This study, using a test group of five colleges--2 state colleges, a liberal arts college, a multipurpose college, and an urban community college--investigated the effects of student feedback on improving teaching at the college level. Students rated instructors with a 23-item "Student Instructional Report," which included course objectives, instructor preparation and organization, student-faculty interaction, student effort, and course difficulty and scope. Three test groups--the feedback (treatment) group, the nofeedback (control) group, and the posttest group were analyzed using multivariate analysis of variance (MANOVA) and regression analyses. Information retrieved included effects of subject area of the course, sex of the instructor, and number of years of teaching experience. Results of the study indicated that after reviewing student feedback, instructors who rated themselves more favorably than had their students were more apt to effect change in their teaching patterns. Evidenced also was the tendency of greater instructor change with interpretive assistance in analyzing results, and with more time in which to implement the change. While actual results and interpretation were left to the instructor, the ease with which student ratings can be employed seemed to warrant continuation of this method of improving college teaching. (JCW)