From codes to themes: Experimenting new code arrangements

From codes to themes: Experimenting new code arrangements

13.06.2018, 21:01

Hello everybody,

A very popular technique of qualitative data analysis is known as thematic analysis, which - put simply - posits that the analyst should define themes after generating a code scheme. Codes are more descriptive (as demonstrated in MAXQDA tutorials) while themes are more abstract issues. Themes are not upper level codes. They should synthesise the meaning of the dataset, and the process to identify themes usually involves multiple reviews. [For those interested in the method, Braun and Clarke (2006) provide a helpful overview on thematic analysis.]

This introduction was necessary for putting my question which is not methodological but related to the software. In order to identify themes one will need to play around with data, and explore different relationships among the coded data. For instance, if a subcode appears in different parts of the material, one may think that something related to this subcode is a theme.

I will provide a hypothetical example to make my point clear.
For example, the code scheme around "Activities" seems like:

- "I love it"
- Obligation
- ...
- Enjoyable
- Healthy
- Need to get to places
- ...
- Since childhood
-- at school
- World cup
- ...

Someone may rearrange the coded segments and the links between codes from the perspective of a theme ("Pleasant" or "Seek for pleasure") which could be framed as a combination of "I love it", "enjoyable", and "since childhood". It may be it doesn't fit so well, then another grouping will be tested and so on, in an iterative process.

How is the best way to do this rearrangement with proper visualisations, preferably keeping the original code scheme?

As another user pointed out, Sets are not compatible with MaxMaps (

Many thanks for any help here!


Version: MAXQDA 2018
System: Windows 10
Posts: 12
Joined: 09.12.2017, 23:17

Re: From codes to themes: Experimenting new code arrangement

14.06.2018, 18:00

Hello TGR,

thanks for your question and for pointing out Fred's thread, I just realized I never answered him... probably hoped someone else would and no one did. I find these questions rather difficult to answer because my perspective on the whole process is, I'm afraid, rather unorthodox. But I'll give it at try..

For me, at least in the beginning, this rearranging that you describe, the iterative process, happens with tons of sheets of paper (preferably A3) where I scribble many ideas about how all of my data is (on a descriptive level) or might (on an abstract level) be connected. The reason why I don't do this with any software is because in the beginning, most of my ideas seem inadequate, and no software or visualization-tool in the world can keep up with pen&paper when producing and discarding inadequate ideas (in my opinion). The "rearranging"-part itself, the making-sense of the data, the shattering of existing images to produce new mosaics, this in my eyes is a creative part which can not adequately be described as just following a set of rules and methods, and so I wouldn't look for a software for that. Even if one could only use methods recommended by a certain methodology – how could one really know that this methodology is the right one? It must always be based on assumptions that can either be theoretically falsifiable (making them, following Popper, scientific), or they are either in theory or practice not open for discussion, not falsifiable, and then we are in the realm of religion. (As a sidenote, I spent some time reading Fleck, Popper, Kuhn, and, with the utmost delight, Feyerabend.)

But to come back to the original question: It just makes me crazy to make a nice visualization and realizing afterwards that it's all utter nonsense, and at some point, it might even tempt me to stick with it simply because the deadline is near. And since (at least in my experience, but I think I could successfully generalize this to some extend, unfortunately) no one but me is interested in most of my discarded ideas and how and why I came to discard them, I really don't see the point.

So usually before I would really make a visualization of a theme or continue with some "axial" or "selective" coding, I checked my coded segments hundreds of times, recoded, wrote summaries etc. pp., until I have a _good_ idea about it.. until it's rearranged in my head. _Then_ I think it's a good idea to use MAXMaps to solidify this and make it transparent to other people (although I probably still discard a few maps before I'm satisfied). To visualize a theme, one could use a freeform object (like a very big circle) to mark its boundaries (what belongs to the theme, what doesn't, what other themes are there, how is the relation between them etc.). Also, I would always not be shy to use free text fields, e.g. when you want to have a subcode that you want to put in two places at once.

So to come to an end, my honest answer to your question:

How is the best way to do this rearrangement with proper visualisations, preferably keeping the original code scheme?

Would be to use visualizations when you have a (one) good (doesn't need to be final of course) idea of what your data is about, and if you use it for rearranging, then I would do it at a later stage.

It's also much nicer and more healthy to think while pacing up and down the garden then it is to stare at a display all the time.

I hope this helps somehow or at least shows my good will ;)
MAXQDA Support Team
Andreas V.
Posts: 274
Joined: 13.04.2017, 16:23

Return to Technical Questions

Who is online

Users browsing this forum: No registered users and 4 guests

We use cookies to improve your experience on our website. By clicking OK or by continuing to browse the website, we’ll assume that you are happy with their use. Click here to review our Cookie Policy. OK