoods, commu-
nities, and urban marginality (pp. 77–109). Palgrave Macmillan. [Link]
10.1007/978-­3-­319-­96466-­9_5
Haslerig, S. J., Vue, R., & Grummert, S. E. (2020). Invincible bodies: American sport
media’s racialization of Black and white college football players. International
Review for the Sociology of Sport, 55(3), 272–290. [Link]
1012690218809317
Krippendorff, K. (2018). Content analysis: An introduction to its methodology. Sage.
Lapchick, R., Bowman, D., Ewing, S., Forbes, A.J., Green, A., Jackson, B., Johnson-­
Schmeltzer, B., Middleton, T., & Richardson, K. (2021). The 2021 Sports Media
Racial and Gender Report Card. The Institute of Diversity and Ethics in Sport.
Retrieved December 4, 2021, from [Link]
Maddox, K. B. (2004). Perspectives on racial phenotypicality bias. Personality and Social
Psychology Review, 8(4), 383–401. [Link]
McCombs, M., Llamas, J. P., Lopez-­Escobar, E., & Rey, F. (1997). Candidate images in
Spanish elections: Second-­ level agenda-­ setting effects. Journalism & Mass
CommunicationQuarterly,73(4),703–[Link]://[Link]/10.1177/107769909707400404
McCombs, M. E., & Shaw, D. L. (1972). The agenda-­setting function of mass media.
Public Opinion Quarterly, 36(2), 176–187. [Link]
McCombs, M. E., Shaw, D. L., & Weaver, D. H. (2014). New directions in agenda-­
setting theory and research. Mass Communication and Society, 17(6), 781–802.
[Link]
Miles, M. B., Huberman, A. M., & Saldaña, J. (2018). Qualitative data analysis: A
methods sourcebook. Sage.
198 Sara E. Grummert and Siduri J. Haslerig

Moy, P., Tewksbury, D., & Rinke, E. M. (2016). Agenda-­setting, priming, and fram-
ing. In K. B. Jenson & R. T. Craig (Eds.), The international encyclopedia of com-
munication theory and philosophy (pp. 1–13). John Wiley & Sons. [Link]
10.1002/9781118766804.wbiect266
Piaget, J., & Inhelder, B. (1967). The child’s conception of space. W. W. Norton & Co.
Volner, D. (2016, December 6). More than 179 million fans watched 100 billion minutes
of college football games on ESPN’s TV networks during the 2016 college football
season; 15 million unique devices streamed ESPN games. ESPN MediaZone.
Retrieved October 23, 2018, from [Link]
2016/12/179-­million-­fans-­watched-­100-­billion-­minutes-­college-­football-­games-­
espns-­t v-­n etworks-­2 016-­c ollege-­football-­s eason-­1 5-­m illion-­u nique-­d evices-­
streamed-­espn-­games/ (Not accessible as of [28-06-2024])
Woodward, J. R. (2004). Professional football scouts: An investigation of racial
stacking. Sociology of Sport Journal, 21(4), 356–375. [Link]
ssj.21.4.356
12
REFLECTIONS ON CONDUCTING
TEAM-BASED QUALITATIVELY
ORIENTED MIXED METHODS
RESEARCH ABOUT STUDENTS WITH
DISABILITIES IN STEM CLUBS
Peggy Shannon-­Baker, Karin Fisher, and Kania Greer

Qualitatively oriented mixed methods research uses quantitative research

procedures to aid in a study that otherwise prioritizes qualitative research
practices and related epistemologies (Cameron, 2016; Hesse-­Biber, 2010,
2018; Poth & Shannon-­Baker, 2022; Toledo & Shannon-­Baker, in press).
Researchers use this type of mixed methods for many reasons, including to
center the perspectives and experiences of marginalized populations in the
research process and research products, emphasize more inductive data
analysis practices, and use researchers’ qualitative research skills (Poth &
Shannon-­Baker, 2022). Qualitatively oriented mixed methods also provide
researchers with more flexibility to make changes to the research design
based on dynamic contextual changes in contrast to the relatively linear way
that mixed methods research is often described (Hesse-­Biber, 2018; Poth &
Shannon-­Baker, 2022).
Mixed methods research is increasingly conducted in teams because such
collaborations allow individual researchers to bring their specific method-
ological expertise to examine the complexities of the given study (Aarons
et al., 2019; Bazeley, 2018; Hemmings et al., 2013; Hesse-­Biber & Johnson,
2013). In these instances, research teams may include individuals with more
skills and experience in a single methodology and research applications
within a given field or discipline (Hemmings et al., 2013; Jordan & Bartel,
2023). Additionally, having multiple methodological and disciplinary per-
spectives can aid a mixed methods research team in making their work more
culturally responsive, accessible to diverse audiences, and cross-­disciplinary
(Hesse-­ Biber & Johnson, 2013; Jordan & Bartel, 2023; Plano Clark &
Ivankova, 2016; Shannon-­ Baker & Martinez, 2022). However, mixed

DOI: 10.4324/9781003254041-13
200 Peggy Shannon-Baker et al.

methods research teams face unique challenges (Jordan & Bartel, 2023),
such as opposing views about the value and rigor of other research practices.
Mixed methods research teams need to engage in ongoing reflection as indi-
viduals and as a group about the research process, plan the research study in
dialogue, and work intentionally to negotiate consensus and build collabora-
tion (Aarons et al., 2019; Hemmings et al., 2013; Jordan & Bartel, 2023).
Additionally, examples of mixed methods studies that prioritize the qualita-
tive components are challenging to find (cf. Poth & Shannon-­Baker, 2022).
Few examples detail how to analyze qualitative data and meaningfully com-
bine those data with quantitative data in a team-­based study. “Such an
example,” such as the one presented here may help increase the use of quali-
tatively oriented mixed methods, better prepare researcher teams to work
effectively together, and improve the rigorous application of such studies.
Thus, the purpose of this chapter is to describe the collaborative and ana-
lytical practices in our qualitatively oriented mixed methods investigations
into the experiences of students with disabilities (SWD) in science, technol-
ogy, engineering, and math (STEM) clubs (Fisher et al., 2021, 2022). Study
1 (Fisher et al., 2021) investigated the features of STEM clubs in rural-­based
schools and the credentials of teachers who lead STEM clubs in working
with SWD in informal learning environments. We found that the STEM
clubs were offered with an average of 20 students in each and that most
teachers did not know how many SWD were in their clubs. Moreover, none
of the teachers had professional development to work with SWD in infor-
mal environments. Study 2 (Fisher et al., 2022) explored students’ and their
parents’ experiences related to STEM clubs. This study found that students
joined and stayed in clubs for multiple reasons (e.g., liking the subject mat-
ter, because parents and teachers encouraged them). The barriers students
and parents shared included teachers or clubs not making accommodations
for SWD and parents feeling overwhelmed trying to ensure their children
could fully participate (e.g., transportation and advocating for accommoda-
tions). These studies were essential to understanding the school-­based and
interpersonal contexts for the participation of SWD in STEM clubs.
Moreover, studying this from the perspectives of students, parents, and
teachers across the two studies allowed us to investigate current practices,
gaps in knowledge and accommodation practices, and barriers that inhibit
SWD from participating fully in STEM clubs. By acknowledging potential
barriers, schools can begin to provide services that increase access to SWD
and provide teachers with the professional development they need to imple-
ment inclusive practices in STEM clubs.
This chapter discusses our team-­based analytical practices across these
two mixed methods studies. In the following section, we describe our posi-
tionalities and methodological expertise that we, as authors, brought to this
 Qualitatively Oriented Mixed Methods, Students with Disabilities 201

line of inquiry. Next, we provide further details about our two studies. With
both studies, we focus on our team-­based qualitative analyses and provide
our current reflections on those practices. Our discussion of Study 2 also
describes how we created visualizations that compared the qualitative and
quantitative findings and the challenges of publishing this article. We con-
clude with our reflections on the process and recommendations for mixed
methods teams that want to emphasize their qualitative practices.

Our Positionalities as Collaborative Researchers

Throughout the two studies, the research team consisted of three to four
people. Each person brought a unique perspective to the study and had a
background that enabled them to view the information differently. The
inquiry began in 2016 when Karin and Kania collaborated on a curriculum
review for after-­school robotics teams. The beginning ideas of the first study
materialized through discussions about curriculum and after-­school clubs.
Understanding the need for more input from others to participate in the
initial study, Karin invited Peggy to join them in their discussions.
Karin is an associate professor of special education at Georgia Southern
University and the parent of a son with autism. As a former high school
special education teacher and robotics coach, Karin studied the impact of
extracurricular STEM activities on the science performance of SWD for her
doctoral dissertation (Fisher, 2016). She used an exploratory single-­survey
quantitative research methodology to determine the correlation of test
scores with survey results. She teaches a variety of special education courses
to undergraduate and graduate students. Karin was the content expert and
primary investigator of the project.
Kania coordinates experiences for SWD in informal learning environ-
ments. She is the Coordinator for the Center for STEM Education within
the College of Education (COE) at Georgia Southern University. Kania has
a background in working with SWD in the transition out of high school and
studies the intersection of formal and informal education in developing sci-
ence identity. She has conducted numerous mixed methods studies on how
people understand and use science. She approached our data analysis with a
lens considering context, human nature, and language.
Peggy is an associate professor of educational research. They teach grad-
uate courses in mixed methods, qualitative, and action research, among
other research methods courses. They joined the studies as a research meth-
odologist trained primarily in qualitative, arts-­based, and mixed methods
research practices. Their role entailed translating the study goals into feasi-
ble research studies, grounding the methods in the methodological literature,
and leading the mixed methods analyses.
202 Peggy Shannon-Baker et al.

Study 1: Characteristics of Rural-Based STEM Clubs and

Teachers’ Training
The purpose of Study 1 was to identify the types and characteristics of
STEM clubs in two rural counties and teachers’ professional development to
work with SWD in informal learning environments. We used a sequential
explanatory mixed methods design with a survey and interviews (Creswell &
Plano Clark, 2017). Sequential explanatory designs begin with an initial
quantitative phase followed by a qualitative phase. Typically, these phases
are connected by using quantitative results to select participants or direct the
qualitative data collection (e.g., by asking participants about surprising
quantitative findings) (Creswell & Plano Clark, 2017), both of which apply
to this example study.

Initial Quantitative and Qualitative Data Collection

We distributed a survey using Qualtrics to STEM club advisors/coaches in
two rural counties. Rurality was defined based on population size (cf. Fisher
et al., 2021). A total of 17 participants responded to the survey. The survey
consisted of 19 questions, including two open-­ended questions asking par-
ticipants to describe the professional development they received for working
with SWD in an extracurricular environment and why teachers think SWD
do not participate in extracurricular activities as much as their peers. For a
full copy of the survey, see Fisher et al. (2021). Only two respondents
answered the first open-­ended question. For the second question, eight
responses varied in word length from one word (“transportation”) to com-
plete sentences such as “I think the one student that was in my club did not
show up but maybe once because he did not make it on time.” We did not
position these open-­ended questions as “qualitative” data because they
asked simple questions, lacked direct participant engagement, and did not
generally encourage richness or complexity, which are characteristics of
qualitative research (Hatch, 2002). Given the small number of responses to
the first open-­ended survey question, we simply reported them in our first
article (Fisher et al., 2021). Because we were focused on the student experi-
ence at the time of publication, we did not report the teacher responses to
the second open-­ended question in the survey.
The second phase of our study explored teachers’ training and their per-
ception of students’ interest in extracurricular activities via interviews with
the teachers/coaches who responded to the initial survey. Because the primary
purpose of the quantitative survey was to help us select teachers for qualita-
tive interviews, we placed more emphasis on identifying teachers across
unique settings, individual demographics, and STEM club types. Eleven
teachers agreed to participate in follow-­up interviews (Fisher et al., 2021).
 Qualitatively Oriented Mixed Methods, Students with Disabilities 203

The researchers went in pairs to each participant’s school to conduct semi-­

structured interviews (Fisher et al., 2021). For a full copy of the interview pro-
tocol, see Fisher et al. (2021). In the teams, one person focused on asking
questions and talking with the teacher, while the other focused on taking field
notes about the conversation (Yin, 2016). For example, Kania concentrated on
the interview conversation based on her experience in informal STEM environ-
ments, whereas Peggy focused on the field notes to capture important points
from the conversation (Document 12.1). Having a second person taking notes
during the interviews helped address any potential challenges the research pair
might have faced, such as fast-­talking participants, participants who opted not
to be audiorecorded, and technical glitches with recorders. The field notes
served as supplementary data to support the interview transcript data (cf. King
et al., 2019), though we analyzed both the transcript and field note data during
our analysis phase. In total, we collected 144 pages of interview transcripts and
36 pages of field notes. All transcripts were stored as hard copies in a locked file
box in the lead researcher’s office. All code lists, transcripts, and meeting notes
for this first study were stored together in the lock box.

[SEE DOCUMENT 12.1 at eResource: Excerpt of Handwritten Field Notes from

Study 1 Interview].

Collaboratively Analyzing the Interviews and Field Notes through

Shared Meaning Making
After we transcribed the interviews, the research team had a day-­long meet-
ing to discuss how to code the transcripts and field notes. To prepare for the
meeting, all researchers read hard copies of the interview transcripts inde-
pendently to refamiliarize themselves with the data. Some researchers made
notes in the margins to ask questions about the data or indicate possible
descriptive codes (see Document 12.2, which displays some example margin
notes).
Then, during the meeting, we opened the discussion by asking questions
like “What did you see in the transcripts? What did the teachers say?” Based
on our discussions, we identified several common topics that we then used as
codes, such as awareness of other clubs, comfort with recruiting SWD, work-
ing with SWD in general, and understanding of modifications for SWD in
informal learning environments. By collaboratively identifying codes as a
team, our goal was to identify a list of codes we could each then look for in
the data. However, we were concerned about having too many codes to look
for, especially because some of the team members were newer to qualitative
coding. We then grouped our codes into four broad, color-­coded categories:
(a) recruitment of SWD in STEM clubs (blue), (b) their retention and
204 Peggy Shannon-Baker et al.

participation in STEM clubs (green), (c) professional development for teach-

ers to work with SWD in informal learning settings (orange), and (d) teach-
ers’ knowledge of club participants and requirements for SWD in clubs
(pink). Document 12.2 also shows the application of these color schemes.

[SEE DOCUMENT 12.2 at eResource: Excerpt of an Interview from Study 1 with

Margin Notes and Color-­Coded Categories].

Next, each team member coded a hard copy of the data individually during the
meeting. Coding the data like this allowed us to discuss example quotes and
overlapping codes as a group. Each coder brought their unique perspectives to
the coding process, which we saw as an asset to the analysis process. For exam-
ple, former k-­12 teachers or team members with professional development or
STEM club-­related experiences could potentially identify more data related to
a category than others without those experiences. During and after coding, we
talked as a team. These discussions allowed each team member to explain their
coding and thought process and gave us space to make decisions as a group. For
example, we decided that keeping documentation of data coded with multiple
categories was important, rather than identifying a single code for a piece of
data, because we saw a conceptual overlap between the categories. Any data not
coded within the allotted time was coded independently after the meeting.
Because each person’s expertise brought a unique lens to the coding, the process
was very collaborative. Each person was given the opportunity to make a case
for how they coded data in a particular way. Instead of trying to identify one
way to code a piece of data, any piece of data could have multiple codes to
capture the whole team’s understanding of how to code it.
The research team then met to refine the categories into themes under
Kania’s leadership. Kania collated our coded data into areas centralized around
themes that most related to our study’s original purpose: (a) Area 1: Awareness
of other clubs, teachers’ comfort with recruitment, and their professional devel-
opment; (b) Area 2: Barriers to SWDs’ participation and teachers’ awareness
of modifications and accommodations; and (c) Area 3: Ideas for recruitment.
For example, Table 12.1 displays how we reviewed example data under the
third area. We organized the data into a table to show example quotes each
researcher coded for a particular area theme. Seeing everyone’s example-­coded
data allowed us to talk about the context for each quote because some coded
data pieces were longer than the others. It also helped further dialogue about
our interpretations of the coded data and its relevance for a given theme.

