There is a growing body of published research and theory/philosophy which suggests that generalizability from qualitative research is both legitimately possible and is 'done'. What I'm looking for now is the contemporary examples from the research literature that shows how and where it is done. I'd appreciate any examples.
Surely generalisability is dependent on the representativeness of your study samples. It wouldn't make sense to generalise your findings across populations you have not sampled from. i think, therefore, that you answer lies within a discussion of sampling - what kind of sampling do we use in qualitative research?
Qualitative research often uses purposive sampling, singling out certain information-rich individuals as study participants. Even if it uses a randomly selected study sample, this is usually small, with no statistical power. My point is - qualitative samples are not very representative of a population, therefore generalising qualitative findings, and suggesting they are a representation of what the population under study would say, had they all have been asked, seems a bit dodgy to me.
I prefer to use qualitative research as a scoping tool which helps direct me toward the most important quantative studies. Which are generalisable, if they are done right.
Thanks Tony for your response. Your lost point about scoping is one I am interested in. Have you got examples that you can point me to?
I prefer the idea of "transferability" to "generalizability" when it comes to qualitative research. For better or worse, the term generalizability will probably always be associated with large random samples. As an alternative, transferability asks us to make statements about the other circumstances in which our results might apply.
Thanks David, that may be so, but have you examples from your own published research that shows how that transferability took place and to what end?
My best example dates back to 1989 and unfortunately I don't have a copy of the article. At the time, there had been a strong of failures to demonstrate that involved in social networks would improve wellbeing. We did focus groups with recent widows to find out about the role that social networks played in their coping with this stressful life event.
What we found was, back then, quite unexpected: that negative relationships, although rare, had a strong impact on how well these women were coping. In particular, problematic relations with their adult children not only deprived them of support but produced major sources of stress.
Where transferability comes in was our argument that similar issues would apply to other stressful life events, beyond widowhood, and that these issues would be especially acute when close family members were involved.
See attached. There is an example of what might be called a scoping exercise, using qualitative research, on page 2, box 1.
For me this is honest qualitative research, doing what it does best, scoping out possibilities, directing further inquiry, through more research; and ultimately, as in the attached, fueling a drive toward large experimental trials focusing on real-world applications.
Great example thanks Tony. Definitely one of the ways that qual research is used in a process towards generalisation.
- Badges
- Science method
- Similar topics
- Philosophy