Monday, September 15, 2008

Qualitative Research, Part 3

The objectives of sampling in qualitative research differ from those in quantitative research. The reason for that difference is that with qualitative research we are not concerned with our ability to generalize the information we collect to a larger population, which is the entire point of the sampling methods used in most quantitative studies. The sampling in qualitative research is called “purposive.” This means that it is intended to describe the processes of a phenomenon, and not its distribution. Our goal is to obtain information-rich cases that allow us to look in-depth into the phenomenon our study is examining. Again, the entire idea is to look at interpretation and lived experience.

Liamputtong and Ezzy (1) list a number of different sampling methods that may be used in qualitative research. They include the following:

Extreme or deviant case sampling: where we select cases that are unusual or distinctive in order to illustrate the processes under examination.
Maximum variation sampling: where we try to select cases that provide for wide ranges in the experience or process under examination.
Homogenous group sampling: here we select a sample to minimize variation and to maximize homogeneity to getter better depth in the process under examination.
Typical case sampling: where we select a case because it is not in any way unusual or atypical.
Critical case sampling: where cases are selected to illustrate processes where the processes would be thought least likely.
Criterion sampling: all cases meet some set of pre-specified criteria. This allows in-depth study of a phenomena, providing rich information.
Stratified purposive sampling: where we select cases from some previously identified subgroups.
Snowball or chain sampling: where after identifying an initial group of respondents, we ask them to suggest other possible participants.
Opportunistic sampling: where as a project develops and as new information is discovered, opportunities to go in new directions arise that are taken advantage of.
Convenience sampling: where we invite whomever we can easily locate and involve. It is often confused with purposive sampling, but Liamputtong and Ezzy (1, p. 49) note that this is a mistake because it does not theorize the sample.
Volunteer sampling: where we obtain participants by advertising for them.
Triangulated sampling: we can combine any of the above as strategies for creating our project develop.

As always, the best method to use is the one that best helps you answer the questions you are asking. It is the best tool for the situation, and the decision to use sample method as opposed to another will be driven by the needs of your project.

1. Liamputtong P, Ezzy D. Qualitative research methods, 2nd edition. South Melbourne, Australia; Oxford University Press, 2006

No comments: