Welcome to the second module in the Overview of Qualitative Research Methods series.

My name is Julie Stoner and I am a Professor at the University of Oklahoma Health Sciences
Center.

This introductory series is based on a graduate course that I developed with Dr. Toby
Hamilton from the OUHSC College of Allied Health. Much of the content was developed by
Dr. Hamilton and is driven by her work as an occupational therapist and qualitative
researcher.

1
In this second module, we will focus on features that distinguish quantitative and
qualitative research projects.

2
Let’s begin to compare features of a quantitative and qualitative research paradigm.

In quantitative research, the paradigm is a positivistic view, focused on finding the truth by
making objective measures and controlling extraneous variables. In contrast, a qualitative
research paradigm is a naturalistic one, that embraces multiple realities, perspectives and
contexts in trying to understand meanings of experiences. These are different paradigms,
like comparing apples to oranges. Let’s consider some differences.

(click)

The philosophy that we follow is that there is a single truth or reality, for example, there is
a true, but unknown, prevalence of smoking among adult males in a particular population,
say, 30% of males smoke in the entire population. Our goal is to collect data on a sample of
adults from the target population and use that sample to derive an estimate of the true
population proportion. We want to use research design and analysis approaches that
minimize error or bias to ensure that we generate an estimate that is “right” or as close to
the truth as possible. We aim to use objective methods and avoid subjectivity that may
lead to bias or measurement error.

As a model, we view health in terms of biological and physiological factors – features that

3
we can measure, such as blood pressure, age, and amount of moderate physical activity per
day, as examples.
(click)

In contrast, in qualitative research, we recognize multiple truths and multiple realities. There
is value in considering multiple perspectives to understand factors that influence health
outcomes. We want to understand relations among different realities. There is no “bias” in
qualitative research because there is no single “truth”.

In qualitative settings, subjectivity is accepted and honored. We don’t have objective

measures using instruments or assays. Instead, typically data collection involves interviews,
discussions, drawings, or direct observations.

In terms of models, we view health as a humanistic discipline.

3
The approach to inference differs between the quantitative and qualitative approaches.

(click)

In a quantitative setting, we collect samples from the target population of interest and use
the sample to make inference to the population at large or infer generalizations to future
situations, for example, in the setting of predicting health outcomes for a given patient.
We consider physical, biologic, psychologic, and environmental causes of disease or health
status.

(click)

In contrast in a qualitative setting, we are interested in how humans experience health and
illness and the meanings they attribute to these experiences. We have a broader view of
factors impacting health status, including social determinates of health and how social,
political, and economic contexts influence human experiences and behaviors.

4
 Deductive Reasoning Inductive Reasoning
Quantitative Methods Qualitative Methods

Assumes positivistic paradigm Assumes naturalistic paradigm

(one reality) (multiple realities)

Begins with theory and uses Begins with collecting lots of data
experimental design based on accounting for the top‐heavy
testing one hypothesis of the triangle
theory
Analyzes and interprets the data
Tests the hypothesis by collecting to come up with findings and
lots of data and findings – the theory – note that the data are
triangle broadens to represent reduced, not broadened, to its
lots of data conclusions

Results expressed in numerical Used to develop theory, not test it

statistical quantitative findings
that can be generalized to lots of Results expressed in organized
people or contexts and accounts word‐based findings specific to
for the breadth of the bottom of those studied
the triangle 5
We also recognize that quantitative and qualitative methods use
different reasoning approaches.

(click)

In quantitative research, we use a deductive reasoning approach. We

begin with a theory, for example, based on basic science animal
research, we have a theory that exposure to high levels of sugar, or
glucose, in the blood leads to cell and vascular damage. Based on
this theory, we would develop a hypothesis that an intensive
treatment protocol that targets maintenance of low blood sugar
levels will result in a lower risk of vascular damage leading to heart
disease, eye disease, and kidney damage among patients with Type 2
diabetes. We would then design a randomized controlled clinical trial
comparing health outcomes between participants randomized to an
intensive treatment approach to a standard of care approach. We
would then collect data to test the hypothesis and make inference to
the target population of patients with Type 2 Diabetes.

