Req: RDQA's 'relation' function

[glocalglocal]

Looking at RQDA's functions listed here one function that stands out as not being supported by qualcoder is 'relation' between two codings.

I am not quite sure how people use it. I think it is about proximity, which then can be used for hypothesis testing and theory building. If these data are exported, researchers can then perform statistical analysis (eg coding A appears near coding B that much more often than coding C) and somewhat bridge the gap between qual and quant, possibly for corpus analytics. People describe this as one of RQDA's strengths, and I can see why, but I am not an expert in this kind of qual/quant method and don't have a working RQDA setup so can't help much. Perhaps worth exploring, esp when people single this out as important.

[RadicalLibre]

Yes that's a very useful function. And also the Profile Matrix - coding function of RQDA, it would be useful to add it to QualCoder, too. Regards, El 4/8/20 a las 8:21 p. m., glocalglocal escribió:

Looking at RQDA's functions listed here <https://www.rdocumentation.org/packages/RQDA/versions/0.3-1> one function that stands out as not being supported by qualcoder is 'relation' <https://www.rdocumentation.org/packages/RQDA/versions/0.3-1/topics/relation> between two codings. I am not quite sure how people use it. I think it is about proximity, which then can be used for hypothesis testing and theory building. If these data are exported, researchers can then perform statistical analysis (eg coding A appears near coding B that much more often than coding C) and somewhat bridge the gap between qual and quant, possibly for corpus analytics. People describe this as one of RQDA's strengths, and I can see why, but I am not an expert in this kind of qual/quant method and don't have a working RQDA setup so can't help much. Perhaps worth exploring, esp when people single this out as important. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#244>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAFJMKL32UISVYIEXIU2MBTR7CQY7ANCNFSM4PU5QWTA>.

-- Omar Bautista González Colaborador en autogestión comunitaria desde República Dominicana Docente de la Universidad Autónoma de Santo Domingo

[menelic]

I believe this is also used for RQDAs visualisation of code relations, which is why implementing this feature would also address #182 and strengthen QualCoders interoperability with R. It appears that an SQL database that can be queried from R is what enables RQDA to use igraph to visualise code relations. RQDA creates more interactive graphs that can be "sorted" in more insightful ways that qualcoder graphs at this point. I believe including relational data would enable QualCoder to implement better visualisation as one of the last missing features.

[ccbogel]

@glocalglocal, I have begun this, please check it out. Hope I am on the right track with this.
Menu Reports > Code Relations
Select 2 or more codes. Then Press Calculate.
The Relation results are shown in a table.
Hover mouse over cells in the table to show more details. E.g. code id is in the table, but hover over it and you will see the code name.
It only shows relations for the codes within the same text file, which makes sense. However, all text files are examined.

Union positions shows the lower and upper union of 2 codes (as in RQDA)
Overlap positions shows the positions from the lower position of the lower code to the upper limit of the overlapping code (as in RQDA)

Also can export a detailed results csv file

[ccbogel]

@glocalglocal @menelic @RadicalLibre I have implemented the relation function, so will close this issue.
Menu Reports > Code Relations
Select 2 or more codes. Then Press Calculate.