Qualitative content analysis – checklist of methodological decisions

The label “qualitative content analysis” encompasses many variants and characteristics. How can I then describe my methodical approach as well as possible? Of course, one option is always to follow established procedural suggestions. You then work“according to author:in XY“, which is okay in principle. Beyond that, however, as already indicated in other articles (e.g. here), there are much more precise and confident ways of describing your “own” qualitative content analysis. This includes citing the many small methodological decisions that are made in such an analysis.

Since I like to make things easy for myself, I have created a small overview of possible methodological decisions – based on an article by Margrit Schreier on FQS – without claiming to be exhaustive, of course. Margrit Schreier’s“Qualitative Content Analysis in Practice” can also be recommended to all those who are interested in this approach in the sense of a “toolbox”.

Aim of the evaluation and planned way of presentation of results

The reference point for all further questions is, of course, the research question. So: What do I need to learn from the field or from the material in order to be able to answer my research question? This indicates the type of results report that will be produced even before the analysis.

• Do I need case descriptions to describe concrete processes and patterns of interpretation?
• Or perhaps rather content-semantic summaries to be able to describe larger amounts of material?
• Or do I need an overview of the connections between different topics?
• Or am I interested in answer patterns and may have to form types?

Selection of the material

“All is data” is an infamous Glaser quote, but what data should or must be taken into account? And even if it is clear that I am using interviews, it is not clear. Therefore, one should not stop at “I do interviews” for the sake of argument.

And where does my material come from?

• Which material is suitable to answer my research question? Do I need image material, text material, observations, interviews?
• If interviews: How can I organize them so that the interesting topics come to light?
• How do I get the material, which interview partners are suitable and how do I select them?
• Can I give criteria for how much material I need? Or at least criteria for when I realize that I have collected enough material?

Of course, this is closely linked to the question of what actually interests me about the material:

• semantically content-related mentions,
• Argumentation patterns
• Narrative process, dramaturgy
• Choice of words, linguistic peculiarities
• Linking of topics

Creation of the category system

How do I arrive at my category system? Certainly one of the core questions in qualitative content analysis. As a rule, main categories are derived from the research question. (But even that is not set in stone.) The following questions can be answered differently and separately for each main category.

How do I get to my categories?

• Do I have concepts that I can use in advance? (A-priori, theoretically defined categories)
• Where would like to form categories on the material.

When I create categories on the material, I can choose different strategies here:

• I summarize the text passages (step by step)?
• I sort the text passages directly according to similar statements (subsumption)
• I first use open coding as is usual in grounded theory, i.e. I write memos on the text passages and then look for similar concepts in the answers.
• I form contrasting categories, so I always look for two text passages that are as opposite as possible and form a category from what they have in common?

In connection with the previous point, the question arises as to what I actually want to sort into the categories:

• I focus on semantic content in general, i.e. “what is said”?
• I am interested in actions or descriptions of actions?
• I am interested in emotions, ratings?
• I want gradations of characteristics?

Once I have coded, the next step is to test the code system. This can be understood in two ways. I can test whether the code system can be sensibly applied to the other material and I can check whether my code system is sensibly and logically structured.

• As a separate work step, or iterative alternation between coding and testing?
• I don’t do trial coding because I include all the material straight away?
• I do this alone, with coding discussion with others, I calculate a coefficient (if so which one?), both
• Trial coding with how much of a fraction of the material?

And finally, we need to think about the subdivision into units of meaning. So how do you actually know how much text is marked during encoding?

• How do I define coding units, in advance or during coding?
• Which definitions do I use for coding units (what is marked), individual words, sentences, paragraphs, semantic units, entire text…

 

Basic idea: Think from the result.

That was it. You certainly don’t have to work through this list, but you can certainly use excerpts from it as a source of inspiration or ideas. The basic idea behind this is to think as early as possible about what the result might look like. Some questions have to be answered quite pragmatically: If I work alone, then perhaps I don’t have the luxury of an interpretation group. Nevertheless, whatever the decision, as long as you write something about it, the procedure remains transparent, comprehensible and open to discussion. And is therefore much more meaningful than“I work according to XY“.