Sampling in Qualitative Research

Dr. Rania Albsoul

May 2023
2 Learning objectives

By the end of this lecture, you will be able to:

 Differentiate between sampling in qualitative research and quantitative

research.

 Identify approaches to sampling in qualitative research.

 Identify different types of purposive sampling.

May 2023
 3 Introduction
 A sampling plan is the design for how to specifically choose sources for your data.

 A sampling plan is a formal plan specifying a sampling method, a sample size, and procedure for recruiting
participants.
o Recruitment refers to the process whereby the researcher identifies and invites (recruits) participants to join the
study.
 A qualitative sampling plan describes how many observations, interviews, focus group discussions or cases are
needed to ensure that the findings will contribute rich data.

 In quantitative studies, the sampling plan, including sample size, is determined in detail in beforehand but
qualitative research projects start with a broadly defined sampling plan.

 The sampling plan in qualitative research is appropriate when the selected participants and settings are sufficient
to provide the information needed for a full understanding of the phenomenon under study.

 Good qualitative researchers, at the very least, engage in purposeful sampling, which means that they purposefully
choose data that fit the parameters of the project’s research questions and goals.
May 2023
 Introduction
4

 While quantitative studies often aim to maximise statistical power through the use of as
large a sample size as feasible, qualitative studies usually work with a small number of
cases that are feasible to study in depth.

 The setting, where sampling is carried out, is described in detail to provide thick description
of the context, thereby, enabling the reader to make a transferability judgement.

May 2023
5 Introduction

 Sample sizes for qualitative research vary by technique but are generally
small.

 Qualitative research involves non-probability sampling, where little attempt is

made to generate a representative sample.

 Participants are always sampled deliberately, not at random in qualitative

research.

May 2023
 6 Introduction
 The sampling process in qualitative research is iterative and is expected to continue to develop and
be refined during the research process.
iterative sampling approach whereby the research team moves back and forth (iterating) between
sampling and analysing data such that preliminary analytical findings shape subsequent sampling
choices.

 Analysis and interpretation of data collected after initial sampling feeds back to influence sampling
methods and decisions regarding sample size.

 As the research progresses, and the sampling of additional data yields no further
themes/ideas/concepts on analysis, the point of data ‘saturation’ is reached and sampling can
cease.

 You review the analysis, findings, and the quality of the participant quotes you have collected, and
then decide whether sampling might be ended because of data saturation. In many cases, you will
choose to carry out two or three more interviews or an additional focus group discussion to confirm
that data saturation has been reached

May 2023
7 Some practicalities
 You do not have to interview everyone (in a community, hospital) to get a “good” sample.

 a critical first step is to select settings and situations where you have access to potential
participants.

 Subsequently, the best strategy to apply is to recruit participants who can provide the
richest information. Such participants have to be knowledgeable on the phenomenon and
can articulate and reflect, and are motivated to communicate at length and in depth with
you.

 Finally, you should review the sampling plan regularly and adapt when necessary.

May 2023
 Types of sampling
8

• Probability sampling means that every member of the

population has a chance of being selected. It is mainly
used in quantitative research.

• If you want to produce results that are representative of

the whole population, probability sampling techniques
are the most valid choice.

• In a non-probability sample, individuals are selected

based on non-random criteria, and not every individual
has a chance of being included.

• Non-probability sampling techniques are often used in

qualitative research. In these types of research, the aim is
not to test a hypothesis about a broad population, but to
develop an initial understanding of a small sample of
population. May 2023
9 Approaches to sampling in Qualitative Research

 Approaches to sample selection in qualitative research fall under two broad

categories;

 non-conceptually-driven approaches (convenience and opportunistic

sampling).

 conceptually-driven approaches (purposive and theoretical sampling).

May 2023
10 Sampling strategies in qualitative research

May 2023
11 Convenience sampling

 In this approach, the potential participants/research settings/ that are most easily
accessible to the researcher are sampled.
 “Because they are there”: people closely surrounding you.
 Its advantages are that it is less expensive and time- and effort-intensive.
 Convenience samples are most appropriate when the priorities are speed and low
cost.
 Example, when you simply ask any patient in your clinic who is willing to participate.

May 2023
12 Convenience sampling examples

 a teacher who wanted to examine the perceptions of teachers about a

policy change and decided to utilize a school within the district he or she
worked in to recruit participants.

 a professional who is a member of a professional organization and wanted

to recruit participants through contact information available to members of
that organization.

 Both examples would be convenient to each researcher but would also

require obtaining permissions to recruit participants (from the district and
professional organization respectively).

May 2023
13 Opportunistic (emergent) sampling

 This sampling method involves the researcher taking advantage of

circumstances that occur as the study progresses, taking up emerging
opportunities for data collection along the way.

