#18 Methods Consult – Thematic analysis
Episode host: Lara Varpio
Enjoy listening to us at your preferred podcast player.
Methods Consult is an inaugural episode where Lara Varpio dig a bit deeper into the some of the science methods and theory in Health professions education. When you need a little help, or a second opinion, or just some advice from an expert colleague, you might call for a consult. These methods consults are precisely that: a little insight from a colleague who has medical education research experience and (some) expertise.
Thematic analysis
Today we are going to talk about thematic analysis. It is a building blocks of many different methodologies and approaches in qualitative research (e.g., it is part of the Grounded Theory process), but it is also—in and of itself—a method for doing qualitative data analysis.
In this episode, we review thematic analysis as an independent data analysis approach, following the tradition that has been laid out by Braun and Clarke in Thematic Analysis: A Practical Guide (2022).
Thematic analysis is the work the researchers does to construct themes in a data set.
To bridge back to Methods Consult #1 (where we examined the post-positivist and constructivist paradigms), it is important to note that thematic analysis is a way of analysing your data that is flexible. It can be used from a post-positive orientation, or a constructivist orientation, or a post-modern orientation, etc.. It is very important to note that you can do thematic analysis in any of these paradigmatic orientations, but the work of doing thematic analysis will look different depending on your orientation. The markers for rigor will also be different depending on your paradigmatic orientation. Therefore, if you are going to do thematic analysis you must first be clear about the paradigmatic orientation that is underpinning your research.
Today, we will only talk about thematic analysis from a constructivist orientation. This is what Braun and Clarke call “Big Q” thematic analysis (2022).
Definition
Thematic analysis (TA) is “a method for developing, analyzing, and interpreting patterns across a qualitative data set which involves systematic processes of coding to develop themes. Themes are the ultimate analytic purpose.”
Note that TA is a data analysis method, not a methodology. A methodology is theoretical and disciplinary informed way of doing research (e.g., Grounded Theory). Methodologies serve specific research purposes. They come from specific disciplines and so shape how you conduct studies and frame what kind of question you can ask. They guide study design.
TA, as we are talking about it today, is a method. If interviews and focus groups are methods for collecting data, TA is a method for data analysis. You use TA to develop, analyze, and interpret pattern in your data set. The end product is the generation of themes.
Big Q TA is called Reflexive Thematic Analysis (Braun & Clarke). It foregrounds the active role of the researcher in coding and theme development. It emphasizes the researcher’s work of reflecting on their personal assumptions and practices and how those assumptions and pratices shape their data analysis.
Doing Reflexive Thematic Analysis
Before we get into the work of analyzing a data set, reflexive TA starts with moments of reflexivity. Reflexivity involves reflecting on your assumptions, expectations, choices, and actions to recognize and be aware of the influence of who you are and your position on the knowledge you are generating. This kind of data analysis recognizes that who the researcher is impacts the research that is carried out and the knowledge that is developed in the study.
There are three kinds of reflexivity you need to conduct before (and throughout!) your reflexive TA work: personal (who you are, your social position and how that shapes your perspectives on the topic of study; e.g., How does my gender identification influence my attitude towards the phenomenon of study?), functional & disciplinary (how you were academically trained, your research preparations; e.g., what do I think good research looks like? What theories am I most comfortable with?), and topic (life experiences; e.g., What assumptions do I have about the topic being studied). You do this work to make explicit how who you are is shaping how you are thinking, your interpretations, and the knowledge being developed.
Phase 1: Familiarization of data. Read and reread the data to become deeply familiar with the data set. The goal of this phase is to write some initial thoughts and insights about what you see in the data.
Phase 2: Generation of codes. Closely and systematically reading the data, identifying segments of data that are important / interesting / meaningful to the research question, and generating statements that describe codes. This is identification work. It is line-by-line reading and involves some interpretation, but it remains rather close to the data. TIP: One code captures one meaning unit.
Phase 3: Combining codes into initial themes. This involves identifying shared meaning patterns across the data set. It requires the researcher to bring together clusters of codes to answer your research question. This work is more abstract and requires the researcher to abstract away from the data to make broader analytic interpretation. TIP: One theme holds together 2 or more codes together.
Phase 4: Reviewing themes. This is the work of verifying that the themes make sense to the codes and to the entire data set, and that they answer the research questions. This is when the researcher pushes their thinking and challenges the themes that have been developed. What is missing? What has been added that doesn’t belong? Are these themes truly different? Should they be combined?
Phase 5: Refine, define, and name themes. This is the work of ensuring that each theme is built around a strong core concept. What does this theme tell me? How does it answer my research question? This involves working to develop the story of the study. What is the “therefore statement” that these themes help me develop.
Phase 6: Reporting of findings. The research weaves together the analytical narrative with compelling data excerpts to answer your research quesiton. You are writing your manuscript here.
NOTE: Phases 2 through 6 happen iteratively. You will need to revisit and repeat the phases over and over again because your thinking will evolve over the course of your analytical work.
Pitfalls to avoid when doing reflexive TA
- Don’t mistake codes for themes. Do sufficient interpretation to generate themes. Don’t stop at codes
- Beware if your themes have significant overlap. You might need to merge the themes together. Make sure you can articulate why they are different.
- There must be a clear connection between your data and your themes!
References
Braun, V., & Clarke, V. (2022). Thematic analysis: A practical guide. SAGE.
Kiger, M. E., & Varpio, L. (2020). Thematic analysis of qualitative data: AMEE Guide No. 131. Medical Teacher, 42(8), 846–854.
Olmos-Vega, F., Stalmeijer, R., Varpio, L., Kahlke, R. (2022). A practical guide to reflexivity in qualitative research. Medical Teacher. Apr 7:1-11.
0 comments