The first part of this study replicates an earlier study (D. Waechter, I. Newman, and L. Rosenkoetter, 1998) that used a procedure for authentic assessment of students in a master's level research class, a class that is often feared and avoided by students, by means of a study on humor, since that content would be somewhat incompatible with anxiety. Students (n=88) completed a questionnaire on personal interpretation of humor, demographic information, and the Temperament Profile Survey. Students also completed a course evaluation indicating subjectively evaluated growth in understanding and critiquing research. Part two of the study used the general linear model to examine the relationships among view of humor, temperament, and course evaluation. Results indicate that there appear to be relationships between how one evaluates a course subjectively and how one evaluates the appropriate use of humor and an individual's temperament. Evaluations are also complicated by the interactions of age and temperament. Effect sizes were small but consistent. These findings increased investigators' awareness of the complexity of looking at students' evaluations of courses. (Contains 20 references.) (SLD)