(click)

In contrast, in the qualitative research setting, we do not begin with a

5
hypothesis or theory. Instead, we begin by collecting a large amount
of rich data related to how humans experience health and illness and
the meanings they attribute to these experiences. We collect data
on factors impacting health status, including social determinates of
health and how social, political, and economic contexts influence
human experiences and behaviors. Then, from the data, which often
takes the form of interview transcripts or focus group transcripts, we
identify themes across the participants and then based on these
findings, we develop theories regarding factors that influence health
outcomes, or barriers and facilitators to recommended treatments or
health behaviors.

Citation: Forman, J., Creswell, J., Damschroder, L., Kowalski, C., &
Krein, S. (2008). Qualitative research methods: key features and
insights gained from use in infection prevention research. American
Journal of Infection Control. , 36( 10), 764‐771.

5
This slide helps explain some differences between deductive and inductive reasoning and
how to use them to our best advantage in research.

It is important to keep in mind that the research question determines the methodology
that we use in research and program evaluation.

(click)

In a quantitative setting, we use deductive reasoning, using a step‐wise progression

beginning with theories that lead to hypotheses. We then collect a sample of data to test
the hypotheses and make inference to the larger population. We use this approach to
answer questions related to “How much?” and “Whether?”.

(click)

In contrast a qualitative study uses inductive reasoning. The approach is iterative in the
design and conduct, as well as the process of identifying themes from the interviews and
discussions. Based on the themes, we develop theories and identify factors or variables
that influence health and behaviors. In our sampling, we don’t aim for a “random” or
“representative” sample, but instead target information‐rich participants using purposive

6
sampling. These are individuals who have experiences that we are interested in, for
example, patients who stop adhering to physician recommendations. Research questions
are holistic in nature and answer questions like “What?”, “How?”, and “Why?”.

(click)

Thinking of the triangles of reasoning in the previous slide, how could you use both
quantitative and qualitative methods in a single or successive study? How could you organize
the pyramids to explore new questions? How could you organize the pyramids to explain
results?

6
Let’s return to the literature example that we introduced in the first module. The investigators were
interested in understanding the experience of low‐socioeconomic status patients during the post‐hospital
transition period.

How does this project illustrate the naturalistic paradigm?

How does this study illustrate the naturalistic paradigm?

(click on each)
What are the multiple realities represented and how is their subjectivity honored in
the populations?
How are the participants’ experiences and meanings attributed and reported?
How did the researchers come up with findings represented in Figure 1?
How is the study’s aim or objective stated and how does the design reflect it?
How would the findings lead to subsequent studies and translation?

• Multiple realities include those of the patients with low SES and the health care providers
offering discharge instructions. The participants met inclusion criteria and were not chosen
randomly, but purposively – on purpose – as experts on the topic of interest because of their

7
 subjectivity. A random selection of participants would not have met the inclusion criteria. Table 1
outlines their characteristics.

• The participants’ experiences and meanings in response to interview questions on how they
experienced and made sense of the discharge process. They are reported in 6 themes that
“emerged” from the data and a priori coding. Researchers came up with the themes by repeated
readings of the interview data and reported with illustrations of relevant quotes. Note that the
data are verbatim quotes that are always paired with the speaker’s key inclusion characteristics
such as sex, age, and insurance status. The naturalistic paradigm is interested in the social
determinants of health and health behavior rather than statistically inferred and generalizable
population‐based epidemiology that cannot address the research question.

• The objective was to study the perceptions of patients of low‐SES to identify common experiences
during the transition process. The stated methodology is a modified grounded theory (theory that
is built on or “grounded in” the data) and usually represented by a model as in Figure 1. The text
of figure 1 explains the findings as a theory resulting in the generation of new hypotheses that
could be tested using quantitative methods in subsequent research.

Reference: Kangovi, S., Barg, F.K., Carter, T., Levy, K., Sellman, J., Long, J.A. and Grande, D.,
2014. Challenges faced by patients with low socioeconomic status during the post‐hospital
transition. Journal of General Internal Medicine, 29(2), pp.283‐289.

7
Often attributed to Einstein: Not everything that counts can be counted, and not
everything that can be counted counts.

(click)

In a quantitative setting, we make inference from a sample of data to the population at

large and focus on causal relationships between exposures and outcomes.

(click)

In a qualitative setting, we identify themes and patterns relevant to specific contexts

that may be transferable to other settings. We focus on the complexity of human
interactions and influence of contextual factors (historical, social, political, or
cultural factors).

(click)

• Note that cause and effect and generalization are not goals of the naturalistic paradigm.
Only experimental designs with randomization and controls can accomplish causal

8
 inference. We will talk more about transferability of results from qualitative studies later.