 This flexible approach lends itself to exploratory field research where little is
known about a phenomenon or research setting.

May 2023
14 Opportunistic (emergent) sampling

 New opportunities to recruit participants or to gain access to a new site may develop after
the fieldwork has begun.

 As the observer gains more knowledge of a setting, he or she can make sampling
decisions that take advantage of events, as they unfold.

 Example

 Interviewing homeless people at a shelter, one man tells you where most of the homeless
people sleep, so you add this site to where you interview.

May 2023
 Purposive sampling
Purposive or judgement sampling is a frequently-applied
conceptually-driven approach.

It involves the researcher deliberately and purposefully

selecting the sample they believe can be the most fruitful
in answering the research question.

This selection process can be guided by consideration of

the variables or qualities of potential participants that
affect the contribution they could provide to the study.

These variables may be simple demographics such as

age, gender and socioeconomic status but can also
include other aspects such as specific attitudes or beliefs.

15 May 2023
16 Example on purposive sampling

 a student who seeks to look at current nurses’ perceptions of leadership

styles within a specific hospital setting.

 This one sentence description alone can already generate two selection
criteria: (a) must be an active nurse and (b) must work at a specific hospital
setting.

 Additional criteria such as number of years in the field or level of nursing

education will ensure participants have a similar foundation.

May 2023
17 Strategies of purposeful sampling

 There are several different strategies for purposefully selecting information-rich cases.

 Maximum variation sampling.

 Homogenous sampling.
 Deviant sampling.
 Typical case sampling.
 Critical case sampling.
 Confirming and disconfirming sampling.
 Stratified purposeful sampling.
 Snowball sampling.

May 2023
 Maximum variation sampling
18
(Heterogenous sampling)
 Entails the recruitment of study participants who vary widely on the dimensions
of interest with the aim of identifying central themes/elements that hold true
across the diverse sample.

 Another definition; researchers access a wide range of data or participants who

will represent wide variations of the phenomena under study.

 This allows for multiple perspectives of individuals to be presented that exemplify

the complexity of the world.

 Any common patterns that emerge from great variation are of particular interest
and value in capturing the core experiences and central, shared aspects or
impacts of a program
May 2023
19 How does one maximize variation in a small
sample?
 One begins by identifying diverse characteristics or criteria for constructing the sample.
 Suppose a state-wide program has project sites spread around the state, some in rural
areas, some in urban areas, and some in suburban areas. The evaluation lacks sufficient
resources to randomly select enough project sites to generalize across the state. The
evaluator can at least be sure that the geographical variation among sites is
represented in the study.

 When selecting a small sample of great diversity, the data collection and analysis will
yield two kinds of findings:
 (1) high-quality, detailed descriptions of each case, which are useful for documenting
uniqueness, and
 (2) important shared patterns that cut across cases and derive their significance from
having emerged out of heterogeneity.

May 2023
20 Maximum variation sampling-Why use this
strategy?
 Often, researchers want to understand how a phenomenon is seen and understood
among different people, in different settings and at different times.

 When using a maximum variation sampling method the researcher selects a small
number of units or cases that maximize the diversity relevant to the research question.

 For example, this strategy was used by Foss and Edson (1989) in their study of women’s
choices about changing their names after marriage. The authors purposefully recruited
three groups of women. Group one included women who adopted their husbands’
names; in group two they kept their birth names; in group three they chose new names.
Sample variation was necessary for illustrating the complex nature of post-marital
naming decisions.

May 2023
21 Homogenous sampling

 In direct contrast to maximum variation sampling is the strategy of picking a small

homogeneous sample.

 aims to select a group of cases with similar backgrounds and experiences,

simplifying analysis and facilitating group interviewing.

 This sampling approach often is used to select focus groups.

May 2023
22 Homogenous Sampling- Why use this method?

 Homogeneous sampling is used when the goal of the research is to understand and
describe a particular group in depth.

May 2023
 Deviant case (extreme instance) sampling
23
 Involves the selection of extreme or outlying cases of the studied phenomenon, such as crises,
exceptions or remarkable failures or successes, in an attempt to glean as much information relevant
to the research question as possible from each case.
 Learning from highly unusual manifestations of the phenomenon of interest.
 For example, in a study of performance of graduate students, a researcher can select the best and
the worse students in class and compare the causes of their performances.
 For example, scholars interested in happiness may choose to interview people who are especially
resilient, energetic, and long-living (Lyubomirsky, 2008), and those interested in crisis sensemaking
may purposefully examine tragic disasters (Weick, 1993).
 Excellent example of extreme group sampling is Angela Browne's (1987) study, When Battered
Women Kill. She conducted in-depth studies of the most extreme cases of domestic violence to
elucidate the phenomenon of battering and abuse. The extreme nature of the cases presented are
what render them so powerful.
 Finding (and even knowing what equates with) “extreme” requires first gathering and then sorting
through a lot of “typical” data (The process of identifying extreme or deviant cases occurs after
some portion of data collection and analysis has been completed).

May 2023
24 Typical case (typical instance) sampling
 focuses on typical/average cases with the aim of building up a profile of a typical case.

 The case is specifically selected because it is not in any way atypical , extreme or
deviant.

 General agreement (consensus) on what constitutes a ‘typical’ case is required for this
approach.

 The researcher should consult several experts in the field of study in order to obtain a
consensus as to what example(s) is typical of the phenomenon and should, therefore, be
studied.

 Another option is to use another sampling technique — like maximum variation sampling
— to identify typical cases prior to choosing cases for your study (Baran, 2016).
May 2023
25 Typical case (typical instance) sampling
 let’s say you were studying violence in schools. The first step would be to list all of
the criteria that define violence for a “typical” school.

 Then you would choose schools that meet that criteria.

 You would want to select schools that are “average” (meeting your selected
criteria) instead of schools with very high or very low violence rates.

May 2023
26 Typical case (typical instance) sampling- Why
use this method?

 Identifying typical cases can help a researcher identify and understand the key
aspects of a phenomenon as they are manifest under ordinary circumstances.

 Providing a case summary of a typical case can be helpful to those not familiar
with a culture or social setting (Helps to give an overview to people with no
background).

May 2023
 Critical case (critical incident
27
sampling) sampling
 selects cases that will produce critical information with maximum generalisability of information
to other cases.
 The process of selecting a small number of important cases - cases that are likely to "yield the
most information and have the greatest impact on the development of knowledge" (Patton,
2001, p. 236).
 A good critical case also permits logical deductions in the form: “If this is (not) valid for this case,
then it is not valid for any (or only a few) cases” (Flyvbjerg, 2011, p. 307)

 Given that the researcher correctly identifies what makes a ‘critical case’, knowledge gained
may be applied to other cases.
 Examples:
 if it happened to so and so then it can happen to anybody, or if so and so passed that exam,
then anybody can pass.
 You want to know how well people understand a new tax law. Ask very educated people -- if
they do not understand it, then probably no one will. Or ask a very uneducated population, if
they understand it, most people will.

May 2023
28 Critical case (critical incident sampling)
sampling
 imagine you are a researcher studying the demise of traditional dinnertime
rituals. You could purposefully choose a critical sample of families who might be
most likely to practice traditional dinnertime rituals (e.g. religious, a stay-at-home
mother). You might find that even these families do not engage in traditional
rituals like saying a family prayer before dinner. In choosing this critical case, you
might be able to play with the claim that, “if dinnertime rituals are fading even in
this critical sample, then such rituals are likely disintegrating among most families.”

 if conservative group adopts new technology, every other group will.

May 2023
29 Critical case (critical incident sampling)
sampling
 Suppose national policymakers want to get local communities involved in making
decisions about how their local program will be run, but they aren't sure that the
communities will understand the complex regulations governing their involvement. The
first critical case is to evaluate the regulations in a community of well-educated
citizens. If they can't understand the regulations, then less-educated people are sure to
find the regulations incomprehensible. Or, conversely, one might consider the critical
case to be a community consisting of people with quite low levels of education: 'If they
can understand the regulations, anyone can.' (Patton 2014: 276).

 In short, choosing a critical sample can help with transferring claims to larger
populations in the long run.

May 2023
30 Critical case (critical incident sampling)
sampling- Why use this method?

 This is a good method to use when funds are limited. Although sampling for one or
more critical cases may not yield findings that are broadly generalizable, they may
allow researchers to develop logical generalizations from the rich evidence
produced when studying a few cases in depth.

 To identify critical cases, the research team needs to able to identify the
dimensions that make a case critical.

May 2023
31 Confirming and disconfirming sampling
 Usually employed in later phases of data collection. Confirmatory cases are additional
examples that fit already emergent patterns to add richness, depth and credibility.
Disconfirming cases act as a means for placing boundaries around confirmed findings.

 involves the selection of a mixture of cases that tie in with expectations or findings up to
that point in the study and cases which deviate from them.

 The confirming cases serve to add depth, detail and enhance credibility while the
disconfirming cases challenge the prevalent narrative and may bring to light alternative
interpretations.

 This approach is generally utilised at later stages of a study when preliminary fieldwork has
already established what qualifies as a ‘confirming case’.

May 2023
32 Stratified purposeful sampling
 Selects participants from specific sub-groups of the population of interest, enabling easier
comparison of the variation across sub-groups.

 Patton (2001) describes these at samples within samples and suggests that purposeful samples can
be stratified or nested by selecting particular units or cases that vary according to a key dimension.