[SEE TABLE 12.1 at eResource: Excerpt of Data Compiled Under the Theme of
“Recruitment” for Study 1].
 Qualitatively Oriented Mixed Methods, Students with Disabilities 205

When considering meeting space, we examined the environment we wanted

to create. We considered meeting online, in person in a classroom, in some-
one’s office, or in a conference room. We wanted to create an environment
that did not promote a hierarchy between researchers and that allowed dis-
cussions to happen naturally. We also needed space to allow all researchers
to bring their laptops, spread out hard copies of transcripts, have room to
work, and be comfortable for an entire day. We decided to meet in a small
conference room, allowing for an intimate and collaborative environment.
The conference room that we used is a work room that has two four-­person
tables aligned in a “T” shape. This setup allowed us to see everyone’s face,
ensured everyone was on the same task, and provided an intimate environ-
ment that promoted whole group discussions.
Reflecting on this process of coding as a team, Peggy said, “coding
together in one room allowed us to immediately talk about a quote or code
as it came up. This meant we could come to consensus or help one another
in the coding process.” Kania elaborated:

The whole process was enjoyable and collegial. Everyone respected the
opinions of each other, and no one dominated the discussions. Having
differing perspectives allowed us each to share what our thought process
was and how we interpreted the statements from the teachers. It was really
great to work together and did a lot to counteract any imposter syndrome
I was feeling as the only non-­faculty member of the group.

Karin agreed:

When I scheduled the meeting, I called it a “coding party.” Not only did
we code, but we also had lunch together in that conference room. We got
to know each other better as colleagues and genuinely had a good time
laughing and talking about life in academia. The time passed quickly
because we were having fun.

Study 2: Experiences of STEM Clubs for Students with

Disabilities and Their Parents
After Study 1 was completed, there were several changes to our research
team: one member left the university to return to k-­12 education (Kelly
Brooksher) and one faculty member joined the research team (Barbara
Serianni, a senior faculty member in the special education program who
asked to join the research team based on her experiences teaching in special
education). We then received funding from the National Youth Advocacy
206 Peggy Shannon-Baker et al.

and Resiliency Foundation for Study 2, which allowed us to hire a graduate

assistant to join the research team.
Whereas Study 1 focused on teachers’ perspectives, the purpose of Study
2 was to explore the views of SWD and their parents on their recruitment to
and retention in STEM clubs (Fisher et al., 2022). Our original research plan
for Study 2 entailed shadowing students as they participated in after-­school
STEM clubs in local school districts. The data would have come from obser-
vations, interviews, and other sources. However, due to the COVID-­19 pan-
demic, schools closed, and many after-­school clubs were canceled. As a
result, we decided as a team to transition from in-­person data collection to
online data collection.
Instead of focusing specifically on students in rural-­based school districts,
we redesigned the study to start with a survey available to US-­based SWD
and their parents. The survey allowed us to collect participant demographics,
the STEM club types, and participation reasons and barriers. For additional
details on our survey questions and procedures, see Fisher et al. (2022).
Following a similar design as Study 1, for Study 2, we chose student and
parent participants from the respondents to the initial quantitative survey.
Participants’ responses on the types of clubs the SWD were involved in and
the nature of their disabilities allowed us to choose participants across race,
class, disability, and STEM club type. Five students and five parents were
interviewed in one of two formats: either individual interviews (one student
and one parent chose this option) or a focus group (four students partici-
pated in a student focus group and four parents in a parent focus group).
These interviews resulted in a total of 57 transcribed pages.

Teaching and Learning Qualitative Coding through

a Group-Based Process
As a team, we decided to take advantage of having more members com-
pared to Study 1 by splitting up the qualitative analyses. This change
allowed us to work in two groups. One group completed the initial inductive
analysis to identify codes and themes within the transcripts, whereas the
second group reanalyzed the transcripts to verify our initial themes. In
Group 1, Peggy met with Karin and a former member of the research team
several times to review how coding entails identifying a word or short
phrase based on a piece of data (Saldaña, 2016). As a teacher of research
methods, Peggy modeled how to code, including using quotation marks for
directly quoted codes, keeping a code list (Saldaña, 2016), and reviewing
the data multiple times to identify more instances of existing codes or new
codes (Creswell & Plano Clark, 2017; Miles et al., 2020; Saldaña, 2016).
During this process, Karin, a more quantitatively oriented researcher, was
 Qualitatively Oriented Mixed Methods, Students with Disabilities 207

concerned about the number of decisions, nuances, and subjectivity that

went into the coding process. However, Peggy reassured her by providing
resources, including articles, book chapters, and qualitative exemplar stud-
ies (e.g., Nelson et al., 2020; Saldaña, 2016). Reflecting on this experience,
Karin shared that Peggy’s guidance and resources helped her feel more con-
fident in the coding process:

Initially, I was uncomfortable with the coding not being as black and
white like quantitative research. This caused me a level of anxiety about
whether we were correctly analyzing the data. After discussing my con-
cerns with Peggy and reviewing the qualitative research resources, I felt
more comfortable and confident in the process.

From Peggy’s perspective: “Because Karin shared her concerns, we could

work through them as a team. This allowed us to go back to the characteris-
tics of qualitative research that emphasize subjectivity and complexity and
demonstrate how these are strengths to qualitative coding.” We found that
working in a smaller group for this study allowed these more intimate discus-
sions about comfort with qualitative coding to surface.

The Qualitative Coding Process by Group

For Study 2, group members decided to forego the hard copy transcripts and
instead try using Google Drive to code. This allowed all group members to
access the documents without having to print, the ability to merge com-
ments, and the ease of access for anyone working from home. Google Drive
is an online cloud system, similar to Dropbox or One Drive, that allows
users to create files (e.g., a Google Doc is the equivalent word processing
system as Microsoft Word) and store files in folders that can be shared
among research team members. We used Google Drive because our univer-
sity provided Google accounts to all users. Karin uploaded all the transcript
files into folders on Google Drive that were then shared with all members of
the research team. This sharing allowed each team member editorial access
to the transcript files located in a folder just for the data from the study.
Separate folders were created in Google Drive for team members to code
copies of the data files so that the raw data files were not altered.
Group 1 members began analyzing the data by rereading all the tran-
scripts at least twice to refamiliarize themselves with the data. Each group
member then individually coded the transcripts at least twice by making a
copy of the raw transcript file and saving the copy in a data analysis folder
located in Google Drive with their initials in the file name. We first coded the
student transcripts to prioritize their experiences in STEM clubs. Coding
208 Peggy Shannon-Baker et al.

from their voices allowed us to return to their experiences even when analyz-
ing the parents’ transcripts rather than vice versa. During this initial round
of coding, we decided to use a more inductive analytical approach to empha-
size participant voices (Poth & Shannon-­ Baker, 2022). These inductive
methods allowed us to better embrace the wholeness and complexity of their
experiences than would have been possible with predetermined themes or
codes. We used a combination of descriptive and in vivo coding. We defined
descriptive coding as using our own words to summarize the main idea or
topic from a piece of data and in vivo coding as using the exact words from
the participant to summarize the main idea or topic from a piece of data
(Saldaña, 2016). Document 12.3 displays how we used comment boxes in a
Google Doc version of a transcript with various codes, such as “parental
involvement” and “I was tapped out,” to track our coding.

[SEE DOCUMENT 12.3 at eResource: Example of Initial Coding from a Member of

Group 1 for Study 2].

Group 1 then met several times to discuss the codes that each team member had
identified. We began each meeting by discussing codes or quotes that stood out
from our most recent coding. This was done by screen-­sharing our coded data
files during Zoom-­based meetings. During these meetings, Karin took notes in
a separate Google Doc located in our shared Google Drive folders. We then
compiled our individual code lists into one long list in a Google Doc to sort into
categories, debating whether a particular code, such as a club being “fun and
creative,” fit into one or more categories. These meetings allowed us to discuss
the different coding processes and specific codes by section of each transcript.
Although we identified many codes and categories, we prioritized categories
that most clearly aligned with our research questions (e.g., the code “advice for
parents”). We then created a list of codes and significant categories (see
Document 12.4). Some categories, such as teacher based, had many descriptive
(e.g., “teachers leave”) and in vivo (e.g., “don’t wanna get sued”) codes.

[SEE DOCUMENT 12.4 at eResource: Excerpt from the Code and Category List for
Study 2].

Next, we grouped the codes and categories into major areas that related to our
research questions: (a) reasons for joining, staying in, and participating in a
STEM club; (b) barriers to joining or staying in a club; (c) students’ general
experiences while in the club; and (d) the impact of COVID-­19. The members
of Group 2 (Kania and another member of the team) then used these major
areas to reanalyze clean copies of the transcripts. These copies were made by
again copying the raw data files located in our shared Google Drive folders
with the research team member’s name in the copied file name. This allowed
 Qualitatively Oriented Mixed Methods, Students with Disabilities 209

Group 2 to focus on the codes that related specifically to the research questions
and ensure previously coded transcripts did not bias them. The second coding
stage entailed assigning colors to each major area and cross-­referencing their
color-­coded transcripts against the original code and category list created by
Group 1. This step aimed to validate the code list through dialogue rather than
through quantitizing forms of intercoder reliability, which can “conceal signifi-
cant disagreements among coders” (Seltzer-­Kelly et al., 2012, p. 264).
We discussed examples of data coded from Group 2 that were not reflected
in the original code and significant category list, which resulted in a few
more code additions. Reflecting the process of comparing coded data
between the groups as a Group 2 member, Kania stated:

This portion of the coding really gave us in Group 2 a unique perspective

of the transcripts. When we had questions, we had to read subsequent
text around the coded pieces of information and try to nuance out what
the other coders were thinking. However, because of my previous experi-
ence in coding with Peggy and Karin, this was much easier to do and
made for a much more focused and streamlined process.

Document 12.5 displays how some codes (highlighted in pink) were added
due to Group 2’s coding. The process helped us identify that “students’ gen-
eral experiences while in the club” as a code was too broad. The highlighted
data from Group 2 was more specific than the codes and categories from
Group 1. We then decided to break out the codes and categories in this area
into the main topics of recruiting and retaining students in the STEM clubs
to refocus on our research questions.

[SEE DOCUMENT 12.5 at eResource: Revised Excerpt from the Code and Category
List for Study 2].

Finally, different team members then each took one major area to reread
example data coded in that area to write a description of the findings. These
descriptions were used to compare against the quantitative results during
our mixed methods integration analysis phase. These descriptions were then
incorporated into our first article as the findings section (Fisher et al., 2022).

Iterative Development of Visuals for the Mixed Methods

Integration Analysis
Integration refers to the processes and practices mixed methods researchers
use to combine or interface multiple methods. Integration can take place
between research phases and at the end of a study (Creswell & Plano Clark,
2017). We used two forms of integration in Study 2. First, we used the
210 Peggy Shannon-Baker et al.

quantitative survey results to identify participants for the qualitative phase,

demonstrating connecting integration. Second, we combined the quantitative
results from parents and students with the qualitative findings, demonstrat-
ing merging integration (Creswell & Plano Clark, 2017). Whereas Study 1
primarily used connecting integration, we added merging integration to
Study 2 to enhance our conclusions.
Merging integration processes often include making a joint display,
which is a visual portraying the relationships between multiple sets of data
or findings (Plano Clark & Sanders, 2015). Peggy was responsible for the
merging integration analyses and write-­ups. They first tried making a table
to compare these findings, which is a common form of joint display for
merging integration (Plano Clark & Sanders, 2015). In Table 12.2, they
created a table to show how quantitative results and qualitative quotes
from students and parents could connect to a meta-­theme for the whole
study (e.g., top reasons for participating in a STEM club) followed by our
inferences and interpretation of that meta-­theme. However, this resulted in
an eight-­page table in landscape orientation due to the complexity of
including data from both students and parents as well as the description of
inferences for each meta-­theme. The team decided the table was too large
to meaningfully work with or portray in a single publication, primarily
because we intended to target a special education journal, not a method-
ological journal. We likely would not have the word count to publish the
whole table.

[SEE TABLE 12.2 at eResource: Excerpt from the Table-­Based Joint Display Draft].

Peggy then began experimenting with different formats for portraying the
merging integration analysis. One design included a circular display to align
similar supporting data to a particular meta-­theme organized within a pie-­
shaped cutout of the circle (see Figure 12.1). The display would have allowed
us to organize meta-­themes by quadrant within a ring of the circle (e.g.,
upper right for “Social and Peer Interactions”). The circular display, how-
ever, did not allow for enough space to include direct quotes from students
and parents (see the abbreviations listed in Figure 12.1 for where quotes
would have been placed). Including direct quotes in the visual was a priority
because our study was designed to center participants’ voices.

[SEE FIGURE 12.1 at eResource: Circular-­Based Joint Display Draft].

The second version of their joint display was more box-­shaped to provide
more space for participant quotes (see Figure 12.2). Each outer square was
labeled in bold with one of the meta-­themes like “Semblance of Normalcy.”
The goal was to provide the supporting quantitative results in the same
 Qualitatively Oriented Mixed Methods, Students with Disabilities 211

space (the inner square) as the quotes to provide a visual summary of each
meta-­theme. This kind of joint display allowed Peggy to use color to differ-
entiate between participant groups (blue for parents, yellow for students).
However, the meta-­theme summary was too long to incorporate into an
outer border around the data while using a legible font size.

[SEE FIGURE 12.2 at eResource: Quadrant-­Based Joint Display Draft].

The final version of the joint display Peggy created instead used different shapes
to indicate different data types and did not include the meta-­theme summary
(see Figure 12.3, which shows the joint display for the reasons SWD joined and
stayed in STEM clubs). They used ovals to indicate data from the quantitative
methods, boxes for quotes from the qualitative data, yellow for SWD data, and
blue for parent data. Peggy then designed two joint displays for these merging
integrations, one for each of the research questions: (1) reasons to join and stay
in STEM clubs and (2) barriers to joining/staying in STEM clubs.

[SEE FIGURE 12.3 at eResource: Final Published Quadrant Joint Display (Fisher
et al., 2022)].

Reproduced with permission from the publisher.

Challenges to Publishing Study 2

We published Study 2 (Fisher et al., 2022) in a journal focused on special educa-
tion research. During the review process, we received feedback that asked us to
compile all our methods into subsections based on the type of method (e.g.,
“Data Collection”) and to remove our research diagram visualizing the research
process between phases. Peggy and Karin emailed the editors to explain that we
used a sequential approach to our methods (findings from one phase influ-
enced the next) and cited a journal article on mixed methods reporting stan-
dards from the American Psychological Association (APA) Publications and
Communications Board task force (Levitt et al., 2018). We argued that follow-
ing the feedback to compile all methods together would misrepresent the study
because APA recommends using that organization for a different type of mixed
methods design. Instead, we suggested changing our heading names to be more
consistent with the recommendations from Levitt et al. (2018) based on our
sequential mixed methods design (e.g., “Phase 1: Participants and Procedures”).
Additionally, rather than removing the research diagram, we asked that it
instead be listed as supplementary material for the article. The editors agreed
with both suggestions, but the journal was unable to produce our joint displays
in the original color-­coding of yellow for students’ and blue for parents’ data
despite having both online and print versions of the journal. As a result, Peggy
212 Peggy Shannon-Baker et al.

had to create a grayscale version of the image. Although we were able to use
these article-­reporting standards to advocate using best practices for writing
about mixed methods, “mixed methods” was changed to “mixed-­methods”
throughout the article during copy-­editing. “Mixed-­methods” with a hyphen is
not used in the field, but this was not caught before publication. This is another
example of the challenge we faced publishing this article where the best prac-
tices for writing mixed methods research were not followed during production
of the publication.

