|
QUALITATIVE RESEARCH PART TWO: SAMPLING AND SELECTION
These terms are used differently by different authors. Basic idea is you study a subsection of a population.
Quantitative research looks at sampling in a probabilistic manner - try to get a representative sample so results will generalize to whole population.
Some qualitative people use sampling in this sense while the choice of defining the kind of people to be studied is selection. Other qualitative researchers look at the issue more globally - any choosing of group to study is sampling.
Regardless, qualitative people add some unusual ways of choosing whom to study. This choosing is a recursive process- dynamic and ongoing. The choices of who to study next are product of what is being found, not the initial plan.
1. Maximum variation - relevant dimensions vary widely in group (clearly see extremes, not averaging together extremes so result represents the "average.") High activity/low activity college students. 2. Snowball approach/networking - each person studied is chosen by previous participant - thus will see linkages between people. (May be the only way to find members of group - such as Christian families that celebrate Jewish holidays). 3. Extreme case - studying one or more people at some extreme. Don't study the average or the opposite. Example: Snake Handling Southern Baptists (probably hard to find!) Possible to have N=1 (with any qual study). 4. Typical case - decide what characterizes typical and then go looking for that person. 5. Unique case - very rare combinations of things - usually discovered fortuitously. (Snake Handling Episcopalian!) 6. Ideal case/bellweather - perfect situation: "If it won't work here, it won't work anywhere." Using humanistic therapy with president of humanist association. 7. Negative case - look for an exception to the emerging rule/hypothesis. Used in analytic induction approach - goal is to refine generalizations by setting out to find when and where it doesn't hold true. Example - when do teachers not pay attention to time spent by child in hall? (I found it was when child was difficult in class, but didn't cause major problems in hall).
Lots of other variations are mentioned in Patton and other sources. The goal is not to know them all, but to see wide variety of possibilities, to open up to even new ways of choosing not in literature that may be best way to study people in a situation. Selection and sampling methods allowed to emerge during study; what is needed, rather than stick with initial choices.
Very important to say how and why you sampled in a given manner - keep extensive notes on decisions and why the decision was made - go into detail, provide a strong rationale for choices made.
I selected 1 school out of 4. Interviewed all 4 principals and toured all 4 schools. Chose school to which I was given most access with fewest restrictions. Also school had widest variations in hall activities (very high to virtually no activity).
Then sampled in hallway - different locations with camera to find most varied activity and least self-conscious/guarded behavior. Where? Turned out to be restrooms/drinking fountain area. (See chart in dissertation, p. 31) Later used snowball/network approach in choosing children to interview - wanted to see them in groups of friends.
Return to Qualitative Research Outline
|
|
|
