Posted by shar on July 14, 2000, at 23:02:19
In reply to Re: Social Phobia --more thoughts, posted by noa on July 14, 2000, at 12:49:32
Regarding the impact of a non-random sample, we always had a statement at the end of the article or presentation that the results should not be considered representative of the population in general due to lack of randomness, etc.
In my grad school, and the vast majority of social psych research I've seen at conferences and seminars, the sample is most often volunteer or self-selected. So, participants may share some common characteristic that differentiates them from people in general. For ex., willingness to give out info about themselves, higher level of self-disclosure. Usually not something that would really separate them as a unique group--which would mean that something about the survey or methodology really drew this particular type of person in (such as all respondents were fire-fighters, or all respondents ran marathons).
The researchers could probably fashion a control group of sorts by giving the survey to a group of students to find out how "people in general" scored on some items, and then do a sort of cross-group comparative analysis.
Generally, everybody ignores the issue of self-selection and continues on with their next study.
That's been my observation at least.
S
poster:shar
thread:40327
URL: http://www.dr-bob.org/babble/20000708/msgs/40481.html