Hello everyone.....!
Can any one suggest me,
How to justify in Multiple case-studies samples...?
For example,
Unit of analysis : Community (Respondent all are part of community)
Case-1:14 respondent,
Case-2: 1 respondent,
Case-3: 1 respondent,
Limitation may include: Travelling, Language, time and financial aspects,
How to justify in argument, ( For internal reliability and external validity)
I will be thankful......!
if some one can share the relevant research (Links, articles) on social science.
Thank you all for contributing your time and effort,
Really appreciate your concern,
Best,
Karun Rawat
The best book I know on comparative case studies in Yin's Case Study Research: Design and Methods, so I would look at it and see what he has to say.
In general the standard in qualitative research is saturation: when have you talked to enough people that you are not learning anything new as you do additional interviews. It is very unlikely that you would achieve saturation with only a single interview.
Hi Karun, I agree with David. If the unit of analysis is community, one respondent is less likely to be representative of the community you are studying. Having said that, with multiple respondents, saturation becomes a good indicative measure to determine the number.
Sorry to say that I think it might be hard to justify just having one respondent for the second two cases. Having said that, it will somewhat depend on the kind of questions you are asking. If your respondent is more of an informant providing you with information concerning e.g. the number of members or other issues that are more factual, one respondent can be sufficient. But if it concerns issues that are more open to interpretation and depedent on subjective feelings, one is seldom enough. ,
@ Fredrik Movitz,
Thank you for your feedback, it really showed me the light,
I know what are you trying to address, i also have the same thought,
I will be more clear if you could provide me some of the relevant study, from where i can see how various researchers have approach this style and analysis part.
Thank you for contributing your time,
Regards,
Karun Rawat
I agree with David re Yin and would also suggest Stake. I also found this article which I found interesting and which might be of use.
http://www.nova.edu/ssss/QR/QR13-4/baxter.pdf
Dear Dr. Rawat,
I have gone through your query and the replies to it.
Unfortunately, Ia am not an academician, I am a Librarian by profession but off and on I do field studies as a part of my other consignments/interests and as I observed that number of requisite respondents often vary according to demographic size & patterns of the requisite subject areas However, it should constitute a certain valid percentage of the population for determination of samples for case studies. For instance, once I had associated with studies of various villages of a particular District in West Bengal, a state in India. Incidentally, the districts had boundaries long with Bangladesh (another Country) and three different Districts and we had to divide the District geographically to under stand the typical issues of concerns the rural people had been facing, In one part, people were concerned with trans border infiltration and burglary. In one part, the principle concern was erosion of highly fertile land due to change of flow of river. For one part, recurrence of flood and its impact on agriculture had been the main concern while the for last part, drought and water shortage of water (both potable and for agriculture!! We had to choose sample size on the basis of population of each of the four regions and naturally they had been uneven. For Household survey, the "number of Households covered" were naturally different but, we collected population data of of all the four regions and covered a Fixed specific percentage of the population for each of the four regions and naturally the number of respondents varied for because of the variance of households.
I hope You can understand my point.
Siddhartha S. Ray
- Badges
- Science topic
- Similar topics
- Social Science