Reflections and Recommendations for Qualitatively Oriented

Mixed Methods Teams
The purpose of this chapter was to describe the collaborative processes and
analytical practices of our qualitatively oriented mixed methods studies into
SWDs’ recruitment and retention in STEM clubs (Fisher et al., 2021, 2022). In
addition to detailing how we analyzed our qualitative data in the two studies,
we described our mixed methods integration analysis procedures in detail for
one study, outlined the challenges faced in publishing that study, and shared
our reflections on these processes throughout the chapter. Ongoing reflections
are essential to team-­based and qualitatively oriented mixed methods research
(Aarons et al., 2019; Hemmings et al., 2013; Mason, 2006; Poth & Shannon-­
Baker, 2022). This last section offers final recommendations for researchers
engaging in team-­based, qualitatively oriented mixed methods work.
First, we recommend that research teams explore and embrace each team
member’s specific research strengths. For example, Kania had a detailed
knowledge of the local school districts and experience with mixed methods
and informal STEM learning environments. Because Karin’s strengths are
leadership and organization, she organized team meetings, took detailed
notes during each session, and shared all communication with the team.
Peggy then brought their methodological and teaching strengths to mentor
other group members in qualitative analysis and visually represent the mixed
methods analyses. Both studies benefited from the diversity of the strengths
of each team member. We recommend that research team leaders be inten-
tional when recruiting team members based on these complementary
strengths. Karin purposefully identified potential colleagues based on their
strengths to complement her weaknesses. At least one team member should
have organization and communication skills. We also recommend all team
members have a reputation for being collaborative and open to new ideas.
Second, we recommend using integration approaches that emphasize quali-
tative data when doing qualitatively oriented mixed methods research. There
are many ways to integrate data in mixed methods that prioritize quantitative
approaches, such as turning qualitative data into quantitative data, known as
 Qualitatively Oriented Mixed Methods, Students with Disabilities 213

quantitizing (Sandelowski et al., 2009). Emphasizing qualitative practices here

means embracing inductive approaches, messiness, and nonlinearity (Poth &
Shannon-­Baker, 2022). In practice, this emphasis starts with aiming to center
participant voices during analyses through detailed quotes, sharing stories spe-
cific to participants’ unique identities, and describing the study location in its
social and political context. Researchers should use integration practices that
preserve and amplify participants’ experiences, such as writing vignette case
studies (e.g., Ling & Pang, 2022), creating visuals (Shannon-­Baker, 2021), dance
(e.g., Kawano, 2018; Wilson & Moffett, 2017), and community performances
(e.g., Bagley & Castro-­Salazar, 2012). Additionally, focusing on the qualitative
data when sharing integrated findings demonstrates the qualitative emphasis
(e.g., Fisher et al., 2021, 2022; Love et al., 2018; Shannon-­Baker, 2015).
Ultimately, using such practices demonstrates the benefits of having a qualita-
tive orientation to a mixed methods study (Poth & Shannon-­Baker, 2022).
Third, we recommend that researchers familiarize themselves with the
journal article reporting standards for qualitative and mixed methods
research (cf. Levitt et al., 2018). These standards were established to identify
the elements that should be showcased in a publication to emphasize its qual-
ity and rigor. However, meeting these standards, particularly in a content
area journal as opposed to a methodological journal, can be difficult due to
the complexity of mixed methods studies and overall findings needing addi-
tional space to explain in a detailed and accessible manner. Mixed methods
scholars recommend using the language of the field to clarify the research
design and integration analyses (e.g., Levitt et al., 2018; Poth & Shannon-­
Baker, 2022), which requires additional space to explain in non-­methodological
journals where readers may be unfamiliar with such specific terminology.
In our experience, publishing qualitatively oriented mixed methods
research was challenging when working within the publication word count
limits. Lower word count limits made it more difficult to include rich evi-
dence from both qualitative and quantitative data to support a meta-­theme.
We also faced difficulties publishing visuals aimed at making our procedures
clearer and thereby more credible.
Researchers must consider the feedback received during the review pro-
cess and advocate for the terminology, writing organization, and visualizing
practices that will best communicate their research accurately. Sharing the
reporting standards with reviewers and editors is one way to address feed-
back to support such advocacy. Additionally, we recommend researchers use
other mechanisms available in publishing, such as asking for online versions
to be in color or moving visuals to supplementary material if necessary. As
this volume demonstrates, more details about researchers’ practices add
credibility and rigor to studies as well as make them useful for teaching how
to apply such practices to readers.
214 Peggy Shannon-Baker et al.

Acknowledgment
We would like to acknowledge the contributions of Dr. Kelly Brooksher,
Dr. Barbara Serianni, and Kayla Fitzgibbons for their work on the studies
and previous publications. Although they were not involved in this publica-
tion, their work was vital to the studies described here.

KEY WORKS GUIDING OUR DATA ANALYSIS

Saldaña, J. (2016). The coding manual for qualitative researchers (3rd ed.). Sage. (Chapter 4:
“After first cycle coding,” pp. 212–232).
Although this whole text is beneficial for emerging qualitative researchers,
this chapter, in particular, is beneficial for these researchers and those who plan
to apply qualitatively oriented mixed methods approaches. Saldaña’s discus-
sions of code mapping and code charting are similar to the practices we used
during our second study when analyzing our codes and how we integrated
qualitative and quantitative data.
Levitt, H. M., Bamberg, M., Creswell, J. W., Frost, D. M., Josselson, R., & Suárez-­Orozco,
C. (2018). Journal article reporting standards for qualitative primary, qualitative
meta-­analytic, and mixed methods research in psychology: The APA Publications
and Communications Board task force report. American Psychologist, 73(1), 26–46.
[Link]
This article reviews the standards for publishing both qualitative and mixed
methods research as set by the APA. Although the standards do not generally
address more creative and arts-­based research, they are helpful when publishing
more traditional reports of these types of research. This article helped us advocate
for having more details about our mixed methods process in our second study.
Hemmings, A., Beckett, G., Kennerly, S., & Yap, T. (2013). Building a community of
research practice: Intragroup team social dynamics in interdisciplinary mixed meth-
ods. Journal of Mixed Methods Research, 7(3), 261–273. [Link]
1558689813478468
This article analyzes the social dynamics and work of a research team as a com-
munity of research practice. Dynamics were characterized by processes of the forma-
tion of the group social identity that enabled members to cross discipline-­bordered
traditions and produce interdisciplinary mixed methods combinations.
Recommendations for other teams are provided.

References
Aarons, G. A., Seijo, C., Green, A. E., Moullin, J. C., Hasson, H., von Thiele
Schwarz, U., James, S., Ehrhart, M. G., Ducarroz, S., Sevdalis, N., & Willging, C.
(2019). Fostering international collaboration in implementation science and
research: A concept mapping exploratory study. BMC Research Notes, 12(1),
Article 778. [Link]
Bagley, C., & Castro-­Salazar, R. (2012). Critical arts-­based research in education:
Performing undocumented historias. British Educational Research Journal, 38(2),
239–260. [Link]
 Qualitatively Oriented Mixed Methods, Students with Disabilities 215

Bazeley, P. (2018). Integrating analyses in mixed methods research. Sage.

Cameron, R. (2016). The role of qualitative methods in mixed methods designs. In
K. Townsend, R. Loudoun, & D. Lewin (Eds.), Handbook of qualitative research
methods on human resource management: Innovative techniques (pp. 13–29).
Edward Elgar.
Creswell, J. W., & Plano Clark, V. L. (2017). Designing and conducting mixed methods
research (3rd ed.). Sage.
Fisher, K. (2016). The relationship between extracurricular STEM activities and per-
formance on the Florida Science Assessment. [Unpublished doctoral dissertation].
University of Central Florida.
Fisher, K. M., Shannon-­Baker, P., Brooksher, K., & Greer, K. (2021). Characteristics
of rural STEM clubs and implications for students with disabilities. Special
Education Research, Policy & Practice, 5, 15–39. [Link]
[Link]/teach-­elementary-­facpubs/37
Fisher, K. M., Shannon-­Baker, P., Greer, K., & Serianni, B. (2022). Perspectives of
students with disabilities and their parents on influences and barriers to joining
and staying extracurricular STEM activities. The Journal of Special Education,
56(2), 110–120. [Link]
Hatch, J. A. (2002). Doing qualitative research in education settings. SUNY Press.
Hemmings, A., Beckett, G., Kennerly, S., & Yap, T. (2013). Building a community of
research practice: Intragroup team social dynamics in interdisciplinary mixed
methods. Journal of Mixed Methods Research, 7(3), 261–273. [Link]
10.1177/1558689813478468
Hesse-­Biber, S. (2018). Toward an understanding of a qualitatively driven mixed
methods data collection and analysis: Moving toward a theoretically centered
mixed methods praxis. In E. Flick (Ed.), The SAGE handbook of qualitative data
collection (pp. 545–564). Sage.
Hesse-­Biber, S. N. (2010). Mixed methods research: Merging theory with practice. The
Guilford Press.
Hesse-­Biber, S., & Johnson, R. B. (2013). Coming at things differently: Future direc-
tions of possible engagement with mixed methods research. Journal of Mixed
Methods Research, 7(2), 103–109. [Link]
Jordan, T. R. & Bartel, M. (2023). Prioritizing cultural responsiveness in mixed
methods research and team science with underrepresented communities. In C. N.
Poth (Ed.), The SAGE handbook of mixed methods design (pp. 233–244). Sage.
Kawano, T. (2018). Transmission of the professional identity through an embodied
artistic ritual: An investigation of a dance/movement therapy welcoming cere-
mony. The Arts in Psychotherapy, 57, 1–10. [Link]
King, N., Horrocks, C., & Brooks, J. (2019). Interviews in qualitative research (2nd
ed.). Sage.
Levitt, H. M., Bamberg, M., Creswell, J. W., Frost, D. M., Josselson, R., & Suárez-­
Orozco, C. (2018). Journal article reporting standards for qualitative primary,
qualitative meta-­analytic, and mixed methods research in psychology: The APA
Publications and Communications Board task force report. American Psychologist,
73(1), 26–46. [Link]
Ling, H. L., & Pang, M. F. (2022). A vignette-­based transformative multiphase
mixed methods interventional study featuring Venn Diagram joint displays:
216 Peggy Shannon-Baker et al.

Financial education with Hong Kong early adolescent ethnic minority students.
Journal of Mixed Methods Research, 16(1), 130–149. [Link]
1558689821989834
Love, C. D., Booysen, L. A. E., & Essed, P. (2018). An exploration of the intersection
of race, gender and generation in African American women doing social justice
work. Gender, Work & Organization, 25(5), 475–494. [Link]
gwao.12095
Mason, J. (2006). Mixing methods in a qualitatively driven way. Qualitative Research,
6(1), 9–25. [Link]
Miles, M. B., Huberman, A. M., & Saldaña, J. (2020). Qualitative data analysis: A
methods sourcebook (4th ed.). Sage.
Nelson, T., Shahid, N. N., & Cardemil, E. V. (2020). Do I really need to go back and
see somebody? Black women’s perception of help-­seeking for depression. Journal
of Black Psychology, 46(4), 263–286. [Link]
Plano Clark, V. L., & Ivankova, N. V. (2016). Mixed methods research: A guide to the
field. Sage.
Plano Clark, V. L., & Sanders, K. (2015). The use of visual displays in mixed meth-
ods research: Strategies for effectively integrating the quantitative and qualitative
components of a study. In M. T. McCrudden, G. Schraw, & C. W. Buckendahl
(Eds.), Use of visual displays in research and testing: Coding, interpreting, and
reporting data (pp. 177–206). Information Age Publishing.
Poth, C. N., & Shannon-­Baker, P. (2022). State of the methods: Leveraging design pos-
sibilities of qualitatively oriented mixed methods research. International Journal of
Qualitative Methods, 21, 1–11. [Link]
Saldaña, J. (2016). The coding manual for qualitative researchers (3rd ed.). Sage.
Sandelowski, M., Voils, C. I., & Knafl, G. (2009). On quantitizing. Journal of Mixed
Methods Research, 3(3), 208–222. [Link]
Seltzer-­Kelly, D., Westwood, S. J., & Peña-­Guzman, D. M. (2012). A methodological
self-­
study of quantitizing: Negotiating meaning and revealing multiplicity.
Journal of Mixed Methods Research, 6(4), 258–274. [Link]
1558689811425798
Shannon-­Baker, P. (2015). “But I wanted to appear happy”: How using arts-­informed
and mixed methods approaches complicate qualitatively driven research on cul-
ture shock. International Journal of Qualitative Methods, 14(2), 34–52. [Link]
org/10.1177/160940691501400204
Shannon-­Baker, P. (2021). The palimpsest as an arts-­based integration strategy for
mixed methods research. Caribbean Journal of Mixed Methods Research, 2(1),
1–19. [Link]
Shannon-­Baker, P., & Martinez, C. (2022). Integrating arts-­based approaches in
mixed methods research with diverse stakeholders. In J. H. Hitchcock & A. J.
Onwuegbuzie (Eds.), The Routledge handbook for advancing integration in mixed
methods research (pp. 372–387). Routledge.
Toledo, C. & Shannon-­Baker, P. (in press). Choosing a qualitatively oriented mixed
methods research approach: Recommendations for researchers. In R. Cameron &
X. Golenko (Eds.), Handbook of mixed methods in business and management.
Edward Elgar.
 Qualitatively Oriented Mixed Methods, Students with Disabilities 217

Wilson, L., & Moffett, A.-T. (2017). Building bridges for dance through arts-­based
research. Research in Dance Education, 18(2), 135–149. [Link]
647893.2017.1330328
Yin, R. K. (2016). Qualitative research from start to finish (2nd ed.). Guilford Press.
13
MIXING METHODS TO ADVANCE
OUR UNDERSTANDING OF
PARENTAL STRESS AND COPING IN
YOUTH SPORT
Sam N. Thrower, Travis E. Dorsch, Camilla J. Knight,
and Chris G. Harwood

Within the field of sport psychology, it is widely accepted that parents play a
critical role in the performance, wellbeing, and long-­term development of
young athletes (Knight, 2019). As such, over the last decade, considerable
attention has been given to examining parents’ roles, involvement, and influ-
ence within the context of youth sport (see Dorsch et al., 2021). One topic
that has received significant academic attention is parents’ experiences of
stress (i.e., the demands that parents experience due to, or associated with,
their children’s engagement in youth sport). Although initial studies (i.e.,
Burgess et al., 2016; Harwood & Knight, 2009a, 2009b; Harwood et al., 2010;
Hayward et al., 2016; Lienhart et al., 2020) have provided invaluable insights
into the stressors faced by parents in sport, they have overlooked other
important aspects of the stress experience (i.e., appraisals, emotions, coping,
and coping effectiveness; Lazarus, 1999). As such, it was proposed that to
fully understand parents’ stress experiences, there may be a benefit to utilizing
mixed methods approaches to capture crucial associations between stressors,
appraisals, emotions, and coping (Lazarus & Folkman, 1984; Lazarus, 2000).
In this chapter, we explain how a mixed methods design was used to
explore psychological stress and coping more fully among sport parents. In
doing so, we place particular emphasis on how qualitative analysis occurred
within the context of a mixed methods research (MMR) study (Harwood
et al., 2019). Because the process of data analysis in MMR cannot be easily
separated from the assumptions and decisions that precede it, this chapter
begins by presenting some of the contemporary critical debates within the
social sciences about MMR, which show that it is a contested and ambigu-
ous topic, particularly for qualitative researchers (Sparkes, 2015).

DOI: 10.4324/9781003254041-14
 Mixed Methods, Parental Stress, and Coping in Youth Sport 219

Mixing Methods: Points of Controversy and

Ways of Moving Forward
In its simplest sense, MMR refers to the process of collecting, analyzing,
and integrating both qualitative and quantitative data within a single study
or series of studies (Creswell & Plano Clark, 2018; Tashakkori & Creswell,
2007). Although this process has been labeled by some scholars as a “third
methodological movement” (e.g., Johnson et al. 2007), using more than one
method of data collection is not a new approach and is often viewed as an
intuitive and practical way of understanding complex human experiences or
practices (Gibson, 2016). In addition, MMR is considered to have the poten-
tial to offset weaknesses of a single approach alone, provide stronger infer-
ences, enhance credibility, produce more robust knowledge claims, generate
a more comprehensive or broader understanding of the research problem,
and inform sampling, hypothesis development, or measurement develop-
ment (Gibson, 2016; Sparkes, 2015). Taking these points into consideration,
it is unsurprising that there has been a huge increase in the popularity of
MMR within the social sciences (McGannon & Schweinbenz, 2011), as well
as a growing demand for interdisciplinary research and MMR from funding
agencies (Gibson, 2016; Teddlie & Tashakkori, 2009).
The rising popularity of MMR has also led to several critical debates within
the social science literature (Sparkes, 2015). For example, within the field of
sport and exercise psychology, there have been disagreements about the chal-
lenges of mixing methods, methodologies, and paradigms from researchers
adopting pragmatic (e.g., Giacobbi et al., 2005; Moran et al., 2011), interpretiv-
ist (e.g., Sparkes, 2015), and critical realist positionalities (e.g., Ryba et al.,
2022). Within these articles, the main point of controversy surrounding MMR
is not about mixing methods per se but issues surrounding ontology and epis-
temology (i.e., the “incompatibility thesis;” Gill, 2011; Howe, 1988).
Beyond epistemological and methodological constraints, researchers
have raised important concerns about the dominance of the quantitative
component in most MMR designs. Some researchers feel there is a “power
hierarchy” (Giddings, 2006), wherein qualitative methods are often used in a
supportive role in experimental or intervention research (Creswell et al.,
2011; Giddings, 2006). Furthermore, there are concerns regarding the ten-
dency for researchers to present qualitative and quantitative results in paral-
lel ways with limited integration (e.g., Bryman, 2007; Cheek, 2021). Such
concerns have led to calls for greater clarity regarding how mixed methods
researchers analyze, interpret, and write up their research in a way that is
“mutually illuminating” (Bryman, 2007; Moran et al., 2011).
These points of controversy within the social sciences present MMR as a
disputed and challenging methodology (Sparkes, 2015). However, a number
of solutions have been presented that have paved the way for conducting
220 Sam N. Thrower et al.

