I have surveyed 100 people. Based on my results I would love to interview 2 people from all the different backgrounds represented in my data. I have representatives from 9 different backgrounds therefore ideally I need to interview 18 people. However, only 15 people provided me with their info and gave me permission to contact them for interview. I can only select 6 people from that sample, so only 3 backgrounds will be represented. I do not have access to the others. Am I doing purposeful sampling or convenience sampling? I am a little confused here.
The defining characteristic of using a purposeful sample is that you used a specific goal (purpose) as the basis for your sample. A convenience sample would not have any purpose other than ease of data collection -- for example, if you used the first six people who responded.
If you decided that you want to compare three backgrounds with two interviews per background, that would indeed be a purposeful sample.
FYI may favorite source on purposeful sampling is M. Patton's book, Qualitative Evaluation, which lists and discusses several different kinds of purposeful sampling.
Dear Yulia,
This is an example of a mixture of purposive, convenience and consent samplings. It is not a pure purposive or convenience sampling because there are several agents who decide how the sample is selected, and not only the interviewer, but also the interviewed ones decide.
Hi Yulia,
it seems much more important to me that you have described exactly what happened rather than applying the 'right' label to it. Far better than the other way round!
Anyway, for my money, I'd say that pretty much all social research has an element of convenience unless participants are compelled (which would probably be unethical anyway), so the idea that not everyone was available or willing doesn't seem to warrant a special label of "convenience". But it is purposive because you are looking for particular cases to drill further into particular concepts from your initial phase.
I don't know if methodologists would agree - I'm just going on how I read it, I'm afraid, so not authoritative.
Thank you, Mariano and Peter.
Since my dissertation requires I label things, can I call it mixed purposeful sample?
What your dissertation requires is a question for your school / supervisor.
Dear Yulia,
You could call it mixed purposeful sample, but the important thing is that with such sample you could not do statistical inferences based on controlled probabilistic properties.
The defining characteristic of using a purposeful sample is that you used a specific goal (purpose) as the basis for your sample. A convenience sample would not have any purpose other than ease of data collection -- for example, if you used the first six people who responded.
If you decided that you want to compare three backgrounds with two interviews per background, that would indeed be a purposeful sample.
FYI may favorite source on purposeful sampling is M. Patton's book, Qualitative Evaluation, which lists and discusses several different kinds of purposeful sampling.
- Badges
- Science method
- Similar topics
- Mathematical Sciences
- Statistics
- Applied Statistics
- Survey Design