STUDY DESIGN
Descriptive/non- Analytical
analytical
Cross-sectional Qualitative Quasi- Non-experimental/
Experimental experimental observational
RCT Cohort
Survey
Community Trials Case-control
Cross-sectional
##A poorly designed study gives misleading results
SELECTING A RESEARCH
DESIGN
FACTORS THAT INFLUENCE THE
SELECTION OF A RESEARCH DESIGN
Nature of the research problem
Purpose of the study
Researcher’s knowledge & experience
Researcher’s interest & motivation
Research ethics & principles
Respondents/participants
Resources
Time
Users of the study findings
Possible control on extraneous variables
RESEARCH APPROACH
1. Qualitative Research
2. Quantitative Research
3. Mixed-Methods Research
Several designs under each research approach
QUANTITATIVE
RESEARCH DESIGNS
Broad Categories Types of Research Design/Main Features
I. Descriptive/Non- 1. Descriptive cross-sectional research design
Analytical studies Accurate description of characteristics of individual, situation, or
group, & the frequency with which a certain phenomenon occurs
without imposing any control or manipulation
Univariant : Studies undertaken to describe the frequency of
occurrence of a phenomenon rather than to study relationship
Exploratory: Investigating the phenomenon & its related factors
about which very little is known
Comparative: Comparing occurrences of a phenomenon in two or
more group
II. Analytical studies 1. Experimental design
Manipulation of independent variable, in the presence of control
group, randomization
2. Quasi-experimental design
Manipulation of independent variable, but absence of either
randomization or control group.
Broad Categories Types of Research Design/Main Features
II. Analytical studies 3. Non-experimental/Observational studies
Examining the relationship between two or more variables without
manipulation or control (cause & effect relationship)
Epidemiological designs
• Cross-sectional study design
• Case-control design
• Cohort study design
Survey study design
Investigation in which self reported data are collected from sample with
the purpose of describing population and associations on some variables
of interest.
• Cross-sectional studies
• Longitudinal studies
III. Other additional 1. Methodological studies
research design Develop, test, & evaluate the research instruments &
method.
2. Meta-analysis/Systematic reviews
Quantitatively combing & integrating the findings of the
multiple research studies on a particular topic.
QUALITATIVE
RESEARCH DESIGNS
Types of qualitative Main features
designs
Phenomenological Phenomenological research examines human
research experiences through the descriptions provided by people
involved.
Ethnographic research Ethnographic research involves the information
collected from certain cultural groups, by living with
people of those groups & from their key informants,
who are believed to be most knowledgeable about the
selected culture.
Case studies Research on a phenomenon by studying in depth a single
case example. The case can be an individual person, an
event, a group, or an institution.
Historical research Systematic collection & objective evaluation of data
related to past occurrences in order to understand these
events that may help to explain present events &
anticipate future events.
MIXED METHODS
RESEARCH DESIGNS
WHAT IS NOT MIXED METHODS RESEARCH?
Simplyusing the name without the rigorous
methods
Having both quantitative and qualitative data
available
Collecting and reporting qualitative and
qualitative data separately without
“combining” them
Using multiple quantitative (e.g experiment,
surveys) or multiple qualitative approaches
(e.g interviews, observations)
WHAT DOES MIXED METHODS RESEARCH INVOLVE?
Collecting and analyzing qualitative and
quantitative data in response to a research
question
Usingrigorous qualitative and quantitative
methods
Combining or integrating qualitative and
qualitative data using a specific mixed
methods design
Framing the mixed methods design within a
broader framework (e.g experiment, theory
or philosophy)
Types of mixed methods Main features
designs
Convergent parallel design Simultaneously collect both quantitative and qualitative data,
–parallel or concurrent merge the data, and use the results to understand a research
problem
Explanatory sequential First collecting quantitative data and then collecting
design – two-phase model qualitative data to help explain or elaborate on the
quantitative results
Exploratory sequential First gathering qualitative data to explore a phenomenon, and
design then collecting quantitative data to explain relationships found
in the qualitative data
Embedded design Collecting quantitative and qualitative data simultaneously or
sequentially, but to have one form of data play a supportive
role to the other form of data.
The two databases may not be easily compared
because the data address different research questions.
