Critics have referred to the influence of an instructor’s personality on student ratings of instruction as the “Dr. Fox effect” or “educational seduction.” In reviewing earlier research, we found inconsistencies from study to study and conducted a meta-analysis to attempt an empirical integration of this literature. Overall, we found that instructor expressiveness had a substantial impact on student ratings but a small impact on student achievement. In contrast, lecture content had a substantial impact on student achievement but a small impact on student ratings. We discuss these findings and offer a new interpretation of the implications of the educational seduction literature for the validity of student ratings

Two experiments were conducted to determine the effects of teacher standards for assigning grades on student ratings (and also achievement). Male and female college students (n = 143) participated in Experiment 1, in which we factorially manipulated instructor expressiveness (high, low), lecture content (high, low), and grading standards (B, C+, C). College students (n = 278) also participated in Experiment 2, in which we factorially manipulated student incentive (high, low), instructor expressiveness (high, low), and grading standards (B+, B, C+, C). Standards failed to affect student achievement in either experiment. We did find significant (p < .05) effects of standards on ratings, but not consistently for all types of ratings, instructors, or differences in standards. Moreover, we consider the size of the differences to be relatively unimportant when ratings are used to make gross distinctions between teachers.

Previous "educational seduction" research suggested that teacher differences in expressiveness controlled the degree to which lecture content affected student ratings differently from student achievement. We attempted to replicate statistically this Expressiveness X Content X Measures interaction in a 2X2X2X2X2 factorial design (N = 245) which investigated four simulated classes. The interaction was found for the high-incentive/no-study-opportunity class and the high-incentive/study-opportunity class, which most resembles typical classes, but not for the low-incentive/study-opportunity class or the low-incentive/no-study-opportunity class, which most resembles educational seduction research. In only the high-incentive/no-study-opportunity class did probes of the interaction replicate education seduction research in which content affected ratings and achievement similarly only for low expressiveness.The validity of student ratings of teacher effectiveness continues to be a major concern. Naftulin, Ware, and Donnelly (1973) found that an entertaining, charismatic lecturer who spoke deliberate nonsense received surprisingly high evaluations from his audience. This phenomenon was labeled educational seduction. Ware and Williams conducted follow-up research (Ware & Williams, 1975, 1977Williams & Ware, 1976, 1977 using a common core design. Students viewed one of several videotaped lectures, which were delivered by a professional actor. The lectures differed on two variables-instructor expressiveness (e.g., enthusiasm and humor) and amount of lec-This research was supported by grants from the University of Manitoba research board and the Canada Council (S74-1078; S76-0345). The authors are indebted to Wenda Dickens for her very able assistance in the collection and analysis of the data. The authors would like to thank W. J, McKeachie and R. Williams for their comments on earlier drafts.

Researchers know little about determiners of section selection in multi-section college courses. Studies on teacher evaluation and on the validity of teacher rating forms have often assumed section to section equivalence of students assigned by customary registration procedures. To investigate the section selection process, a questionnaire containing items on personal history, reasons for section selection, and sources of information about the instructor was administered to 1,188 undergraduate students in multi-section first year and advanced psychology courses. Major findings were: (1) students significantly differed across sections on biographical variables and on section selection reasons, (2) time at which class was scheduled (classtime) and teacher's reputation were the primary reasons for section choice, (3) teacher's reputation was less important than classtime for first year students, but comparable to classtime for advanced students, and (4) reports from other students and published ratings were, respectively, the first and second most frequent source of instructor reputation information.

Instructor reputation, defined as student expectations of an instructor's teaching ability, was investigated to determine its influence on student ratings and achievement. Two teaching behaviors, instructor expressiveness and lecture content, were combined with reputation in a 2 X 2 X 2 factorial design to assess interaction effects. Students read an introductory statement about an instructor (reputation), viewed a videotape lecture, rated the instructor, and took a test. Results indicate that reputation interacted with expressiveness but not content, in which students rated the positive, high-expressive instructor more favorably than the negative, high-expressive instructor. The results suggest a complex relationship in which reputation interacts with some teaching behaviors and affects some student outcomes.Expectancy effects have been reported in a variety of contexts from discrimination-training experiments with rats (Rosenthal, 1969) to impression formation experiments with humans (Asch, 1946;Kelley, 1950). Rosenthal and his colleagues Rosenthal & Rubin, 1971) have created a controversy about expectancy effects in the classroom by showing that teacher expectations of student ability, apart from actual ability, can affect the student at a later stage in his/her academic career. Students labeled as academically gifted performed better at the end