“…A number of extraneous variables are viewed as contributing to this lack of validity. These variables include instructor personality and popularity, sex and rank of the professor, class size, class level, and expected grade (Rotem & Glassman, 1979). For example, ratings were found to be marginally higher but not statistically significant in small classes, discussion classes, classes in the humanities (Seldin, 1993), and electives (Arden, 1989).…”