,
20 tweets,
5 min read
Read on Twitter
THREAD - Here are my thoughts on generating themes in qual research & why they don't emerge. I've included my BAE notes to situate myself. I hope you will read and share. #AcademicTwitter #qualitativeresearch #critqual #highered #highereducation #educolor #dataanalysis #phdchat 
1. First, generating theme is only ONE type of inductive analysis in qual research. It shouldn't be confused as THE path to data analysis in qual research. Thematizing is quite intuitive, despite being presented as a form of systematic approach to data analysis in qual.
2. Also, while some disciplines privilege a systematic approach to qual analysis, because identifying codes, categories, themes can be an extremely intuitive process to which we try to put language, a lot remains unexpressed even after articulation.
3. Generating themes is an inductive process of chunking raw data, analyzing them, grouping things in their semantic units, and trying to find patterns within and across data sources. There are various ways of achieving this, but I will share one of the most common ways.
4. One of the most common ways of generating themes is to code the raw data pieces, i.e. create semantic units of meaning through some analytic process. Saldana has a great book (The Coding Manual) that outlines ways one can code the data (theory, emotions, process, etc.).
5. Note, the process of moving between coding, categorizing (more on this later) and thematizing is anything but linear. It's iterative, messy, and could also contain contradictions and tensions. I only write linearly due to the medium of expression.
6. During coding, one also sharpens the focus to identify similarly coded data for grouping them under another semantic unit of meaning, often called categories. This is also iterative because some codes become categories & some categories become codes in this process.
7. Point of Consideration - A good practice is to be as analytically descriptive as possible to label codes & categories to bring in sharp focus. Generic labels such as family, support, justice, oppression don't work well because the nuanced meaning gets lost in those labels.
8. What is it about "family" that stands out in a code or category? Is it lack of support, or tricky negotiations, or feelings of isolation or rejection? Of course all of this should be juxtaposed with the research questions & the theoretical framework of the study for sharpness.
9. During this process the researcher writes memos, journals about thoughts, hunches, elaboration of codes, categories, relationship with the research purpose, questions, framework, existing literature, stuck places, researcher positionality, subjectivities, etc.
10. This reflexive writing process is the key to language some of the intuitive ways of knowing, analyzing, and sense making. However, most researchers find it difficult to articulate how they arrived at a theme, & so they often say "Themes emerge" or follow established examples.
11. Themes cannot "emerge" because they are not agentic beings who can pop up on your computer screen and scream at you and say, "Hey look at me, I am a theme, I am now emerging. Give me a name. It's my born day."
12. While engaging in analysis, the researcher looks within & across codes, categories to look for patterns. The researcher identifies what these patterns are sometimes through reflexive writing, deep analytical thinking, using research purpose & frameworks, as guidance.
13. So themes do NOT emerge, but they are IDENTIFIED by the researcher and when shared with the participants, co-constructed as verified patterns within and across data sources.
14. I see a lot of people use generic and intuitively obvious titles for themes. Themes reflect a deep analytical insight as represented in an identified pattern in the data. It has to be something that was otherwise not possible to uncover without the researcher's analysis.
15. So, if someone has a theme, "PWI discriminates students of color" it is not that great of a theme, because it does not represent an in-depth understanding of a pattern of data. Could this have been said without doing the study? If yes, then it is not a strong theme.
16. Now, if there are certain patterns in the discrimination that are important to highlight, then that could be valuable insight to share as a theme and unpacked in the thematic elaboration. Perhaps ask, "What surprised me in this study? What did I not know before?"
17. If people need a depiction of this detailed messy process, I tell my students to draw a visual of their iterative messiness and then speak to it, so that it can disrupt some of the assumptions of linearity. They use some conceptual mapping or graphic design softwares.
18. In summary, the researcher identifies a theme, co-constructs the veracity of it by sharing it with the participants (to whatever extent it's possible). Thematic titles should demonstrate critical in-depth insights, & not intuitively obvious statements lacking original ideas.
Wrote this in between meetings. Forgive typos and other mistakes. 🙏🏾
