What kind of sampling frame and research design should be used for study of involuntary bachelorhood in Indian context?
Are you doing qualitative or quantitative research? A small sample would not be a problem with qualitative research, if you select it carefully. But you might indeed have problems getting a large enough sample for meaningful statistical analysis.
Ambarish -
You could research Monroe Sirken's "network sampling," used for hard to reach populations, to see if that works for you.
Cheers - Jim
Or you could adopt a case comparison design where you identify say 200 people who are involuntary bachelors and say 200 to 1000 who are not; that is you oversample on the case outcome ( it is not worth having more than 5 to 1 ratio of the more easily obtained outcome). This design is commonly employed in epidemiology ( and often called a case control design where cases have the disease and controls do not)
https://www.researchgate.net/publication/236671039_Epidemiology_An_Introduction
and here is a paper promoting it in sociology
Lacy, M. 1997. “Efficiently Studying Rare Events: Case-Control Methods for Sociologists” Sociological Perspectives 40:129
https://www.researchgate.net/publication/246035921_Efficiently_Studying_Rare_Events_Case-Control_Methods_for_Sociologists
To be useful you need to have some way of readily identifying cases and comparisons.
Book Epidemiology: An Introduction
Article Efficiently Studying Rare Events: Case-Control Methods for S...
Objective of the study should dictate the type of sample whether qualitative, case comparative or quantitative.
It seems that you are looking for triangulation in which both quantitative and qualitative methods are combined. Actually we need good amount of data to make some meaning out of it. It can be large population with small amount of observations or measurements or small sample with large number of measurements. Besides this we can take quantitative measurements and qualitative verification. in conclusion sample size is not a problem, we need to adjust methodology for answering research questions or testing hypotheses.
If you are using qualitative approach, a small sample would be adequate and you can use snowballing sampling technique. First you find one respondent and ask him/her to refer you to their friends and relatives who experience the same phenomenon. The same technique can be used for quantitative approach if you can obtain a large sample in comparison to the population of your potential respondents within the location of your research.
As your question is about sampling and not analysis (quantitative or qualitative), I would suggest non-probability sampling methods such as purposive sampling or snow balling where you grow your sample from a few known cases of involuntary bachelorhood. Findings are not generalisable but would still be valid given the rarity of the events.
This seems like it might possibly apply here:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1705515/
"Variance Estimation, Design Effects, and Sample Size Calculations for Respondent-Driven Sampling"
Sample sizes are relatively high, and I do not know the limitations with respect to whatever you are collecting, but it seems it might be worth your looking it over.
Random snowball sampling, using social media as an effective and far-reaching platform to obtain access to excessive data.
- Badges
- Science topic
- Similar topics
- Social Science