I have been reading about mixed methods synthesis techniques (Sandelowski et al 2006, 2012 and Joanna Briggs Institute 2014) for systematic review.
I think a segregated synthesis approach will suit my subject well but can find references only to qualitative/quantitative studies - how do I deal with mixed methods studies? Would I separate the data into the above categories and place that data into the appropriate synthesis? I'm not sure if that would work.
I think that a lot would depend on the research design of the mixed methods study you were reviewing. For example, if a study used a preliminary set of focus groups to help develop a piece of survey research, that could be represented as a qual --> QUANT. But the article in question might use focus groups in a way that would be quite different from a full-fledged qualitative study of the same topic. In that case, there would only be one usable set of results, the QUANT from the survey.
At the other end of the extreme are cases where the uses of the two methods are so thoroughly integrated that it would not make sense to speak of separate qualitative and quantitative results.
What it sounds like you have in mind are what would be called low levels of integration in mixed methods, where an article could essentially be divided into qualitative and quantitative findings , which could each be independently of the other. You certainly can find many such studies, but mixed methods studies that go further in integrating the two components are much less likely to fit the model for separate review.
Mixed method approaches to research are tricky all along the research pathway. I've always done the analysis of qualitative data separate from the quantitative data. However, In the report or publication try to show how the results complement, support or contradict each other. Like David said a lot depends on the design.
Maybe I'm best to wait and do the systematic search and see what primary resources I find (not excluding any methodology) then settle on an approach then revise my methods section
Waiting to deal with the mixed methods studies and until you see what you have strikes me as a good idea. Then you could say something like: Of the total number of mixed methods studies found, XX% provided complete results from both qualitative and quantitative methods, YY% from qualitative methods only, and ZZ% from quantitative methods only.
Yes - I think perhaps trying to solve a problem that I am anticipating but haven't yet encountered might actually be asking too much!
- Badges
- Science topic