MMR. The most obvious, and common, solution to the “incompatibility

thesis” (see Gill, 2011; Howe, 1988) is simply for researchers to continue
operating within their paradigmatic position. Philosophical assumptions
informing a paradigm influence its methodology, but they make no claims
that these determine the methods used in any given study (Sparkes, 2015).
An alternative solution has been to reconsider the philosophy of knowledge
by replacing the emphasis on abstract concerns like reality and truth with
inquiry about the human experience (e.g., pragmatism; see Creswell & Plano
Clark, 2018; Feilzer, 2010; Morgan, 2007).
These different solutions to the “incompatibility thesis” (Gill, 2011;
Howe, 1988) have also enabled researchers to shift their focus toward
addressing some of the other more straightforward points of controversy
within the literature. For instance, researchers have developed MMR designs
(see Creswell & Plano Clark, 2018) and shorthand capitalization/symbols to
illustrate if qualitative and quantitative data are given equal emphasis or
priority (e.g., QUANT + QUAL or QUANT → Qual; Morse, 2003). Although
these steps do not necessarily resolve the aforementioned “power hierarchy,”
they do enable researchers to more effectively communicate to the reader
why decisions were made (i.e., transparency) and how qualitative and quan-
titative elements were integrated within a specific study to be greater than the
sum of its parts (Barbour, 1999; Bryman, 2007; Cheek, 2021).
Furthermore, to address concerns regarding the separation criteria
(i.e., the quality criteria for each aspect of the study being considered
separately), specific MMR evaluation criteria have been developed that
place emphasis on integration in MMR, such as relevance to research
questions, transparency, a rationale for using MMR, and the need for
integration of MMR findings (Bryman et al. 2008). However, few MMR
studies state their position, and even fewer engage in thoughtful and thor-
ough methodological discussions about their work (see Ryba et al., 2022).
As a result, within the following sections, we illustrate how qualitative and
quantitative can be integrated within an MMR research design (Harwood
et al., 2019).

Mixing Methods: The Parental Stress and Coping Study

As noted earlier, to extend the literature on sport parent stress experiences
and align more closely with appropriate theories (e.g., Cognitive Motivational
Relational Theory; Lazarus, 1999), researchers needed to better capture the
crucial associations between stressors, appraisals, emotions, and coping in a
more linear manner. In advancing this field of research, we (i.e., Harwood
et al., 2019) aimed to investigate a series of interconnected questions in a
large cross section of parents within one specific sport. Specifically, the
research questions were:
 Mixed Methods, Parental Stress, and Coping in Youth Sport 221

a) What are the prominent stressors that tennis parents experience?

b) How are these stressors appraised?
c) What emotions are associated with such stressors?
d) What coping strategies do parents use?
e) How effective are these coping strategies?
f) Does perceived effectiveness vary as a function of stressor type, appraisal,
and coping strategy used?

Such research questions represented a significant methodological challenge due

to the context-­specific (i.e., British tennis) nature of the stressors and the absence
of measures that capture the complexity and idiographic nature of stressors
and coping. In this respect, the study was conducted from a post-­positivistic
philosophical position, which acknowledges that some aspects of the social
world cannot be directly measured yet seeks to retain an objective approach
that is as free from bias as possible (Weed, 2009) and where quantification of
data is often used to investigate relevant theory (Krane & Baird, 2005). This
philosophical position was determined by the first and last authors who were
responsible for setting up the study and conducting data collection (authorship
references refer to the example paper; see Harwood et al., 2019). The second,
third, and fourth authors were invited to be part of the study due to their quali-
tative, quantitative, and theoretical expertise, respectively. It is important to
note that although the second and fourth authors had different philosophical
positions, they agreed to conduct the study from a post-­positivistic perspective.
Consistent with this philosophical position, a multipart concurrent mixed meth-
ods design (Morgan, 2013; Morse, 2003) was adopted in which qualitative and
quantitative data were collected simultaneously, with both forms of data given
equal emphasis and priority (i.e., QUAL + QUANT; Morse, 2003).

Data Collection within a Mixed Methods Design

When collecting data from the 135 parents who participated in the study, it
was critical to view parents as experts on their own experiences and to orien-
tate them toward key concepts that would feature in the study. To do this, a
short guidebook was developed to enhance parents’ understanding of the
key terms (i.e., stressor, appraisal, emotion, coping, and coping effective-
ness) before using worked examples (Figure 13.1) as an instructional process
that helped parents to complete an open-­ended online survey.

[SEE FIGURE 13.1 at eResource: Example Stressors from the Guidebook given to
Participants].

Within the online survey, quantitative measures captured the appraisal, emo-
tions, and coping effectiveness associated with each stressor and coping strategy
222 Sam N. Thrower et al.

that the parent disclosed, and qualitative data were gathered on parents’ subjec-
tive perceptions of prominent stressors and coping strategies. Specifically, after
reading the working guidebook, parents were asked to share their first of up to
five most prominent stressors they had faced as a tennis parent (see Figure 13.2).

[SEE FIGURE 13.2 at eResource: Example of how Qualitative Data were Collected]

Next, for this particular stressor, drawing on the procedures outlined by

Hanton et al. (2012), parents were asked to record whether they appraised
this stressor as a harm/loss (i.e., damage to goals or values that has already
occurred), threat (i.e., anticipated future damage), challenge (i.e., anticipated
gain), or benefit (i.e., gain that has already occurred; Lazarus, 1999, see
Figure 13.3). Then, based on a brief version of the Sport Emotion
Questionnaire (SEQ; Jones et al., 2005), participants were asked to record
how anxious, dejected, excited, happy, and angry they felt in response to that
specific stressor on a five-­point rating scale of zero (not at all), one (a little),
two (moderately), three (quite a bit), and four (extremely).

[SEE FIGURE 13.3 at eResource: Example of how Quantitative Data were Collected]

Finally, parents were asked to articulate up to three strategies they used to

cope with that stressor in an open-­ended format. Then, following each sepa-
rate strategy box, participants were asked to rate on an 11-­point scale (0–10)
the effectiveness of each coping strategy. In accordance with previous stud-
ies (i.e., Didymus & Fletcher, 2012; Nicholls et al., 2007; Levy et al., 2009),
this assessment of coping effectiveness was used to understand whether the
strategy employed was perceived to be successful in alleviating any negative
outcomes of the stressor. Overall, this process of articulating a stressor in
qualitative terms and then assessing its properties and effects in quantitative
terms before both qualitative and quantitative assessments of coping was
repeated up to five times by the participant. In total, there were 212 separate
entries submitted.

Analyzing Qualitative Data within a Mixed Methods Design

Consistent with the post-­positivistic philosophical position and novel multi-
part concurrent mixed methods design (Morgan, 2013; Morse, 2003), the
qualitative data from the online open-­ended survey were analyzed using
hierarchical content analysis. Hierarchical content analysis organizes and
reduces the content of the data by counting what is in it or by coding and
identifying themes at a semantic or latent level (Sparkes & Smith, 2013).
Specifically, we adopted an abductive (i.e., inductive and deductive) approach
 Mixed Methods, Parental Stress, and Coping in Youth Sport 223

to hierarchical content analysis. This approach had been applied elsewhere

in the literature (e.g., Didymus, 2017; Harwood et al., 2010) to enable the
creation of new ideas (i.e., inductive logic) while also applying theoretical
frameworks or lenses to participants’ experiences (i.e., deductive logic). The
procedures used in hierarchical content analysis aligned with the first and
last authors post-­positivistic commitment to reliability, objectivity, trust-
worthiness, and a psychological reality that can be found and known inde-
pendently of the researcher (Sparkes & Smith 2013; Thomas et al., 2019).
Data analysis began by exporting responses from the Online Surveys website
(formerly known as Bristol Online Surveys) to Microsoft Excel (i.e., sheet one;
see Table 13.1). Although we had access to qualitative coding software (i.e.,
NVivo), data were analyzed using Microsoft Excel so that we could quantify
the qualitative data using pivot tables while also being able to filter and sort the
data by both stressors and coping strategies at different phases of the analysis.
At this point, columns that were not needed for data analysis (e.g., ethical
approval and demographic information; see columns A–H) and 77 incomplete
entries (i.e., parents who had started but not finished the survey; see rows 6–8
for examples) were removed from the dataset. Each remaining participant (n =
135) was then allocated a participant number, and the columns were attributed
a heading label corresponding with the stressor, appraisal, emotions, coping
strategies, and coping effectiveness (Lazarus, 1999).

[SEE TABLE 13.1 at eResource: Raw Data Extracted from the Online Survey].

Because each participant could record up to five stressors, all of which were
located together within the exported data as one row (columns I–V =
Stressor 1, columns W–AK = Stressor 2, etc., see Table 13.1), prior to analy-
sis each stressor was moved to its own individual row. Following the hierar-
chical content analysis procedures outlined by Sparkes and Smith (2013),
the 342 individual stressors were then read to promote content familiarity
and initially coded independently by the second author. Specifically, a new
column was added titled initial coding and each open-­ended response was
inductively coded at a semantic level. For instance, responses from parents
such as “cheating of opponents in terms of bad line calls, and unsporting
behavior” were simply labeled as “bad line calls and cheating” (see row 27,
Table 13.2). Due to the specific instructions given during data collection,
only eight responses were not coded due to limited relevance or coherence
within the responses. Additionally, any response where a participant inte-
grated more than one stressor within a single response was split up into
separate rows (see rows 11–13 in Table 13.2 for an example).

[SEE TABLE 13.2 at eResource: Initial Coding of Participant Stressors].

224 Sam N. Thrower et al.

Following initial coding, column A was then sorted alphabetically to cluster the
same initial codes together and create lower-­order themes (e.g., “negative body
language” and “reluctance to challenge line calls/decisions”). These lower-­order
themes and the remaining initial codes were then carefully reviewed and codes
that were reworded differently (e.g., “poor line calls” and “bad line calls”) but
referred to the same stressor were reworded (e.g., “bad line calls”). Following
this, all lower-­order themes were then re-­reviewed and those that were consid-
ered conceptually similar (e.g., “Aggressive behavior” and “Inappropriate
behavior”) were combined together (e.g., “Aggressive and inappropriate behav-
ior”) to reduce the overall number of lower-­order themes (n = 51). It is impor-
tant to note that participants’ responses to the question “when did/does this
stressor occur?” also provided useful contextual information (see column D,
Table 13.2) to guide initial coding and the identification of lower-­order themes.
Next, additional columns were added, and similar lower-­order themes
were grouped together to create 20 more meaningful higher-­order themes
(e.g., “child’s opponent” and “child’s behavior”) and three general dimen-
sions (i.e., larger more inclusively meaningful clusters; Sparkes & Smith,
2013) (see Table 13.3). At this point, existing stressor classifications (i.e.,
competition, organizational, and developmental stressors; Harwood &
Knight, 2009a, 2009b) were deductively used to inform coding and promote
consistency regarding the terminologies that were used and applied within
published literature. Figure 13.1 shows the final stressor dataset sorted by
column A (i.e., general dimensions), column B (i.e., higher-­order themes),
and column C (i.e., lower-­order themes).

[SEE TABLE 13.3 at eResource: Example of how Qualitative Stressor Data were
Coded at a Lower-­Order, Higher-­Order, and General Dimension Level].

Because parents were able to identify up to three coping strategies for each
individual stressor, a separate Microsoft Excel spreadsheet (i.e., sheet two)
was created to facilitate the analysis of multiple coping strategies (and per-
ceptions of coping effectiveness). Specifically, data from sheet one were cop-
ied and pasted to sheet two and each individual coping strategy used was
then allocated its own individual row (n = 646). Following the same coding
process outlined above, each initial coping strategy was initially coded,
grouped into lower-­order themes, higher-­order themes, and general dimen-
sions. In addition, existing coping classifications such as mastery coping,
internal regulation, and goal withdrawal coping (Nicholls et al., 2016) were
deductively used during the later phases of coding to promote consistent
terminology within the literature (see Table 13.4). Following this qualitative
coding process, both the stressor and coping strategy themes were thor-
oughly reexamined against the original responses by the second author and
then independently crosschecked by the first author who had previously
 Mixed Methods, Parental Stress, and Coping in Youth Sport 225

published within this area (e.g., Harwood & Knight, 2009a, 2009b; Harwood
et al., 2010). The first and second author met to discuss disagreements which
resulted in a small number of lower-­order themes being recoded.

[SEE TABLE 13.4 at eResource: Example of how Qualitative Coping Strategy

Data were Coded at a Lower-­Order, Higher-­Order, and General Dimension
Level].

Next, the number of parents citing each lower-­order theme, higher-­order theme,
and general dimension were calculated for both stressors (sheet one) and coping
strategies (sheet two). Specifically, the second author inserted pivot tables, and
the distinct count function (which removes duplicates) was used to identify the
number of parents citing each general dimension, higher-­order theme, and
lower-­order theme, as well as the percentages. Tables were then created in
Microsoft Word to reflect the hierarchical nature of the results (i.e., most cited
themes first), including the frequencies of each cited stressor or coping strategy
(see Harwood et al., 2019). By keeping the stressor, appraisal, and emotion col-
umns (see columns A–L; Table 13.4) within sheet two, we were also able to use
the filter function to organize the data and explore the descriptive relationships
between stressors, appraisals, coping strategies, and coping effectiveness (see
Tables 4 and 5 in Harwood et al., 2019).
This process of quantifying the qualitative data (i.e., stressor and coping
strategies) in Microsoft Excel also enabled combined analyses to be conducted.
In the first phase, both qualitative and quantitative data were transferred to
SPSS and descriptive statistics (i.e., means and standard deviations) for the
emotions experienced (i.e., anxiety, dejection, happiness, excitement, and anger)
and coping effectiveness were calculated by the third author. Quantitative data
were screened for parametric assumptions and detected the presence of outliers
in happiness and excitement variables. Accordingly, when analyzing the data,
sensitivity analyses were conducted without these extreme cases (i.e., the
removal of four cases each for happiness and excitement), but these exclusions
did not materially change the pattern of results. It is important to note that due
to the journal word limit, we did not examine these cases from a qualitative
perspective. The main and interaction effects were the same, while two addi-
tional pairwise comparisons emerged as significant (these points are highlighted
in the results section of the example paper; Harwood et al., 2019). To comple-
ment the quantitative descriptive statistics, differences in emotions experienced
as a function of stressor category and appraisal were explored using a 3
(stressor: competition vs. organizational vs. developmental) × 3 (appraisal:
harm vs. threat vs. challenge) multivariate analysis (MANOVA). Differences
in coping effectiveness as a function of stressor category, appraisal, and coping
strategy were then assessed using a 3 (stressor: competition vs. organiza-
tional vs. developmental) × 3 (appraisal: harm vs. threat vs. challenge) × 3
226 Sam N. Thrower et al.