Multiphase design Using one of the four designs (convergent, explanatory,
exploratory, or embedded) to examine a problem or a set of
incremental research questions through a series of studies or
separate studies
Transformative design Using one of the four designs (convergent, explanatory,
exploratory, or embedded), but to encase the design within a
transformative framework or lens (feminist, racial, ethnic,
disability, and gay or lesbian) to address a social issue for
marginalized group to bring about change
Convergent mixed methods
Qualitative Data
Quantitative Data
Collection
Collection
Qualitative Data
Quantitative Data
Collection
Analysis
Quantitative Qualitative
Results Results
Interpretation
Quantitative and qualitative results are compared
Explanatory sequential mixed methods
Qualitative Data
Quantitative Data
Collection
Collection
Qualitative Data
Quantitative Data
Collection
Analysis
Quantitative Qualitative
Results Results
Quantitative results are further
Interpretation
explained by qualitative data and results
Exploratory sequential mixed methods
Qualitative Data
Quantitative Data
Collection
Collection
Qualitative Data
Quantitative Data
Collection
Analysis
Quantitative Qualitative
Results Results
Qualitative exploration leading to
Interpretation quantitative test
Quantitative Methods Mixed Methods Qualitative Methods
Nature Pre-determined methods Pre-determined Emergent methods
and emerging
methods
Instrument Closed ended questions Both open- and Open ended
closed- ended questions
questions
Data Performance, attitude, Multiple forms of Interview,
observational and census data drawing on all observation,
data possibilities document and audio-
visual data
Analysis Statistical analysis Statistical and text Text and image
analysis analysis
Interpretation Statistical interpretation Across database Themes, patterns
interpretation and interpretations
SURVEY
RESEARCH DESIGN
WHAT ARE SURVEY RESEARCH DESIGN?
Procedures in quantitative research in which
investigators administer a ‘survey’ to a sample or
entire population
Theaim is to describe the attitudes, opinions,
behaviours, or characteristics of the population
Quantitative, numbered data is collected
They do not involve experimental manipulation of
conditions by the researcher
Mainly focus on learning about a population,
although they often involve correlation of
variables
WHEN DO YOU EMPLOY SURVEYS?
To describe trends
e.g. HIV prevalence in Ghana
To determine individual opinions about policy issues
e.g. NHIS options (fee for service vs capitation)
To identify important beliefs, attitudes and practices
of individuals
e.g. Abortion beliefs, attitudes and practices
May be used to follow up individuals 5, 10,15 years
or more..
e.g. Migration intentions of medical students
TYPES OF SURVEY
DESIGNS
Survey types
Time of Data Collection
Study Over Time Study at One Point in Time
Longitudinal Cross-sectional
Changes in a Changes
Trends in Sub-population
group identified
in the Attitudes
the same Community Program
by a common
same and
population needs evaluation
characteristic people practices
over time
over time over time
Trend Cohort Panel Group National/local
comparisons assessment
SAMPLING IN SURVEY RESEARCH
The Population
is the group of individuals having one
characteristic that distinguishes them from Census
other groups.
The Target Population or Sampling
Frame is the actual list of sampling units
from which the sample is selected.
The Sample is the group of
participants in a study
selected from the target
population from which the
researcher generalizes to the
Random/ target population.
probability
procedures
ENSURING VALIDITY IN SURVEY
RESEARCH
External Validity – extent to which the results can be
generalized to a larger population
Reduce coverage error/bias
Have a good sampling frame list
Reduce sampling error/bias
Select as large a sample from the population as possible
Reduce measurement error (Internal Validity)
Use a good instrument, with clear, unambiguous questions
and response options
Reduce non-response error/bias
Use rigorous administration procedures to achieve as large a
return rate as possible
DATA COLLECTION IN SURVEY
RESEARCH
Who Completes or Records the Data?
Participant Researcher
Questionnaire Interview
Electronic/web-
Mailed Focused
based Individual Telephone
group
WHY WOULD YOU OR WOULD YOU NOT CHOOSE ONE TECHNIQUES
OVER THE OTHER???
STEPS IN CONDUCTING RESEARCH
USING SURVEY DESIGN
STEP 1
• Decide if a Survey Is the Best Design to Use
STEP 2
• Identify the Research Questions or Hypotheses
STEP 3
• Identify the Population, the Sampling Frame, and the Sample
STEP 4
• Determine the Survey Design and Data Collection Procedures
STEP 5
• Develop or Locate an Instrument
STEP 6
• Administer the Instrument
STEP 7 • Analyze the Data to Address the Research Questions or Hypotheses
STEP 8
• Write the Report
QUESTION TIME!!