Code 'tree' in code list

[QC-user]

I've been using QualCoder for a few years now, after the experience of using a proprietary alternative. One of the features I miss is the possibility of organising some of my codes as a 'tree' of codes and subcodes - in a way that when applying a certain subcode, the higher code under which it is nested can be applied automatically.

When we are adding codes in the code list, I suggest that 1) we are allowed to add subcodes under previously created codes and 2) to specify whether we do or do not want that the higher level code is applied automatically when a certain subcode is applied. Currently, QC allow that we organise codes in higher 'categories', but these 'categories' cannot be used as codes.

Suppose I have a code for 'fishing' and subcodes for different fishing techniques. If I am interested in all my coded data regarding fishing (either generic or specific), I could retrieve that by asking a report for the higher code 'fishing', if the suggested feature is implemented. The organisation of codes in the code list and the process of generating code reports would be greatly simplified in that way, in my view.

[kaixxx]

Regarding the hierarchical code system: I would second this idea. I think the decision that only categories can have sub-categories/codes dates back to QualCoder's roots in RQDA. But today, it is a major difference between it and other common tools like NVivo, MAXQDA, Atlas.ti..., that all allow codes to have sub-codes (as far as I can tell).
This breaks workflows for people coming from such tools. Also, it makes importing projects or codebooks from these tools complicated. And it makes it difficult to adhere to inductive methodologies that demand for an option to differentiate a code or phenomenon over the course of your analysis by adding sub-codes, representing aspects/dimensions/variants of a phenomenon.

My suggestion would be to add the possibility to create sub-codes, but leave the current category system in place. This would make migrating old projects frictionless. Also, it makes epistemological sense to have a clear distinction between empirical phenomena with their sub-dimensions represented as codes and sub-codes, and overarching theoretical concepts represented as categories.
As this would require some UI changes, it is not something that can be done quickly. But maybe for one of the next versions.

Regarding the 'auto-coding' with parent codes, I am not so sure. At least this should be optional. I understand the basic logic, but it would mean that we end up with a lot of text fragments that have two, three, four or more codes attached in parallel, depending on the depth of your code tree. Maybe an easier option would be to allow for an easy selection of a code with all its subcodes when generating reports. What do you think?

[QC-user]

Yes, I agree, it would be useful to see a hierarquical code system coexisting with the current category system.

Regarding the following alternative suggestion:

Regarding the 'auto-coding' with parent codes, I am not so sure. At least this should be optional. I understand the basic logic, but it would mean that we end up with a lot of text fragments that have two, three, four or more codes attached in parallel, depending on the depth of your code tree. Maybe an easier option would be to allow for an easy selection of a code with all its subcodes when generating reports. What do you think?

I think that the possibility (I agree that you should be able to opt in or out of this) of auto-coding with parent codes would simplify report generation, specially in the cases concerning my other suggestion (#1209 ). Suppose for exemple that I am interested in sections where a certain parent code (and its subcodes) overlaps with other parent code (and its subcodes).

[ccbogel]

Yes, good food for thought for the future.