(coping strategy: mastery vs. internal regulation vs. goal withdrawal) analysis
of variance (ANOVA). Significant main or interaction effects were followed up
with Bonferroni adjusted pairwise comparisons.
As outlined earlier, MMR requires researchers to integrate the quantita-
tive and qualitative components in an effective and “mutually illuminating”
way (Bryman, 2007; Moran et al., 2011). To achieve this, the results section
was structured based on the cognitive-­ motivational-­ relational theory
(Lazarus, 1999). Specifically, data relating to stressors were presented first,
followed by primary appraisals, emotions, coping strategies, and coping
effectiveness. Specifically, the qualitative data (i.e., stressors and coping
strategies) were written up as a realist tale (Sparkes & Smith, 2013) and the
most prominent higher-­order themes were prioritized. Realist tales are char-
acterized by an absence of the author from the text, focus on what the par-
ticipants do or say, and rely on carefully selected and edited quotes to
represent the findings (Sparkes & Smith, 2013). As such, for each general
dimension (e.g., competition stressors), we included a descriptive overview
of the most prominent higher-­order themes, including the percentage of
parents citing each higher-­order theme (e.g., “child’s opponent [24.44%] was
the most prominent cause of stress for parents during competitions”), and
then carefully selected direct participant quotes to illustrate themes and pro-
vide evidence to support the claims we were making. However, it is impor-
tant to note that we selected quotes flexibly to highlight, illustrate, and bring
to life the points being made, not simply to represent the most frequently
cited stressors or coping strategies (Thorne, 2020). The final subsections of
the results explored the statistical differences in emotions experienced and
coping strategy effectiveness. A similar structure was also used within the
discussion section as it allowed us to discuss how each aspect of the study
(e.g., stressors, appraisals, emotions, coping strategies, and coping effective-
ness) contributed to knowledge individually while also providing a more
complete picture of the whole transactional stress process.

Conclusion
The aim of this chapter was to provide a behind-­the-­scenes look at how
MMR has been used to explore psychological stress more fully among sport
parents. Emphasis was placed on the process of quantifying qualitative data
to enable a combined analysis to be conducted. In doing so, we have high-
lighted how MMR is difficult, and at times messy, to conduct. It follows that
scholars who engage in the design and execution of high-­quality MMR
should constantly grapple with the best ways to access truth and reality in
attendance to a context or phenomenon of interest. As such, we conclude
this chapter both by reiterating the benefits of utilizing MMR as a research
methodology and by considering some of the challenges of its use, for both
 Mixed Methods, Parental Stress, and Coping in Youth Sport 227

neophyte and experienced researchers. As a final offering, we reflect on how

scholars might best employ MMR in their future work.
In recounting the benefits of MMR, it is important to emphasize how
combining qualitative and quantitative research can often foster a more
complete understanding of a research problem than either could alone.
Although it is too simple to ascribe broad knowledge (i.e., generalizability)
to quantitative approaches and deep knowledge (i.e., specificity) to qualita-
tive approaches, utilizing both in tandem within an MMR framework serves
a complementary function. In MMR, multiple experiences, from multiple
individuals, across multiple contexts, utilizing multiple data collection meth-
ods are combined in ways that allow a broader and deeper understanding of
a context or phenomenon.
Despite the potential benefits of utilizing an MMR approach, it is impor-
tant to also consider some of the challenges associated with utilizing such an
approach. First, becoming an expert in quantitative and qualitative method-
ologies is somewhat akin to learning two languages. In terms of MMR,
these languages utilize different terms, cut across sometimes competing phil-
osophical, ethical, methodological, and analytic lines, and tend to prompt
different types of research questions and hypotheses. A second challenge is
that even when a scholar gains expertise in multiple modes of data collection
and analysis, utilizing them in a single project can be costly in terms of time
and resources. Indeed, including two or more methods of data collection,
management, and analysis—and doing it well—can encumber researchers
who are pressured to “publish or perish.” These considerations are particu-
larly important for younger, less-­experienced, researchers who may not yet
be seasoned enough to successfully employ multiple lenses, perhaps in a dis-
sertation, thesis, or early-­career manuscript. Less-­experienced researchers
may also have limited access to time, money, and materials necessary to
execute such a project while fulfilling their many other roles. It can also be
tempting for young scholars to tease out disparate quantitative and qualita-
tive findings in an effort to publish two (or more) papers instead of one!
However, as mentioned above, when qualitative and quantitative elements
are integrated effectively within one paper, the whole can be greater than the
sum of the parts (Barbour, 1999; Bryman, 2007; Cheek, 2021).
Considering these strengths and limitations, we thought it important to
offer some brief reflections for scholars who are considering engaging in
MMR in the future. It is important that all researchers clearly outline the
assumptions that underpin their work and articulate what counts to them as
knowledge. The current chapter offers a detailed insight into how MMR can
be conducted from a post-­positivistic perspective, and we encourage readers
to also explore published MMR from social constructivist (e.g., Papathomas
et al., 2018), critical realist (e.g., Eckardt et al., 2022), and pragmatic posi-
tionalities (e.g., Thrower et al., 2023). It is also vital, in all cases, to recognize
228 Sam N. Thrower et al.

why using an MMR approach aligns with one’s research questions. It must
be clear from the outset why researchers need the combined insights of two
types of data to answer a question of interest. Through scholarly reflection,
understanding the exact role of the quantitative and qualitative datasets will
become clear and can guide researchers to the different data collection
approaches needed. This is particularly important when researchers choose
to employ distinct and varied data collection methods (e.g., quantitative sur-
vey or psychophysiological measures combined with in-­depth and multiple
interviews).
In an academic era where the impact of a scholar’s research study on
society is becoming increasingly important to universities and funding bod-
ies, it pays to consider MMR designs in terms of the power of resultant
research findings to deliver a more integrated, convincing, and balanced set
of findings. In the case of the present MMR study, it allowed many partici-
pants in a single sport to individually share their personal stress, emotion,
and coping journeys in ways that allowed us to better understand the state of
play for parents in youth tennis. The practical implications would not have
been as well informed by applying quantitative or qualitative methods alone.

KEY WORKS GUIDING OUR STUDY AND DATA ANALYSIS

Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods
research (3rd ed.). Sage.
Researchers interested in conducting MMR often begin their journey by reading
a practical how-­to guide for designing MMR studies. Although a number of such
textbooks exist, we found this book to be an excellent starting point for understand-
ing different mixed methods designs as well as providing insightful chapters on how
to collect, analyze, and interpret data in MMR.
Harwood, C. G., & Knight, C. J. (2009a). Stress in youth sport: A developmental inves-
tigation of tennis parents. Psychology of Sport and Exercise, 10(4), 447–456. https://
[Link]/10.1016/[Link].2009.01.005
This seminal paper was the first to examine the stressors parents experienced in
youth sport, through the application of Lazarus and Folkman’s (1984) transactional
approach. In doing so, Harwood and Knight provided valuable insights regarding
how to collect qualitative data about parents’ stressors using a cross-­sectional
online survey as well as how to quantify qualitative data using inductive and deduc-
tive content analysis.
Sparkes, A. C. (2015). Developing mixed methods research in sport and exercise psy-
chology: Critical reflections on five points of controversy. Psychology of Sport and
Exercise, 16(3), 49–59. [Link]
Although we have provided a brief overview in relation to the points of contro-
versy surrounding MMR within this chapter, we direct interested readers to this
paper to gain a more in-­depth and rich insight. We found this paper particularly
insightful and consider it an essential reading for anyone interested in conducting
MMR within the field of sport and exercise psychology.
 Mixed Methods, Parental Stress, and Coping in Youth Sport 229

References
Barbour, R. S. (1999). The case for combining qualitative and quantitative approaches
in health services research. Journal of Health Services Research & Policy, 4(1),
39–43. [Link]
Bryman, A. (2007). Barriers to integrating qualitative and quantitative research. Journal
of Mixed Methods Research, 1(1), 8–22. [Link]
Bryman, A., Becker, S., & Sempik, J. (2008). Quality criteria for quantitative, qualita-
tive and mixed methods research: A view from social policy. International Journal
of Social Research Methodology, 11(4), 261–276. [Link]
13645570701401644
Burgess, N. S., Knight, C. J., & Mellalieu, S. D. (2016). Parental stress and coping in
elite youth gymnastics: An interpretative phenomenological analysis. Qualitative
Research in Sport Exercise and Health, 8(3), 237–256. [Link]
9676X.2015.1134633
Cheek, J. (2021). Maintaining the integrity of qualitatively driven mixed methods:
Avoiding the “this work is part of a larger study” syndrome. Qualitative Health
Research, 31(6), 1015–1018. [Link]
Creswell, J. W., Klassen, A. C., Plano Clark, V. L., & Smith, K. C. (2011). Best prac-
tices for mixed methods research in the health sciences. Office of Behavioral and
Social Sciences Research National Institutes of Health. [Link]
scientific_areas/methodology/mixed_methods_research/[Link]
Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods
research (3rd ed.). Sage.
Didymus, F. F. (2017). Olympic and international level sports coaches’ experiences of
stressors, appraisals, and coping. Qualitative Research in Sport, Exercise & Health
9(2), 214–232. [Link]
Didymus, F. F., & Fletcher, D. (2012). Getting to the heart of the matter: A diary
study of swimmers’ appraisals of organisational stressors. Journal of Sports
Sciences, 30(13), 1375–1385. [Link]
Dorsch, T. E., Wright, E., Eckardt, V. C., Elliott, S., Thrower, S. N., & Knight, C. J.
(2021). A history of parent involvement in organized youth sport: A scoping
review. Sport, Exercise, and Performance Psychology, 10(4), 536–557. https://
[Link]/doi/10.1037/spy0000266
Eckardt, V. C., Dorsch, T. E., & Lobinger, B. H. (2022). Parents’ competitive stress-
ors in professional German youth soccer academies: A mixed-­method study.
Psychology of Sport and Exercise, 58. Advance online publication. [Link]
org/10.1016/[Link].2021.102089
Feilzer, M. Y. (2010). Doing mixed methods research pragmatically: Implications for
the rediscovery of pragmatism as a research paradigm. Journal of Mixed Methods
Research, 4(1), 6–16. [Link]
Giacobbi, P. R., Poczwardowski, A., & Hager, P. (2005). A pragmatic research phi-
losophy for sport and exercise psychology. The Sport Psychologist, 19(1), 18–31.
[Link]
Gibson, K. (2016). Mixed methods research in sport and exercise: Integrating quali-
tative research. In B. Smith & A. C. Sparkes (Eds.), Routledge handbook of quali-
tative research in sport and exercise (pp. 404–418). Routledge. [Link]
org/10.4324/9781315762012
230 Sam N. Thrower et al.

Giddings, L. S. (2006). Mixed-­ methods research: Positivism dressed in drag?

Journal of Research in Nursing, 11(3), 195–203. [Link]
1744987106064635
Gill, D. L. (2011). Beyond the qualitative-­quantitative dichotomy: Notes from a non-­
qualitative researcher. Qualitative Research in Sport, Exercise and Health, 3(3),
305–312. [Link]
Hanton, S., Wagstaff, C. R., & Fletcher, D. (2012). Cognitive appraisals of stressors
encountered in sport organizations. International Journal of Sport and Exercise
Psychology, 10(4), 276–289. [Link]
682376
Harwood, C., Drew, A., & Knight, C. J. (2010). Parental stressors in professional
youth football academies: A qualitative investigation of specializing stage par-
ents. Qualitative Research in Sport and Exercise, 2(1), 39–55. [Link]
10.1080/19398440903510152
Harwood, C., & Knight, C. (2009a). Stress in youth sport: A developmental investi-
gation of tennis parents. Psychology of Sport & Exercise, 10(4), 447–456. https://
[Link]/10.1016/[Link].2009.01.005
Harwood, C., & Knight, C. (2009b). Understanding parental stressors: An investiga-
tion of British tennis-­parents. Journal of Sport Sciences, 27(4), 339–351. https://
[Link]/10.1080/02640410802603871
Harwood, C. G., Thrower, S. N., Slater, M. J., Didymus, F. F., & Frearson, L. (2019).
Advancing our understanding of psychological stress and coping among parents
in organized youth sport. Frontiers in Psychology, 10, 1–17. [Link]
10.3389/fpsyg.2019.01600
Hayward, F. P. I., Knight, C. J., & Mellalieu, S. D. (2016). A longitudinal examina-
tion of stressors, appraisals, and coping in youth swimming. Psychology of Sport
and Exercise, 29, 56–68. [Link]
Howe, K. R. (1988). Against the quantitative-­qualitative incompatibility thesis or
dogmas die hard. Educational Researcher, 17(8), 10–16. [Link]
013189X017008010
Johnson, R. B., Onwuegbuzie, A. J., & Turner, L. A. (2007). Toward a definition of
mixed methods research. Journal of Mixed Methods Research, 1(2), 112–133.
[Link]
Jones, M. V., Lane, A. M., Bray, S. R., Uphill, M., & Catlin, J. (2005). Development
and validation of the sport emotion questionnaire. Journal of Sport and Exercise
Psychology, 27(4), 407–431. [Link]
Knight, C. J. (2019). Revealing findings in youth sport parenting research. Kinesiology
Review, 8(3), 252–259. [Link]
Krane, V., & Baird, S. M. (2005). Using ethnography in applied sport psychology.
Journal of Applied Sport Psychology, 17(2), 87–107. [Link]
10413200590932371
Lazarus, R. S. (1999). Stress and emotion: A new synthesis. Springer.
Lazarus, R. S. (2000). How emotions influence performance in competitive sports.
The Sport Psychologist, 14(3), 229–252. [Link]
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Springer.
Levy, A., Nicholls, A., Marchant, D., & Polman, R. (2009). Organisational stressors,
coping, and coping effectiveness: A longitudinal study with an elite coach.
 Mixed Methods, Parental Stress, and Coping in Youth Sport 231

International Journal of Sports Science & Coaching, 4(1), 31–45. [Link]

org/10.1260/1747-­9541.4.1.31
Lienhart, N., Nicaise, V., Knight, C. J., & Guillet-­Descas, E. (2020). Understanding
parent stressors and coping experiences in elite sports contexts. Sport, Exercise,
and Performance Psychology, 9(3), 390–404. [Link]
spy0000186
McGannon, K. R., & Schweinbenz, A. N. (2011). Traversing the qualitative-­
quantitative divide using mixed methods: Some reflections and reconciliations for
sport and exercise psychology. Qualitative Research in Sport, Exercise and Health,
3(3), 370–384. [Link]
Moran, A. P., Matthews, J. J., & Kirby, K. (2011). Whatever happened to the third
paradigm? Exploring mixed methods research designs in sport and exercise psy-
chology. Qualitative Research in Sport, Exercise and Health, 3(3), 362–369. https://
[Link]/10.1080/2159676X.2011.607843
Morgan, D. L. (2007). Paradigms lost and pragmatism regained: Methodological
implications of combining qualitative and quantitative methods. Journal of
Mixed Methods Research, 1(1), 48–76. [Link]
Morgan, D. L. (2013). Integrating qualitative and quantitative methods: A pragmatic
approach. Sage.
Morse, J. M. (2003). Principles of mixed methods and multimethod research design.
In A. Tashakkori & C. Teddlie (Eds.), Handbook of mixed methods in social and
behavioral research (pp. 189–208). Sage.
Nicholls, A. R., Polman, R., Levy, A. R., Taylor, J., & Cobley, S. (2007). Stressors,
coping, and coping effectiveness: Gender, type of sport, and skill differences.
Journal of Sports Sciences, 25(13), 1521–1530. [Link]
02640410701230479
Nicholls, A. R., Taylor, N. J., Carroll, S., & Perry, J. L. (2016). The development
of a new sport-­specific classification of coping and a meta-­analysis of the rela-
tionship between different coping strategies and moderators on sporting out-
comes. Frontiers in Psychology, 7, Article 1674. [Link]
2016.01674
Papathomas, A., Petrie, T. A., & Plateau, C. R. (2018). Changes in body image per-
ceptions upon leaving elite sport: The retired female athlete paradox. Sport,
Exercise, and Performance Psychology, 7(1), 30–45. [Link]
10.1037/spy0000111
Ryba, T. V., Wiltshire, G., North, J., & Ronkainen, N. J. (2022). Developing mixed
methods research in sport and exercise psychology: Potential contributions of a
critical realist perspective. International Journal of Sport and Exercise Psychology,
20(1), 147–167. [Link]
Sparkes, A. C. (2015). Developing mixed methods research in sport and exercise psy-
chology: Critical reflections on five points of controversy. Psychology of Sport &
Exercise, 16(3), 49–59. [Link]
Sparkes, A. C., & Smith, B. (2013). Qualitative research methods in sport, exercise and
health: From process to product. Routledge.
Tashakkori, A., & Creswell, J. W. (2007). Exploring the nature of research questions
in mixed methods research. Journal of Mixed Methods Research, 1(3), 207–211.
[Link]
232 Sam N. Thrower et al.

