THE EFFECTS OF STUDENT, COURSE, AND INSTRUCTOR CHARACTERISTICS ON STUDENT RATINGS OF INSTRUCTION
Abstract
The present study was conducted to determine if there are systematic biases affecting students ratings of instruction. Stepwise multiple regression analysis was employed to examine the influence of eighteen student, course, and instructor characteristics on student ratings of five aspects of instruction. All the students enrolled at Grinnell College in the spring semester of 1978 were the subjects. The instrument was an end-of-course evaluation rating form administered the end of the semester to all classes (207) except practica; therefore the entire faculty at Grinnell College participated in the study. Five criterion variables were derived from a factor analysis of the end-of-course evaluation instrument. The factors were designated Course Organization, Rapport and Assistance, Stimulation, Administrative Details, Presentation, and Administrative Details. The predictor variables used in the regression analysis were the student characteristics Expected Grade, Year in School, Goals Shared, Required vs. Elective, Difficulty, Pace, Workload, Percent Class Meetings Attended, Percent Reading Completed, Percent Other Requirements Completed, Grade Point Average, and Major Course; the course characteristics Class Size, Course Level, and Subject Area and the instructor characteristics Years of Teaching Experience, Rank, and Tenure. The eighteen predictor variables were used to calculate five regression equations to predict the ratings for each of the five aspects of instruction (factors). (1) Overall, the student variable Goals Shared was found to be the most potent predictor of ratings. Sharing the instructor's goals was predictive of higher ratings on the Course Organization, Rapport and Assistance, Stimulation and Presentation factors. (2) The student variable Pace played a small but significant role in prediction for Course Organization, Stimulation, and Administrative Details. Faster paced courses received higher ratings. (3) The course variable Course Level was also found to be a significant predictor. Lower level courses received higher ratings on the Stimulation factor. (4) The course variable Class Size and the student variable Expected Grade were significant predictors only for the Rapport and Assistance factor, with smaller classes and higher expected grades being predictive of higher ratings. (5) The student variable Percent Other Requirements Completed, the instructor variable Years of Teaching Experience, and the course variable Subject Area were all significant predictors only for the Stimulation factor. Students completing more of the course requirements or teachers having more years of experience was predictive of higher ratings. The results of this study showed that a combination of student, course, and instructor characteristics could explain a significant amount of the variance for all five aspects of instruction measured by the student ratings. The evidence suggests that the five factors are not influenced by the same variables, and that the influence of a variable is not necessarily in the same direction for each factor. The results also suggest that two reasons for some of the contradictory results in the literature may be (1) the use of overall ratings rather than ratings of more specific aspects of teaching behavior for the dependent variable, and (2) confounding of commonly studied predictor variables.
Degree
Ph.D.
Subject Area
Higher education
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.