 The purpose of a stratified purposeful sample is to capture major variations rather than to identify a
common core, although the latter may also emerge in the analysis.

 Each of the strata would constitute a fairly homogeneous sample.

 This strategy differs from stratified random sampling used in quantitative research in that the sample
sizes are likely to be too small for generalization or statistical representativeness.
 If you want to study university students, pick a certain number of students from each of the 4 years
(sample of freshmen, sophomores, juniors, and seniors).
 one may purposefully sample primary care practices and stratify this purposeful sample by
practice size (small, medium and large) and practice setting (urban, suburban and rural).
May 2023
33 Snowball sampling (FRIEND OF FRIEND)
 Can also be called as chain sampling.

 involves identification of participants by a technique known as ‘snowballing’ whereby

initially identified participants are asked to suggest other possible candidates.

 Researchers begin by identifying several participants who fit the study’s criteria and then
ask these people to suggest a colleague, a friend, or a family member.
Start with a few respondents and then ask them who else might have _____ or know about
____?

 Find a few diabetic patients and then ask them who else they know that has diabetes.
 This is especially useful when the studied population is hard to access, and/or may not
publicly signal that they belong to the group of interest (e.g. drug-users).
 One downside to snowball samples is that they can quickly skew to one type of group or
demographic (as participants tend to suggest others who are similar to themselves).
 A potential solution is to recruit a handful of participants who represent a maximum
variation,
Mayand
2023 then to generate several smaller snowballs from that diverse initial sample.
34 Theoretical sampling
 Theoretical sampling is an approach where sampling decisions are guided by the theory that
starts to emerge from the collected data.

 The process of data collection for generating theory whereby the analyst jointly collects,
codes, and analyzes his data and decides what data to collect next and where to find
them, in order to develop the theory as it emerges” (Glaser and Strauss, 1967)
 The goal of sampling is to collect data that either further develops or challenges existent
hypotheses. Initial cases selected have similar characteristics and are studied in depth. The
researcher then samples outlying cases to see whether the developing hypothesis ‘holds up’
to these.
 Once no new insights are derived from further data collection, sampling is ceased. This
approach necessitates that data analysis and coding commence while data collection is still
ongoing.

May 2023
35 Theoretical sampling

 In theoretical sampling, the actual number of cases studied is relatively unimportant.

 What is important is the potential of each case to aid the researcher in developing
theoretical insights into the area of social life being studied.
 After completing interviews with several informants, you consciously vary the type of
people interviewed until you have uncovered a broad range of perspectives held by
the people in whom you are interested.

 You would have an idea that you had reached this point when interviews with
additional people yield no genuinely new insights.

May 2023
36 SELECTING INFORMANTS

 Qualitative interviewing calls for a flexible research design.

 The researcher starts out with a general idea of which people to interview and how to
find them, but is willing to change course after the initial interviews.

May 2023
37 SELECTING INFORMANTS

 Those new to qualitative research usually want to know exactly how many
people they need to interview to complete a study. This is a difficult
question to answer prior to conducting some research.

 As Kvale (1996) pointed out: To the common question, “How many

interview subjects do I need?” the answer is simply, “Interview as many
subjects as necessary to find out what you need to know.” (p. 101).

 The size of the sample in an interviewing study is something that should be

determined toward the end of the research and not at the beginning.

May 2023
38 SELECTING INFORMANTS

 Although qualitative researchers generally cannot determine the sample size prior
to conducting a study, people preparing proposals for dissertations or grants are
usually expected to specify the number of informants or settings they intend to
study.

 IRBs might also require this. You should be prepared to indicate your sample size in
proposals, adding that this might change as you start collecting and analysing
data.

May 2023
39 SELECTING INFORMANTS
 Informants can be found in a number of ways.
 on pre-fieldwork, one of the easiest ways to build a pool of informants is snowballing—
getting to know some informants and having them introduce you to others.
 A potential drawback of the snowball technique is that it can limit the diversity of your
informants (Cannon, Higginbotham, & Leung, 1988).
 Therefore you need to be prepared to use a range of different approaches to
identifying people.
 You can locate potential informants through the same sources the participant
observer uses to gain access to private settings: checking with friends, relatives, and
personal contacts; involving yourself with the community of people you want to study;
approaching organizations and agencies; advertising in media sources; and
announcements through the Internet.

May 2023
40 CONCLUSION- FLEXIBILITY IN QUALITATIVE
SAMPLING
 A flexible research and sampling design is an important feature of qualitative
research.

 When little is known about a phenomenon or setting, a priori sampling decisions

can be difficult. In such circumstances, creating a research design that is flexible
enough to foster reflection and preliminary analysis may be a good idea.

May 2023
41 CONCLUSION

May 2023
 42

Thank you

May 2023