Teddlie, C., & Tashakkori, A. (2009). Foundations of mixed methods research:

Integrating quantitative and qualitative approaches in the social and behavioural
Sciences. Sage.
Thomas, O., Thrower, S. N., Lane, A., & Thomas, J. (2019). Types, sources, and debil-
itating factors of sport confidence in elite early adolescent academy soccer play-
ers. Journal of Applied Sport Psychology, 33(2), 192–217. [Link]
10413200.2019.1630863
Thorne, S. (2020). On the use and abuse of verbatim quotations in qualitative research
reports. Nurse Author & Editor, 30(3), 4–6. [Link]
Thrower, S. N., Spray, C. M., & Harwood, C. G. (2023). Evaluating the “Optimal
Competition Parenting Workshop” using the RE-­AIM Framework: A 4-­year
organizational-­level intervention in British junior tennis. Journal of Sport and
Exercise Psychology, 45(1), 1–14. [Link]
Weed, M. (2009). Research quality considerations for grounded theory research in
sport & exercise psychology. Psychology of Sport and Exercise, 10(5), 502–510.
[Link]
14
FINAL REFLECTIONS
Áine M. Humble and M. Elise Radina

Pulling this book together and working with the chapter authors was both
an exciting and rewarding experience. We feel privileged to have been given
an insider’s view into their work and to have been able to engage with them
to help them really bring their data analysis processes alive for our readers.
In the introduction to our first edition (Radina & Humble, 2019), we
shared observations in five areas about our experience of editing the book.
Below, we revisit four of these areas that we also saw reflected in this
expanded edition. We then discuss some new observations from working on
this expanded edition.

Observations from the First Edition

Our first observation focused on the key works mentioned at the end of
each chapter. In the first edition, we noted the great variety of resources
cited, but also noticed how the only resources repeated across the chapters
were ones about grounded theory. In this expanded edition, we see similar
variety in key resources, likely influenced by our deliberate decision to have
as much methodological variance in the book as possible. Our current
authors have given the readers valuable resources for topics about embod-
ied methodologies, institutional ethnography, New Materialism, narrative
analysis, photovoice, genograms, community-­based participatory action
research, longitudinal qualitative research, and mixed methods research,
just to name some. In terms of repetition across the chapters in this
expanded edition, we saw only two resources mentioned twice. The first
resource was a seminal article about the constant comparative method by

DOI: 10.4324/9781003254041-15
234 Áine M. Humble and M. Elise Radina

Glaser (1965) whereas the second resource was a popular book focused on
coding strategies by Saldaña (2016).
In this review of key works, we notice, again, the predominance of books,
book chapters, and articles from methodologically focused journals such as
the International Journal of Qualitative Methods and the International
Journal of Social Research Methodology. In our opinion, this is a continued
reminder for both novice and experienced qualitative researchers to not limit
their understanding of qualitative methodologies to methods textbooks or
to the brief methods sections in journal articles reporting on empirical
research. This predominance of books, book chapters, and methodologi-
cally-focused journal articles also supports our contention that most journal
articles do not provide the level of detail needed for researchers to get the
guidance necessary to conduct high-­quality qualitative research.
A second observation from the first edition was that many chapter con-
tributors described how important it was to talk about their data analysis
with other people (ethically, of course). Talking through their ideas with
colleagues, whether in a cafeteria, in the hallway, or through emails, helped
them develop clarity about their analysis. We see this theme in this expanded
issue as well. For example, Ferdinands and Raine (Chapter 2) describe how
important it was to talk to others about their approaches to institutional
ethnography, and even though Underhill (Chapter 1) used autoethnography
to reflect on her own experiences, she describes how various people helped
her to understand different components of her analysis.
Our third observation related to the challenge of describing one’s data
analysis and the learning curve that was involved. Not surprisingly, this was
present in this edition as well. We often had to probe authors for more detail
on their analysis, asking for specific examples so that readers could have as
deep an insight as possible into our researchers’ experiences. We would ask
questions such as “What do you mean by that? Can you give a specific exam-
ple?” Asking questions such as these were our attempts to uncover what is
typically hidden. For example, if an author said they struggled with review-
ing summary tables of participants, we asked them what was difficult about
it and how they dealt with it. For those who noted that they did not use a
software program’s intercoder reliability function (more on that later), we
asked them to describe why that was. At times, we also solicited additional
visual aids for the eResource—a feature we felt was essential to help readers
have the experience of “being in the same room” as the researchers.
We recognize the inherent challenges that the broader research context
places on researchers in terms of thinking about how we describe our quali-
tative research. Scholars are socialized by academic journal articles to pro-
vide sound bites about their data analysis instead of in-­depth detail. Some
journals publish guidelines for writing up qualitative findings, which may
 Final Reflections 235

include suggestions about what percentage of an article should be devoted

to methods. For example, Goldberg and Allen (2015) suggest no more than
20% of an article should focus on methods, which includes other important
components such as sampling procedure, data collection, and reflexivity. We
are often not pushed to describe our data analysis or show our work, so to
speak, unless it is a thesis or dissertation. It is not surprising, therefore, that
Ferdinands (Chapter 2) read many dissertations to learn more about “how
to do” institutional ethnography.
Our final observation was about the phrase “themes emerged.” The title
of our book was deliberately chosen to build on the work of Richards
(2014), in which she talked about how data is made, and the work of Dickie
(2003), who has stated that codes and themes do not magically emerge from
data analysis, dropping into our lap as a gift from the sky. We want to
emphasize the hard cognitive work that goes into synthesizing and cohering
tremendous amounts of data. We want to recognize the agency of the
researchers. The phrase “themes emerged” is very ingrained in our scholarly
vernacular, and we occasionally had to remind authors in both editions of
this book not to use this phrase and to find other ways to describe the analy-
sis, such as “we identified themes….” We encourage our readers to avoid
using this phrase and to reframe their processes in a way that makes their
active engagement with the material more obvious.

Observations from This Second Edition

In the remaining section of this concluding chapter, we share six additional
observations from reading the chapters for this expanded book edition.
These observations relate to:

1. the importance of supervisors and course experiences for early-­career

academics,
2. asking questions of data,
3. questioning how qualitative research is conducted,
4. the role that various technologies play in analysis,
5. the different ways that data can be synthesized and
6. experiences with peer review in the publication process.

Early Mentoring Experiences for New Scholars

We deliberately wanted to have our chapter authors represent a broad range
of experiences and, as a result, many of our contributors are early-­career
academics. Relatedly, we observed the importance of supervisors and course
experiences in several chapters. New scholars often have many questions
236 Áine M. Humble and M. Elise Radina

(Alarcão, 2018), such as whether they have completed their analysis or when
they know that they have it “right” (Radina & Sharp, 2019). It was not sur-
prising, therefore, that these more novice researchers discussed the impor-
tance of soliciting and receiving feedback from their supervisors, mentors,
and instructors. For example, Hackett and de Medeiros (Chapter 4) and
Erasga and colleagues (Chapter 5) both noted how their studies developed
out of graduate courses and how, as they progressed in their courses, they
worked closely with their instructors and mentors. Hackett describes how it
was important to draw on de Medeiros’ expertise in narrative analysis to
learn, for example, what to look for in the narratives and to “defend” her
coding choices, which gave her increased confidence in her analysis. In
Chapter 5, Cleofas, the instructor of the course and second author of the
chapter, notes how he emphasized the equal contributions that the PhD stu-
dents and course instructor brought to the analysis and that “his role as a
mentor was to facilitate, evaluate, affirm, and challenge their analytic
insights” (p. 80).

Asking Questions of the Data

Qualitative researchers often ask questions of their data to develop addi-
tional insight. Asking specific questions helps researchers move beyond ini-
tial or surface analyses, and qualitative researchers have pointed out the
helpfulness of this process. In his coding manual, Saldaña (2016) offers gen-
eral questions for researchers as they code, such as “What do I see going on
here?” (p. 22). Another example is Richards and Morse (2013), who identify
questions helpful for specific methodologies. For example, they note that
questions such as “‘What are the characteristics of…?’ ‘What are the types
of…?’ and ‘How does… relate to…?’” (p. 189) can assist in comparative
analysis in ethnographic research.
We saw many examples of this in our chapters and in different ways.

• Underhill (Chapter 1) describes how she brainstormed questions to

“thicken the descriptions and really begin analyzing the forces” (p. 17)
that she saw in various aspects of her analysis, which then led to further
memories that she could investigate.
• Ferdinands and Raine (Chapter 2) describe how they followed the advice
of McCoy (2006) to engage with their transcripts as though they were
conducting an interview with the data itself. They further note how peers
asking Ferdinands how she knew what she knew helped her to under-
stand where the analysis might have been underdeveloped.
• Hackett and de Medeiros (Chapter 4) followed a structured approach to
identify the specific components of their narratives, asking what elements
 Final Reflections 237

of the specific model they were using appeared in their narratives and how
they knew if “this element is truly one element or another one” (p. 65).
• Erasga et al. (Chapter 5) used questions from Wang and Hannes’ (2020)
photovoice analysis guidelines to analyze their photovoice images.
• A consultant on Fort et al.’s study (Chapter 8) was helpful for posing
questions to the team that they had not yet considered.
• Rudes and colleagues (Chapter 9) note how brainstorming research ques-
tions helped identify papers and presentations that could be developed,
and they also identify two key questions all qualitative researchers should
ask of their data, which are “(a) Under what conditions did something
occur? and (b) What are the mechanisms by which that something
occurred?” (p. 152).
• Vogl et al. (Chapter 10) describe a team process whereby team members
would take on a “Devil’s advocate” role in questioning other team mem-
bers’ interpretations. This process took time but was very helpful because
it forced individuals to provide clear rationales for their interpretations.
• Grummert and Haslerig (Chapter 11) describe how they used the memo
function in Dedoose to keep track of questions about their codes.
• To begin team meetings, Shannon-­Baker and colleagues (Chapter 12)
would ask general questions such as “What did you see in the transcripts?”
(p. 203), which helped in collaboratively creating codes.

Asking Questions about Qualitative Research Practices

We observed our chapter contributors asking questions about the way quali-
tative research is carried out. This observation was reflected in how the
researchers questioned (a) coding reliability tools, (b) the ways in which
methodologies were conducted, and (c) various terms used to describe data
analysis.
Three authors questioned the usefulness and appropriateness of using
intercoder/rater reliability functions in software programs, declining to use
them. Shannon-­Baker and colleagues (Chapter 12) describe how they chose
to talk through their coding to gain consensus rather than implement a
quantified measure of intercoder reliability. Rudes et al. (Chapter 9) describe
how the team calculated reliability factors “manually” rather than using
[Link]’s intercoder reliability function, which they felt was too restrictive.
This is a relevant concern, given that some qualitative data analysis software
(QDAS) program functions may be overly sensitive to coding differences.
For example, some QDAS programs may report that there is inconsistency
in coding if coding selections differ by just one punctuation mark (O’Connor
& Joffe, 2020). Grummert and Haslerig (Chapter 11) specifically use the
term interrater reliability instead of intercoder reliability to emphasize the
238 Áine M. Humble and M. Elise Radina

positivistic underpinnings of the concept (see Morse, 1997). Relatedly,

Edwards-­Bianchi (Chapter 7) notes how she responded to a journal reviewer
who asked why she did not have a second coder, arguing successfully that
prioritizing intercoder agreement might have silenced the Black women’s
voices in her study.
Readers may want to become familiar with discussions on this topic. For
example, Clark (personal communication, October 3, 2019, as cited in
Morse, 2020) notes that a focus on intercoder reliability can result in shallow
and invalid analyses:

Multiple coders and inter-­coder agreement for reliability puts agreement

about mnemonic tags for data ahead of thoughtful discussion of the
import and nuanced meaning of those data. If we’re haggling over titles
of tags that label those data, we’re sacrificing rich interpretation for tidy
classification.
(p. 3)

O’Connor and Joffe (2020) also summarize the debates about using inter-
coder reliability calculations in qualitative research.
Many of our chapter contributors noted that there were different
approaches to carrying out a particular analysis. On the one hand, this could
be a broader discussion about methodology. Hackett and de Medeiros
(Chapter 4) distinguish between a structured and unstructured approach to
narrative inquiry, choosing to use the more structured approach. Ferdinands
and Raine (Chapter 2) note that there are different ways of carrying out
institutional ethnography. Erasga and colleagues (Chapter 5) discuss differ-
ent ways of using photovoice. Boone Blanchard et al. (Chapter 6) identify
different ways of carrying out phenomenology. Rudes et al. (Chapter 9)
write about how they used a semi-­grounded theory approach instead of a
traditional approach. And both Shannon-­Baker et al. (Chapter 12) and
Thrower et al. (Chapter 13) apply a different focus in their mixed methods
research—the former prioritizing the qualitative component and the latter
prioritizing the quantitative component.
On the other hand, questioning sometimes related to more specific data
analysis processes, whether it was being flexible in using a QDAS program,
such as Thrower and colleagues (Chapter 13) who moved their analysis
from NVivo to Excel at one point, or arguing why it is important to code
directly to video rather than to video transcripts (Grummert and Haslerig,
Chapter 11).
We also noticed some of our contributors questioning terms typically
used by others. For example, in new materialistic methodology, Schadler
(Chapter 3) uses the terms tagging and boundaries instead of coding and
 Final Reflections 239

codes, the latter of which she sees as too positivistic and interpretive in their
intent. Ferdinands and Raine (Chapter 2) deliberately invoke Dorothy
Smith’s language of informant from institutional ethnography rather than
participant.
We believe it is a positive sign that our contributors are questioning
approaches to qualitative and mixed methods research. Scholars will benefit
from familiarizing themselves with different methodological approaches and
the epistemological underpinnings of various data analysis approaches.
Moreover, questioning the way things are carried out is a natural reflection of
the dynamic and ever-­changing nature of qualitative research (Morse, 2020).

