Comparison of course evaluations by random and volunteer student samples.

Abstract

Ratings of the effectiveness of courses are usually based on students' voluntary responses to evaluation questionnaires. It is a rare course for which all students complete evaluation forms. This self-selection may result in an overrepresentation of students who are highly pleased or greatly disgruntled, thus biasing the evaluation. Responses of an unbiased random sample of students were compared with the responses of a self-selected sample for the evaluations of two medical school courses. There was no evidence of bias introduced by self-selection for either mean ratings or the variance of responses on a question-by-question basis for eight questions included in the evaluations of both classes. When responses were investigated across items, the patterns of differences between the random and voluntary samples appeared to be unique to each class.