17th February 2010 at 10:59 am #4997Santosh MishraParticipant
Most of the times, the onus is on the researcher to have findings that can be generalized. It has been a popular belief that higher generalizability means a better research.
Thanks.17th February 2010 at 11:08 am #5011Luca GhirottoMember
That’s a very interesting discussion. I am used to have such conversation with colleagues, mostly from psychology and cognitive science. But they do not recall what Wundt said about generalizability and qualitative research. For him, qualitative research reaches a even more deep comprehension than quantitative one. Interesting!17th February 2010 at 6:37 pm #5010Alan BrownParticipant
To quote Stanley Lieberson “Happiness is Variance Explained”–I have never understood this critique of qualitative research. As a methodological approach, it does not pretend to be “:generalizable”–so why should we fault it for not doing something that it doesn’t claim to do? It is like going to a cafe wanting a cup of coffee, but instead ordering a glass of milk then complaining about how unlike coffee the milk is.
In Lieberson’s book “Making it Count” he urges folks to “do the doable” and focus on asking better questions and letting that decide the methodological approach as opposed to high power quantitative crutches that are elaborate, sophisticated and, often times, inappropriately used,.17th February 2010 at 7:10 pm #5009Pat BazeleyParticipant
Maybe the issue isn’t so much how can we come up with something that is widely applicable, but: what of significance have I learned from undertaking this research? does it answer the question I set out with (or an ‘updated’ version of it)?17th February 2010 at 11:32 pm #5008
Dear Santosh Mishra,
I think that is right the consideration regarding that new findings should be generalized. So that it is interesting and good.
Giuseppina17th February 2010 at 11:49 pm #5007Bob DickParticipant
Speaking for myself, here’s my brief comment on an issue that I think deserves a longer account.
First, I don’t believe we should give up on generalisability. Of course we can generalise, with due caution. If we couldn’t, we wouldn’t be able to learn from our experience. Experience would count for nothing.
So for example I would maintain that it’s possible to generalise from qualitative research by conducting multiple studies. We can also use the literature to help us decide how broadly our research conclusions can be applied.
For me, the question isn’t “Do these findings generalise”. It’s “Where is the boundary of application of these findings?” — Where else can I apply them, and with how much confidence?
Second, generalisability is valuable, but often less valuable than some of its advocates claim.
Suppose I conduct a careful quantitative study under controlled conditions. Then I can generalise my finding from that study into the “wild” — but only when exactly and only the same variables apply that I controlled or measured in my study. That’s not often.
I think of it like this. To some extent there’s a trade-off between generalisability and relevance. Depending on the purpose of my research I can decide on what balance I’ll seek. Within that balance I can then optimise both generalisability and relevance as much as possible.18th February 2010 at 5:14 am #5006Santosh MishraParticipant
Thanks Giuseppina. I agree with Mr. Bob Dick’s argument, especially the second one, and the very tactical reply by Mr. Allan Brown. In most of the qualitative studies, the variables are often overlooked when the findings are posed as arguments for furthering any new research in conditions often alien to the former researcher who came up with the findings (may be) in a different context and the gap between the two research contexts may play a deciding factor in guiding/misguiding the factor of generalizability.
Researcher in English Language Education18th February 2010 at 8:06 pm #5005
Dear Bob Dick,
I am in accordance with You. In particular, all depend on the purpose of the research and we can decide on generalisability and relevance.
Giuseppina18th February 2010 at 8:38 pm #5004
Dear Santosh Mishra,
I thank you for the answer. I agree with you and Bob.
Giuseppina21st February 2010 at 10:05 pm #5003Jenny HallMember
If it helps at all in some literature it is suggested we should not use the term generalizable for qualitative at all but use ‘transferability’ instead, to lessen any confusion. Therefore is the research transferable to a similar group of people?
jenny23rd February 2010 at 8:03 pm #5002
Dear Luca Ghirotto,
I think that your discussion regarding the qualitative search is very interesting.
Giuseppina24th February 2010 at 2:10 am #5001Julia ThorntonMember
That’s interesting. What literature? There is a real ontological difference between “generalisation” and “transferability”. “Generalisation” suggests a realist model, “transferability” a constructivist model of the world.
Julia24th February 2010 at 10:40 pm #5000
Dear Santosh Mishra,
I think that your words are real and right, but each search is useful and qualitative correct. So, it is important to consider the field of search application. It is the most important to make qualitative search in each field.
Giuseppina25th February 2010 at 9:32 pm #4999
I think that is a quite difference between “generalisation” and “transferability”. I think that “generalisation” is linked to a great general meaning but the “transferability” is a more complex model.
Giuseppina4th August 2010 at 7:38 pm #4998Mario CardanoMember
The discussion on generalization in qualitative research is always interesting. I do agree with Bob Dick on the necessity do not renounce to the goal of extension of the predicability of statements built through qualitative research. Nevertheless, the solution of Guba and Lincoln to the problem of the predicabilty extension (I prefer this expression instead of generalization, or extrapolation – proposed by Alasuutari) through the way of transferability do not convice me. Actually the transferability solution moves the burden of proof of the adequacy of the predicability extensions from the researcher to the reader. The reader must decide about the suitability of the “transfer” from the sending to the receiving context comparing the “thick descripions” of the objects set in the opposite shore of the river that transferability must help to cross (a recent book of Martin Hammersley, Questioning qualitative inquiry, presents an interesting critique of the notion of thick descripion). So I think we have to follow another way (that do not include the solution offereb by Glaser and Strauss of “theoretical saturation”). I find myself quite confortable with the idea of Jennifer Mason of “sampling strategically” (see Qualitative research, second edition), binding this approach with the use of instruments of rhetorical argumentation to defens the quality – as Howard Becker says – of our synedoche. We can renounce to the theory of probability and fill up the void left by this last with the theory of argumentation: we can exchange Pascal with Aristotle…
- The forum ‘Default Forum’ is closed to new topics and replies.