Use of Various Technologies

Technology plays a central role in our qualitative researcher’s experiences,
whether it is using Microsoft Word documents, QDAS programs, or other
technologies. Many QDAS programs are on the market (see also the
CAQDASNetworkingProject,[Link]
data-­analysis), and the following programs were used by our contributors:

1. [Link]: Schadler (Chapter 3, before switching to MAXQDA), Fort

et al. (Chapter 8), Rudes et al. (Chapter 9), and Vogl et al. (Chapter 10,
before switching to MAXQDA)
2. Dedoose: Hackett and de Medeiros (Chapter 4) and Grummert and
Haslerig (Chapter 11)
3. MAXQDA: Schadler (Chapter 3, switching from [Link]) and Vogl
et al. (Chapter 10, switching from [Link])

Although the underlying structures of QDAS programs differ, they typically

perform many similar functions related to coding, memoing, diagramming,
and so on (di Gregorio, 2011). Moreover, software developers are watching
each other to see emerging functions that they can also produce in their own
programs (di Gregorio, 2011). Some differences do exist among the pro-
grams, such as Dedoose being web-based with the ability for people to work
simultaneously on the analysis.
What is not typical is for researchers to talk about how a QDAS program
is used (Humble, 2012, 2023), and we encouraged our contributors to give us
more detail about this, such as if they used a querying, memoing, or inter-
coder reliability function. In some cases, we also asked them to provide a
rationale for why they used or switched to a particular program or for why
they did not use a particular function (e.g., intercoder reliability calcula-
tions). Our questioning was not meant to suggest that there was one correct
way of using a program. Instead, it was meant to bring to the surface the
240 Áine M. Humble and M. Elise Radina

reasoning behind certain choices. Sometimes, it feels right for researchers to

temporarily move away from a QDAS program as well, such as Fort et al.
(Chapter 8) did when examining hand-­drawn genograms and eco-maps to
look for patterns. Likewise, Rudes et al. (Chapter 9) used [Link] for some
of their coding but notes that they coded their policies “by hand.” In other
cases, researchers using QDAS programs can export out content in other
formats for examination, such as Vogl et al. (Chapter 10) using an export
function in MAXQDA to export retrieved segments into a Word document
for further analysis. Thinking intentionally about software usage is impor-
tant in qualitative data analysis (Gilbert et al., 2013; Humble, 2012, 2019),
and using a different approach (i.e., stepping away from a software program)
to think about one’s analysis can sometimes result in helpful, alternative
insights into one’s data (Gibbs, 2013; Humble, 2009).
Of course, QDAS programs are not the only technologies used in qualitative
research. Microsoft Word can be used in various ways, such as using comment
boxes to denote codes (Ferdinands & Raine, Chapter 2; Fort et al., Chapter 8).
Several chapters used spreadsheet programs. For example, Ferdinands and
Raine (Chapter 2) used Microsoft Excel to organize their indices and quotes,
Thrower et al. (Chapter 13) used a pivot table option in Microsoft Excel to
quantify qualitative data, and Erasga et al. (Chapter 5) used Google Sheets.
Communication platforms such as Zoom, Teams, and Facebook messenger
also helped team members work from different locations.
In some cases, researchers use more than one software program or technol-
ogy. For example, Rudes et al. (Chapter 9) used [Link] and Microsoft Excel,
and Grummet and Haslerig (Chapter 11) used Dedoose and Microsoft Excel.
Free cloud-­based programs and instant messaging programs were particularly
helpful to Erasga et al. (Chapter 5), who used Google Drive, Google Docs,
Google Sheets, and Facebook messenger to carry out their photovoice study
during the pandemic. The COVID-­19 pandemic intensified use of communica-
tion technologies, as we expect to see this continue in the future.
Technology is helping qualitative researchers in many ways, but, as noted
earlier, it is still important to think critically about how it is being used and
to describe how it is being used, particularly with QDAS programs (Humble,
2019). Transparency about the role of technology in qualitative analysis will
be particularly important in the future with the fast-­developing advance-
ment of generative artificial intelligence (AI) technologies. Such technologies
are being incorporated into QDAS programs as of 2023, although qualita-
tive researchers have already been using AI in other ways (Christou, 2023).
It used to be that those writing about QDAS emphasized how the researcher
still controlled the interpretation of their analysis when using QDAS programs
(see, for example, di Gregorio, 2011; Humble, 2012, 2019; Richards, 2021;
Silver, 2018). That “control” may be questioned with the advancement
 Final Reflections 241

of AI functions that can quickly generate code names, summaries, and inter-
pretations. Moreover, in the quest to speed up our analyses and rid ourselves of
the more mundane aspects of qualitative data analysis, we lose the benefit of
prolonged and sustained engagement with our data in which we “conceptualize
our data deeply” (Morse, 2020, p. 2). How researchers respond to technological
changes is a key challenge for the future (Silver, 2018).

Ways of Synthesizing Data

One of the central tasks of qualitative researchers is to synthesize large
amounts of data so that it can be more clearly understood, and qualitative
methodologists such as Miles et al. (2019) have written about this task.
Additionally, in our previous edition, three of our authors carried out a
“dialogue” on this topic (Castellanos et al., 2019), noting how creating tables
and diagrams helped them develop deeper understandings or alternative
views of their data. Our new contributors continue to show how such syn-
thesis is vital to their data analysis processes. They describe these processes
in their chapters and provide examples of them in the eResource.
Several authors describe the process of creating case summaries, whether
that is about individuals, couples, families, or organizations. Ferdinands and
Raine (Chapter 2) created summaries of participants and Boone Blanchard
and colleagues (Chapter 6) created both couple and family summaries.
Creating summaries of individuals and couples was also key for Vogl and
colleagues (Chapter 10), who were dealing with large amounts of data from
their longitudinal study. Their summaries included socio-­demographic data
as well as other specific details about each individual and couple as they
related to jobs, parental leaves, and other characteristics. Additionally, they
describe how they revised their summaries over time, demonstrating the
malleability required of qualitative researchers, noting that their “different
lines of comparisons between cases, couples, and across time fostered abduc-
tive thinking” (p. 167).
Visual images serve many functions in qualitative analysis (Miles et al.,
2019; Richards, 2021). For example, “diagrams, pictures, images, photo-
graphs, conceptual maps, matrices, tables and charts not only serve as visual
representations of what is being discovered through analysis but also as gen-
erative/analytical techniques and communicative tools” (Buckley & Waring,
2013, pp. 148–149). Braun and Clarke (2022) note how visuals can be par-
ticularly helpful when exploring themes or considering ways to present an
overall story. Whiteboards provide an immediate way to visualize data and
modify interpreted relationships in real time with colleague input. Ferdinands
and Raine (Chapter 2) used whiteboards to carry out the process of map-
ping in their institutional ethnography. Fort et al. (Chapter 8) focus on the
242 Áine M. Humble and M. Elise Radina

visuals they developed to understand their data: eco-maps, genograms, and

timelines. Rudes et al. (Chapter 9) note how the creation of numerous dia-
grams, including flow charts, helped them make sense of their complex data.
Shannon-­Baker et al. (Chapter 12) provide several examples of their attempts
to integrate their qualitative understandings with supplemental quantitative
data in a joint display.

The Publishing Process

Research results should be disseminated, and several of our chapter contrib-
utors discussed their process of getting their work published. The first step,
of course, in getting a qualitative (or mixed methods) study published is to
write a high-­quality article that demonstrates interpretive insight into a topic
(Clark et al., 2022). Some scholars provide suggestions on how to write good
qualitative studies, such as various editorials in health and business journals
(e.g., Morse, 2021; Reay, 2014) and publications by other scholars (e.g.,
Goldberg & Allen, 2015, in the field of family science). Checklists have also
been created, such as one by Tong et al. (2007). But checklists are problematic
for several reasons, such as the assumption that a general checklist can be
meaningfully applied to all types of qualitative research and that adherence
to the checklist ensures a good, insightful qualitative study (see Morse, 2020).
Once submitted, an article is typically blind peer reviewed, a process seen
as contributing to higher-­quality publications and integral to academia (Ali
& Watson, 2016), and it is more common to have a paper accepted after at
least one revision (i.e., a revise and resubmit) than to have a manuscript
accepted without any revisions (Milardo, 2015). Boone Blanchard et al.
(Chapter 6) describe the lengthy process of getting their paper published,
consistent with Milardo’s (2015) note that perseverance is necessary in get-
ting published and that many scholars eventually find a “home” for their
work after it has been rejected by other journals. Qualitative researchers may
also experience additional challenges in the peer review process, some of
which can be due to reviewers with positivistic epistemological leanings or
lack of understanding about a particular qualitative methodology. When
this happens, qualitative researchers are called upon to defend or provide
rationales for their choices.
Carefully crafted responses to reviewers are vital in the peer review pro-
cess (Milardo, 2015), and we see several examples in this book of how our
contributors had to respond to reviewers’ criticisms of their analysis.
Ferdinands and Raine (Chapter 2) describe how they defended their
approach to institutional ethnography (IE), arguing that they had based
their study design on previous reputable IE studies (and citing examples of
 Final Reflections 243

those). Similarly, Edwards-­Bianchi (Chapter 7) drew on the work of Morse

(1997), a very well-­known qualitative researcher, to rationalize why she did
not use multiple coders. Shannon-­Baker and colleagues (Chapter 12) recom-
mend to readers that they be familiar with reporting standards for mixed
methods research in case they need to mention that in responses to review-
ers. And both Schadler (Chapter 3) and Shannon-­Baker et al. (Chapter 12)
remind us that researchers can always make requests to editors for accom-
modations such as (a) extra words or pages for the additional detail that is
beneficial for qualitative research findings, (b) posting content to online
supplemental material, or (c) requesting that online versions of diagrams or
images be in full color.

Conclusion
In this concluding chapter, we have briefly described some processes that we
saw in our contributors’ data analysis experiences. These processes relate to
the value of (a) mentorship, (b) asking questions of data, (c) asking ques-
tions about how things are “typically done,” (d) QDAS and other technolo-
gies, (e) synthesizing data, and (f) learning from people’s publication process
experiences.
Qualitative research has so much to offer scholars and practitioners, and
there is also much to understand about the background work that brings
researchers’ insights to the surface. Insights (or themes) do not magically
“emerge” from the data analysis in a sudden “a-­ha” moment (although there
may certainly be a-­ha moments that occur in analysis). Qualitative researchers
engage in complex and detailed processes to make sense out of large amounts
of data, and we hope this book has highlighted the agency and the hard work
that underpins the excellent qualitative work that we see published.
Again, we thank our contributors for their hard work in helping us to
understand their data analysis processes. We recognize the inherent difficulty
in trying to present an often nonlinear, complicated process in a linear man-
ner. Five of our chapter contributors commented on this, such as Grummert
and Haslerig (Chapter 11), who stated, “The process of data analysis was
iterative, introspective, and interactive. As such, though we must write about
it linearly, it was anything but [that]” (p. 187).

References
Alarcão, I. (2018). “Dilemmas” of early career researchers from “dilemmas” to prob-
lems. In A. Pedro Costa, L. Paulo Reis, F. Neri de Souza, & A. Moreira (Eds.),
Computer supported qualitative research (Second International Symposium on
Qualitative Research (ISQR 2017)) (pp. 28–38). Springer.
244 Áine M. Humble and M. Elise Radina

Ali, P. A., & Watson, R. (2016). Peer review and the publication process. Nursing
Open, 3(4), 193–202. [Link]
Braun, V., & Clarke, V. (2022). Thematic analysis: A practical guide. Sage.
Buckley, C. A., & Waring, M. J. (2013). Using diagrams to support the research pro-
cess: Examples from grounded theory. Qualitative Research, 13(2), 148–172.
[Link]
Castellanos, D., Ball, M., & Zvonkovic, A. (2019). Dialogue about data display. In A.
M. Humble & M. E. Radina (Eds.), How qualitative data analysis happens: Moving
beyond “themes emerged” (pp. 226–230). Taylor & Francis. [Link]
10.4324/9781315171647-­17
Christou, P. (2023). A critical perspective over whether and how to acknowledge the
use of artificial intelligence (AI) in qualitative studies. The Qualitative Report,
23(7), Article 10. [Link]
Clark, D. A., Brook, R. K., & Doney, E. (2022). Recommendations for improving
peer review of qualitative research in conservation journals. Conservation Biology,
36(1), e13823. [Link]
Dickie, V. A. (2003). Data analysis in qualitative research: A plea for sharing the
magic and the effort. The American Journal of Occupational Therapy, 57, 49–56.
[Link]
di Gregorio, S. (2011). Comment: KWALON conference: Is qualitative software
really comparable? Reflections on “the experiment”: An “expert” view. Forum
Qualitative Sozialforschung/Forum: Qualitative Social Research, 12(1). [Link]
org/10.17169/fqs-­12.1.1635
Gibbs, G. R. (2013). Using software in qualitative analysis. In Uwe Flick (Ed.), The
SAGE handbook of qualitative data analysis (pp. 277–294). Sage.
Gilbert, L. S., Jackson, K., & di Gregorio, S. (2013). Tools for analyzing qualitative
data: The history and relevance of qualitative data analysis software. In J. M.
Spector, M. D. Merrill, J. Elen, & M. J. Bishop (Eds.), Handbook of research on
educational communications and technology (pp. 221–236). Springer.
Glaser, B. G. (1965). The constant comparative method of qualitative analysis. Social
Problems, 12(4), 436–445. [Link]
Goldberg, A. E., & Allen, K. R. (2015). Communicating qualitative research: Some
practical guideposts for scholars. Journal of Marriage and Family, 77(1), 3–22.
[Link]
Humble, Á. M. (2009). Technique triangulation for validation in directed content
analysis. International Journal of Qualitative Methods, 8(3), 34–51. [Link]
org/10.1177/160940690900800305
Humble, Á. M. (2012). Qualitative data analysis software: A call for understanding,
detail, intentionality, and thoughtfulness. Journal of Family Theory & Review,
4(2), 122–137. [Link]
Humble, Á. M. (2019). Computer-­aided qualitative analysis software. In P. Atkinson,
S. Delamont, A. Cernat, J. W. Sakshaug, & R. A. Williams (Eds.), SAGE research
methods foundations. Sage. [Link]
Humble, Á. M. (2023, November 8–11). Qualitative research and use of qualitative
data analysis software (QDAS) increasing in family science research: A decade
review. [Poster presentation]. National Council on Family Relations 85th Annual
Meeting. [Online].
 Final Reflections 245

McCoy, L. (2006). Keeping the institution in view: Working with interview accounts
of everyday experience. In D. E. Smith (Ed.), Institutional ethnography as practice
(pp. 109–125). Rowman & Littlefield.
Milardo, R. M. (2015). Crafting scholarship in the behavorial and social sciences:
Writing, reviewing, and editing. Routledge. [Link]
Miles, M. B., Huberman, A. M., & Saldaña, J. (2019). Qualitative data analysis: A
methods sourcebook (4th ed.). Sage.
Morse, J. (1997). “Perfectly healthy, but dead”: The myth of inter-­rater reliability.
Qualitative Health Research, 7(4), 445–447. [Link]
700700401
Morse, J. (2020). The changing face of qualitative inquiry. International Journal
of Qualitative Methods, 19, 160940692090993. [Link]
6920909938
Morse, J. M. (2021). The misfit between qualitative research and journal require-
ments. Qualitative Health Research, 31(7), 1371–1373. [Link]
10497323211010725
O’Connor, C., & Joffe, H. (2020). Intercoder reliability in qualitative research:
Debates and practical guidelines. International Journal of Qualitative Methods,
19, 1609406919899220. [Link]
Radina, M. E., & Humble, A. M. (2019). Introduction: Real stories of how this vol-
ume happened. In A. M. Humble & M. E. Radina (Eds.), How qualitative data
analysis happens: Moving beyond “themes emerged” (1st ed., pp. xix–xxix). Taylor
& Francis.
Radina, M. E., & Sharp, E. A. (2019). Being lost and becoming reoriented with the
massive landscape: Students and instructors’ reflections on family science qualita-
tive research methods courses. Family Science Review, 23(1), 54–77. [Link]
[Link]/?s=radina
Reay, T. (2014). Publishing qualitative research. Family Business Review, 27(2), 95–
102. [Link]
Richards, L. (2014). Handling qualitative data: A practical guide (3rd ed.). Sage.
Richards, L. (2021). Handling qualitative data: A practical guide (4th ed.). Sage.
Richards, L., & Morse, J. M. (2013). Readme First for a user’s guide to qualitative
methods (3rd ed.). Sage.
Saldaña, J. (2016). The coding manual for qualitative researchers (3rd ed.). Sage.
Silver, C. (2018). CAQDAS at a crossroads: Choices, controversies and challenges. In
A. Pedro Costa, L. Paulo Reis, F. Neri de Souza, & A. Moreira (Eds.), Computer
supported qualitative research (Second International Symposium on Qualitative
Research (ISQR 2017)) (pp. 1–13). Springer.
Tong, A., Sainsbury, P., & Craig, J. (2007). Consolidated criteria for reporting quali-
tative research (COREQ): A 32-­item checklist for interviews and focus groups.
International Journal for Quality in Health Care, 19(6), 349–357. [Link]
org/10.1093/intqhc/mzm042
Wang, Q., & Hannes, K. (2020). Toward a more comprehensive type of analysis in
photovoice research: The development and illustration of supportive question
matrices for research teams. International Journal of Qualitative Methods, 19,
1–15. [Link]
INDEX

Pages in italics refer to figures and pages in bold refer to tables.

