1. Faculty Ratings and Student Grades: A Large-Scale Multivariate Analysis by Course Sections.
- Author
-
Brown, David Lile
- Abstract
The purpose of this dissertation is to provide evidence bearing on the question of the influence of the grades students receive on their ratings of the college teachers who gave them those grades. Specifically, certain characteristics of the grade distribution within each course section are evaluated as predictors of the students' ratings of the teacher of that course section. Multivariate techniques were employed to evaluate data across an entire university. Over 30,000 anonymous student ratings of 2,360 course sections were collected after students had received final course grades, and without student or administrator knowledge that the ratings would be used in the study. Factor analysis was used to reduce the eight-item rating instrument to a single criterion variable. Subsequently, stepwise multiple regression analysis was used, both to reduce an initial battery of predictors to an optimally reduced subset, and to test the incremental importance of certain grading variables as predictors of the criterion.. The primary implication of the study is that there is a relationship between grades and ratings, but it only accounts for about nine percent of the variance in the ratings. There is a significant bias, but factors other than grades must also be influencing student ratings. Whether or not these other factors are valid measures of teaching effectiveness remains to be determined, but one seemingly invalid factor (the grading bias) has been identified. (Author/RC)
- Published
- 1974