Tuesday, June 17, 2003

for sociological geeks only

damn, i am good! check out my response to a post about working with a population with a small N. the person was concerned about utilizing an action research approach with a group of 20 people and having their participation in the process contaminate the data. they were not only going to assist in the research design, but were also going to be the "subjects"

"I don't think the issue is with the size of the population. I think is with the notion that by following certain procedures (not including "subjects" in the design), bounded by certain parameters (populations need to be larger than "X" or the results are not generalizable) that one arrives at something called, "knowledge" or "truth."

Another way to approach it would be to consider the entire endeavor to be the inquiry. I suspect you could learn a whole heck of a lot about this population by having them participate in the design. I try and work from the perspective that "everything is data." The question is, "when do I recognize that some phenomena are data?" This is a decision I have to make. So, does the data collection begin when I administer the survey? Or does it begin when I say, "let's work on this together?" The data are there all the time
-- it is NOT the methodology that captures the data. As such, how can it ever be "contaminated?" "Contamination" is just data that is not deemed to be relevant to the researcher.

I bet you would get much more out of it by simply being curious about these folks. Invite them to participate from day one, watch what they do, the decisions they make, the questions that they choose as best, etc. You can still administer the survey (or whatever) after it is completed. This will provide you with far more information in the long run.

p.s. in the event that you find the above utter nonsense, check out the comparative qualitative analysis, small N site .

No comments: