We have carried out in-depth interviews of participants during and after the COVID-19 lockdown in India. We are using thematic analysis to understand people's experiences.
At the moment, we are writing the results in 2 parts: Phase 1 results and Phase 2 results with separate discussions and a general discussions. Each Phase has generated 5 themes. Making it a total of 10 themes in the entire study. I understand that many researchers and authors suggest to not go over 5 themes in a thematic analysis.
However, does it have merit to have 10 themes in this case? Or is it recommended that we further reduce the themes. What would be the optimum number?
Edit: My first phase has 46 participants and the second phase has 35 participants
Fahreen - themes are what they are. Personally, I don't go for the 'too little, too many' debate. If the themes have naturally evolved - they are merely a representation of the needed narrative. One cannot 'take out' themes if they already exist. I tend to be more critical the other way. If very few themes - I might argue - was an appropriate interview schedule used (assuming semi-strructured)? You talk about two phases - so I assume it's a multi-methods study? If so - then the data is seperate - but related to the whole. There is also the question of 'are they all main themes' - or might some be sub-themes?
The atatched chapter may assist.
Fahreen - themes are what they are. Personally, I don't go for the 'too little, too many' debate. If the themes have naturally evolved - they are merely a representation of the needed narrative. One cannot 'take out' themes if they already exist. I tend to be more critical the other way. If very few themes - I might argue - was an appropriate interview schedule used (assuming semi-strructured)? You talk about two phases - so I assume it's a multi-methods study? If so - then the data is seperate - but related to the whole. There is also the question of 'are they all main themes' - or might some be sub-themes?
The atatched chapter may assist.
Hi there. Depends on your "phases". If your phases are examining different aspects and your themes are emergent of those, then its fine. I also would pay attention to the "size" of the themes. Larger themes can be standalone, whilst smaller themes can/may become subthemes of the larger ones (if they are related). Enjoy.
Dr Meraj:
How many themes should you have in thematic analysis?
IMO it is better to have 6-10 broad themes, rather than lots of really detailed ones. Once you have applied the framework, you can then read through the material that has been coded under each theme and identify any further nuances or differences within them.
https://www.researchgate.net/post/When_using_thematic_analysis_for_multiple_interviews_do_I_apply_themes_for_each_interview_individually_or_collectively
The main problem I have when there are more than 5 themes is that I simply cannot keep track of them all. In my opinion, there are two functions for themes: first, to represent what is in the data, and second to communicate that content to the reader. This second function is often over looked in at attempt to be "true to the data," but if you cannot give the reader a clear sense of what your interpretation was, then why bother?
One way around this dilemma is to include a table (or figure?) that summarizes each set of themes.
But I am still not clear about why you have two separate sets of data and two separate sets of themes, which also leads to the question of why there is no overlap between the two sets of themes.
As far as applicable, the generation of themes is not predetermined prior to data analysis, though one has to be guided by the existing (formulated) objectives of the study. Indeed, it will be advisable for the researcher take care of new issues which may not have been thought about prior to the data collection. In this regard, such new issues will be presented and analysed under a new theme. Hence, I will advice that let your issues guide you in the determination of the number of themes to be developed
Kindly check https://journals.sagepub.com/doi/full/10.1177/1609406917733847
Also check https://www.scribbr.com/methodology/thematic-analysis/
It is tricky to know without the full context of things, but we are also often guided by the word limit constraints of papers. You will want to be able to provide enough detail and quoted examples within each theme. Having lots can mean things are considered at quite a shallow or descriptive level. Sometimes we might choose to focus on themes that are most prevalent or say something very important about our research question. This isn't to say we ignore that others were there, but that we choose to focus in more detail on certain ones for the purpose of a paper. I wonder if things feel too restrictive, if this might actually form two papers looking at each phase. Or - it may even be the case that both phases could actually be considered together in the analysis write up, but you specifically consider when an analytic point is most related to one phase or the other. As often with qual approaches, things are often a judgement call. But at least these methods offer us flexibility. Best of luck!
What I believe it is very important in every research or analysis/article is that you have a very clear idea what is your main problem that you want it to be solved by your research. You need to focus on what you want to do it, how you want to do it and what is your goal.
As these data are not available and also many others, it is quite hard to answer your question. At the end of the day the number of themes might be irrelevant.
Good work!
Thank you for your question Farheen Meraj .
I think there are no ideal numbers of themes. The most important thing is (1) Not to be too many, that will confuse the picture. (2) Ensure that the process is meaningful, (3) Ensure that important themes are not missed. Many times I edit or peer-review papers, and I find them fragmented and not meaningful. This is very important when you conduct such studies.
- Badges
- Science topic
- Similar topics
- Psychology