• Note that the naturalistic paradigm searches for patterns attached to specific contexts
about complex human and contextual issues.

8
In a qualitative setting, we cannot generalize findings to populations, but instead, recognize
complexity and contextual issues.

(click)

In a quantitative setting, we focus on causal relations linking exposures to outcomes. We

have few variables and large sample sizes that inform the generalizations and inference
that we draw.

(click)

In a qualitative setting, we recognize the complexity of human interactions and

influence of contextual factors (historical, social, political, or cultural factors). We
collect information from a small number of participants, but collect a lot of
information, many variables, from each participant.

(click)

9
• Again note the naturalistic paradigm’s emphasis on complexity and contextual issues.

• Especially note the effect on variables and numbers when the study does not infer
generalization from a representative sample to a population or require statistical power
requiring large numbers.
The relationship is an inverse one
• Studies in the positivistic paradigm using quantitative methods require study of a
limited number of well‐controlled dependent variables in large numbers to have
the statistical power to make inferences.
• Studies in the naturalistic paradigm may be exploring previously unknown
variables that cannot be limited by design. Qualitative studies result in massive
amounts of data and so are limited by resources to a few cases. The number of
participants can be small because results are not quantified and findings are not
expected to be generalized. Instead, results are context‐specific.

9
But can you really generalize with quantitative methods?

Consider the limitations of the implied reductionism. One factor is that disease,
diagnosis, or pathology is not the same as the illness, which is the patient’s and
family’s experience. One person’s disease affects many others.

Think of the phrase “Everybody has his own stroke” (etc.) in your field of interest.
What does it mean? Can we really assume that everyone with a given diagnosis has
similar physiology, needs, goals, functional life challenges as a result of the
impairment?

Factors such as sociocultural, economic factors, family roles & expectations, personal
priorities & values, meaning of the injury, and weather all are contextual factors that
may influence experience and health status (Hammel & Carpenter, 2004, p. 3)

Reference: Hammell, K.W. and Carpenter, C. eds., 2004. Qualitative research in

evidence‐based rehabilitation. Elsevier Health Sciences.

10
Where do research results from come from in quantitative and qualitative studies?

(click)

In quantitative studies, results are represented by data that are numerical, quantitative,
and are collected in a database of measurable outcomes. The data are analyzed
using statistical methods. Rigor relates to internal validity (minimizing bias),
external validity (maximizing generalizability), and maximizing reliability (minimizing
variation or error in measurement) with an eye towards achieving statistical
significance.

(click)

In contrast, in qualitative settings, data are more descriptive, including transcripts

from interviews or discussions, quotes, direct observation, or pictures and
drawings. These pieces of information (e.g., quotes) and transcripts can be stored
in a database. The analysis is interpretive, with a goal to identify themes. When we
think of rigor, we think of characteristics of credibility, dependability, confirmability

11
and transferability. This is a new set of terms for those who are more familiar with
quantitative research methods.

(click)

Qualitative studies actually provide raw data in the peer‐reviewed article as quotes under a
theme or in a model so readers can determine where the findings came from. Quantitative
studies rarely present raw data.

What advantages do readers have in seeing raw data? We have a better sense of “where the
findings come from”.

11
When reading a publication or public health program report, what sections do we look for?
What information is important to consider when interpreting results from the study or
evaluation?

With quantitative studies, we review the following:

Research question
Study design
Data collection
Data analysis
Data Interpretation; Findings
Conclusions; Limitations

What information would we look for in a qualitative study?

(click)

The same categories as in a quantitative study.

12
To summarize our discussion, we see that qualitative research offers a variety of methods
for identifying what really matters to patients and their families, community members,
providers, and public health program officials, detecting obstacles and facilitators to
changing performance or behavior, and explaining why improvement does or does not
occur.

Reference: Pope, C., van Royen, P., & Baker, R. (2002). Qualitative methods in
research on healthcare quality. Quality and Safety in Health Care, 11(2), 148‐152.
doi: 10.1136/qhc.11.2.148

13
This module introduced the naturalistic paradigm and philosophy and approach of
qualitative research. We discussed differences in scientific paradigms of quantitative and
qualitative methods. We recognized the importance of identifying an appropriate approach
to address different types of research questions / topics and desired research outcomes or
inference. We recognize that both quantitative and qualitative methods are important for
program evaluation and research.

14
15
16
17
18