abductive approach 167 145, 147–155; visualizing

abstract, narrative analysis 62 relationships in diabetes and
AC see Administrative Custody hypertension management 131,
Administrative Custody (AC) 143 136–137; writing New Materialist
advocacy, balancing education with, ethnography 48, 50, 52
subtheme 120 audit trail 86, 103
After Death Communication (ADC) Augustine, S. M. 18, 21
60–72; codebook development autoethnography 9–24; confessions
64–65; identifying narrative data 19–23; theoretical shaping 11–14;
62–63; narrative approach 63–70; where to start 10–11; writing 14–19
overview 60–62; rigor and axial coding 152
trustworthiness 70–72
after revision 57 Barad, K. 13, 45–48
Agenda-­Setting 181–183, 188, 195 before revision 57
Allen, K. R. 3, 115, 235 bias, preparation for 101
analysis, IE 31–38; analytic chunks Bisaillon, L. 32–34, 38
35–36; indexing 33–34; listening Black consciousness 113, 116
32–33; mapping 36–37; narrative Black feminist theory: coding and
accounts 35; reading and 115–119; doing research with Black
deliberating 32–33; talking with women 114; engaging
critical friends 37–38 intersectionality 113–114; intercoder/
analysis of variance (ANOVA) 226 interrater reliability 121; navigating
analytic chunks, IE analysis 35–36 analysis process 119–121; overview
analytic drift 39 110–111; positionality statement
ANOVA see analysis of variance 111; sociopolitical context of
assemblage 13 111–112; theoretical perspective
[Link] 237, 239–240; centering 112–113
people in Restricted Housing Units Black Lives Matter (BLM) 111
 Index 247

Black people: interviewing parents couple analysis 163–171

93–106; motherwork analysis COVID-­19, single mothers during:
110–122; in sport media 179–196 employing “following a thread”
blind spots, dismantling 114 79–86; ensuring trustworthiness
BLM see Black Lives Matter 86–89; overview 75; project context
BMI see body mass index 76–77; research team meetings 80,
body mass index (BMI) 28 87; role of participants in shaping
boundaries, New Materialism 47 key issue 78–79
bracketing 95–98; epoché process 96; Crenshaw, K. 113, 150
reflexive 96–98 critical friends, talking with, IE analysis
Braun 80 37–38
Breuer, F. 6 Cronin, A. 79, 83–85
Britzman, D. 20 cultural socialization 101
Bronfenbrenner, U. 89 Cuthbert, D. 6
Burris, M. A. 75
DC see Disciplinary Custody
centering 157; major concerns 146–156; Dedoose 64, 179, 237, 239; after death
overview 142; RHU study context communication 62, 64–68; media
142–143; studies 143–146 priming in sports media 179, 184,
Chadwick, R. 16 186–196
Changing the (w)Hole Mind Project 144 deductive approach 63
Charmaz, K. 115–117, 144–148 DeLyser, D. 6
Cho, J. 134 DeVault, M. L. 30
clusters of couples 172 diabetes and hypertension management,
coda, narrative analysis 63 family engagement in: codebook
codebook development 64–65, 130–131 development 130–131; defining key
codes, reflective thematic analysis 80–83 messages 137; drawing on team
coding: agreement 190–191; axial 152; strengths 127–129; initial interview
cycles of 116–119; development of analysis 129–130; reflections and
188–190; first cycle coding 116–118; recommendations 138–139; research
getting started with data analysis team meetings 136–137; small group
115–116; hand-­coding 145–146; in meetings 134–135; study background
vivo 116; line-­by-­line 152; process 126–127; transcript coding 136–137;
116; second cycle coding 115, 118; visual diagrams 132–134
term 33; third cycle coding 115, 119; Dickie, V. A. 2, 235
video data 191–192 direct-­coding video data 186–187
COE see College of Education directed content analysis 94, 101
collaborative analysis 203–205 Disciplinary Custody (DC) 143
College GameDay 179, 182–183 discourses, IE 27–28
College of Education (COE) 201 dissonance, triangulating perspectives
complementary accounts, triangulation 170
171 distinct count 225
complicating action, narrative analysis Dropbox 147
62–63 dual pandemics 112
components, video data 185 dyadic analysis 165–171
connecting integration 209–211
constant comparative process 3, 149, eco-­map 132–134, 138
188–192 ECQ see enhanced community
content analysis: directed 94, 101; quarantine
hierarchical 222–223; interpretive egalitarian parenting 44
182 embodied methodology 16
core category 85 emplotment, narrative analysis 62
248 Index

ending, narrative analysis 63 Hannes, K. 81, 237

enhanced community quarantine (ECQ) Hanton, S. 222
76 head notes 145
epoché process see bracketing hierarchical content analysis 222–223
evaluation, narrative analysis 63 Hierarchical Mapping Technique 60, 63
everyday acts of resistance 113 hierarchical parenting 44
Excel see Microsoft Excel horizontalization process 99
expressive rationales 154, 155 Husserl, Edmund 95–96
Express Scribe (program) 31
ICR see intercoder/interrater reliability
familiarization 187–188; reflective imposter syndrome, IE 38–40
thematic analysis 80 impression management 168–169
Families and Societies study 162–163 inaccessible rights, concentration of
family support, types of 134, 134 reported instances of 153
Federally Qualified Health Center 127 indexing, IE analysis 33–34
Feely, M. 17 index person 129
Filipino mothers, research on see inductive approach 80
COVID-­19, single mothers during injury-­causing impacts 193–195
first cycle coding 116–118 institutional capture 29
first edition, observations from 233–235 institutional ethnography (IE) 26;
First Nations Community analysis 31–38; concepts of 27–28;
HealthSource (FNCH) 126–127, 138 data generation 30–31; developing
first study, SWD: collaborative analysis research problematic 28–29;
203–205; initial quantitative and imposter syndrome 38–40;
qualitative data collection 202–203 positionality for 27
Floyd, George 111 institutions, indexing 33
FNCH see First Nations Community instrumental rationales 154, 155
HealthSource integration 209–211
following a thread, employing 79; intercoder/interrater reliability 121,
creating data repertoire 84–85; 148–150, 190–191, 209, 237–238;
identifying promising threads 83–84; prioritizing Black women’s voices
initial analysis per dataset 80–83; over 121
synthesizing followed threads 85–86 in vivo code 116
forces 17 interpretive content analysis 182–183
framing 182 intersecting relations of ruling,
acknowledging 38–39
Garcia-­Coll, C. G. 105 intersectional consciousness 114
gender studies 44–45 intersectionality, engaging 113–114
Genogram Analytics (program) interviews, reflecting on 129–130
132–133 intra-­acting 47
genogram 132–134
Glaser, B. G. 144, 149, 188, 191–192, Joffe, H. 148, 237–238
195–196, 234 joint display 209–211
Goldberg, A. E. 3, 235, 242
Google Docs 88, 149 kicking back 47
Google Drive 77–78 Knoll, K. R. 20
Google Sheets 78
grand tour question 30 Labov, W. 62–66
Guba, E. G. 86, 102–103 Latino Americans 111
Lesane-­Brown, C. L. 101
half-­life 194 Lincoln, Y. S. 86, 102–103
hand-­coding 145–146 line-­by-­line coding 152
 Index 249

listening, IE analysis 32–33 Societies study 162–163; longitudinal

longitudinal analysis 171–174 analysis 171–174; multiple-­
perspective analysis 165–171;
making data 3 overview 160–161; research 161–162
MANOVA see multivariate analysis multivariate analysis (MANOVA) 225
mapping, IE analysis 36–37 mystical connections 36
Material/Discursive Continuum 17
MAXQDA 48, 165, 239–240; MPQLIs narrative accounts, IE analysis 35
and the transition to parenthood narrative analysis: identifying narrative
165–166, 169–170; writing New data 62–63; narrative approach
Materialist ethnography 48, 53, 55 63–70; overview 60–62; rigor and
McCoy, L. 28, 30, 32–33, 39 trustworthiness in 70–72
McPhail, D. 11 narrative data, identifying 62–63
media priming see priming New Materialism 2, 13–14, 45
memo writing 67, 118, 169, 188 New Materialist Ethnography 44–45;
merging integration 209–211 analysis from New Materialist
Messenger (Facebook) 87 perspective 45–49; paper revision
#MeToo, tracing of self in relation to 56–58; project purposes 49–50;
9–10, 23–24; confessions 19–23; referencing 52–54; starting to write/
theoretical shaping 11–14; where to pausing research 55–56; tagging data
start 10–11; writing 14–19 50–52
Microsoft Excel 34, 147, 153, 164, 179,
223–226 obesity alarmists 27
Microsoft PowerPoint 155 O’Connor, C. 148, 237–238
Miles, M. B. 131, 137, 188, 206, 241 Online Surveys (program) 223
Miller, E. 4 open coding 3
mixed methods research (MMR) and Others: including in autoethnography
parental stress 218, 226–228; process 20–21; understanding 19
analyzing qualitative data 222–226;
data collection 221–222; overview of parental care plans 173
218; parental stress and coping in parental stress and coping, study on:
youth sport 218–228; parental stress analyzing qualitative data 222–226;
and coping study 220–221; points of data collection 221–222; mixed-­
controversy 219–220 methods context 218–220; overview
MMR see mixed methods research 220–221; reflections 226–228
modified van Kaam method 98 Pareto’s principle 153–154
Morse, J. M. 2, 121, 222, 236–239, 241 peer review, dynamics of 88–89
motherwork 110, 113, 116–118 people, indexing 33
Moustakas, C. 95–99 personal reflexive journal 31
MPI see multiple perspective interview perspectives, dissonance in triangulating
MPQLIs see multiple-­perspective 168
qualitative longitudinal interview phenomenology: approaches to 93–94;
multipart concurrent mixed methods transcendental phenomenological
221 design 95–96
multiple-­perspective analysis 165–171 photovoice, remote project: context
multiple perspective interview (MPI) 76–77; doing research during
160–161 pandemic 76; employing “following
multiple-­perspective qualitative a thread” 79–86; ensuring
longitudinal interview (MPQLIs): trustworthiness 86–89; overview 75;
analysis strategies 163–165; role of participants in shaping key
comparison levels 164; concluding issue 78–79
reflections 174–175; Families and pivot table 225
250 Index

polyamorous parents, New Materialist team meetings 203; second study

ethnography on 44; New Materialist 205–212; visual diagrams 210–211
perspective 45–49; paper revision
56–58; project purposes 49–50; racial socialization 100–101;
referencing data 52–54; starting to establishing reliability and validity
write/pausing research 55–56; 102–104; phase two analysis
tagging data 50–52 101–102; rationale for 100–101
poly-­nuclear parenting 44 Rankin, J. 28, 32–33
positionality statement 111 reading and deliberating, IE analysis
post-­positivistic perspective 221 32–33
PowerPoint see Microsoft PowerPoint rebuilding worlds, New Materialism
priming 181–183; see also sport media, 47–49
racialized production decisions in reconceptualization, struggling with
prison housing residency see Restricted 19–20
Housing Units referencing data 52–54
process code 116 referencing, New Materialism 48–49
publication, journey of 104–106, reflexive bracketing 96–98
104–105 reflective thematic analysis 80
reliability, establishing 102–104
QDAS see qualitative data analysis Rempel, G. R. 138
software repertoire, creating 84–85
qualitative data analysis 1–8; research problematic, developing 28–29
autoethnography 9–24; centering research team meetings 80, 87, 135–137,
142–158; directed content analysis 151, 203
94, 101; final reflections 233–243; Restricted Housing Units (RHUs) 143;
grounded theory 163; hermeneutic analysis strategies 151–156; coding
approach 163; institutional strategies 148–151; constant
ethnography 26–40; interpretative comparative process 149; data
content analysis 182; multiple organization 147–148; intercoder/
perspective qualitative longitudinal interrater reliability 148–150; major
interviews (MPQLIs) 160–176; concerns 146–156; research team
narrative analysis 60–72; New meetings 151; sample or subsample
Materialist Ethnography 44–58; 146–147; studies on 143–146; study
phenomenology 93–106; photovoice context 142–143; visual diagrams
75–90; qualitatively oriented mixed 155–156
methods research 199–214; re-­traditionalization 171
quantitatively oriented mixed reviewer feedback/revisions 21–22,
methods research 218–228; semi-­ 39–40, 56–57, 88–89, 104, 105, 211,
grounded theory 144; sequential 242–243
hermeneutics 163; thematic analysis RHUs see Restricted Housing Units
110–122; video data analysis Richards, L. 2–3, 235–236
179–196; visual diagrams 132–134 rigor and trustworthiness, narrative
qualitative data analysis software analysis 70–72
(QDAS) 237–239 rigor, strategies for determining 103, 103
qualitatively oriented mixed methods Roberto, K. 115, 117
research and students with ruling relations, IE 27
disabilities (SWD): collaborative
analysis 203–205; first study Saldaña, J. 1–2, 39, 115–119, 206–208,
202–205; intercoder/interrater 234, 236
reliability 209; overview 199–201; science, technology, engineering, and
positionalities 201; Qualtrics math (STEM) 200; see also students
(program) 202; reflections and with disabilities
recommendations 212–213; research secondary coding 152
 Index 251

second cycle coding 115, 118 superordinate themes 85

second edition, observations from SWD see students with disabilities
236–243; asking questions of data synthetization, threads 85–86
236; data synthesis 241–242; early
mentoring experiences for new tagging data 50–52
scholars 235–236; publishing process tagging, New Materialism 47–48
242–243; qualitative research talk, indexing 33
practices 237–239; use of various team, drawing on strengths of 127–129
technologies 239–241 technologies, using 239–241
second study, SWD: challenges to texts, IE 27–28
publishing 211–212; iterative texts, indexing 33
development of visuals 209–211; thematic analysis: coding (first, second,
overview 205–206; qualitative coding and third cycle) 115–119; doing
process by group 207–209; teaching/ research with Black women 114;
learning qualitative coding 206–207 engaging intersectionality 113–114;
sedimented history 48 navigating analysis process 119–122;
Sellers, R. 100–101 overview 110–111; positionality
semi-­grounded theory 144 statement 111; sociopolitical context
sensitizing concept 115 for 111–112; theoretical perspective
sequential explanatory mixed methods 112–113
design 202 “themes emerged,” avoiding saying 235
shells, IE 29 theoretical shaping 11; autoethnography
SHOWeD method 77 12–13; New Materialism 13–14
Smith, Dorothy 28, 33, 36, 223 thick description, narrative analysis 71
socioecological model 89 third cycle coding 115, 119
Sparkes, A. C. 218–226 threads, identifying 83–84
sport media, racialized production time, comparisons across 171–174
decisions in: constant comparative timeline 132–134
process 188–192; data analysis Together Alone Project 144
187–192; data collection 183–184; transcendental phenomenology:
data management 184–187; concepts tied to 95–96; epoché
intercoder/interrater reliability process 96; see also bracketing; first
237–238; layers of data 184; moving author bracketing excerpt 97;
from analysis to findings 192–196; method 98; phase one analysis
racialized production decisions in 98–100; reflexive bracketing
179–196; study origins 180–181; 96–98
theoretical/methodological Trent, A. 134
alignment 181–183 triangulation 2, 98, 103, 147, 161–162,
Spry, T. 12–13, 16–17 168–169, 171
standpoint, IE 27 trustworthiness: and coder agreement
starting points, coding 115–116 190–191; ensuring 86–89, 102–104,
students with disabilities (SWD): 190–191; narrative analysis 70–72
collaborative analysis 203–205; data Turner, Susan 28, 36
collection on 202–203; first study on
202–205; overview of research about UIHO see Urban Indian Health
199–201; positionalities as Organization
collaborative researchers on 201; Urban Indian Health Organization
reflections and recommendations on (UIHO) 126–127
research 212–213; second study
205–212; sequential explanatory validity, establishing 102–104
mixed methods design 202 video data: components of 185;
subthemes, data analysis notes for 120 direct-­coding 186–187
252 Index

video data, analyzing see sport media, concepts 27–28; IE imposter

racialized production decisions in syndrome 38–40; positionality 27
visual data, components of 185 WFH see work-­from-­home
visual diagrams 132–134, 155–156, Whittemore, R. 103
210–211 work-­from-­home (WFH) 76, 79, 83, 85,
vulnerability, complicating 17–18 87, 89
work, IE 27–28
Waletzky, J. 62 working remotely 87–88
Wang, C. 75, 77, 79, 81 writing 14–15; complicating vulnerability
Weaver-­Hightower, M. B. 4, 19 17–18; never-­ending-­ness 22–23;
weight surveillance work, tracing ruling research as easy part in 15–16;
of 26; analysis 31–38; data understanding Others 18–19
generation 30–31; developing
research problematic 28–29; IE Zoom 87, 208