Within the last several years a consensus of opinion has developed that the use of student ratings as a means of evaluating college teaching is increasing[ 1-41. In some instances student evaluations have become the sole basis for administrative decisions about the effectiveness of college teaching, including decisions related to promotion, tenure and salary increases [3]. Such use of student evaluations has caused considerable concern. Zelby[5, p. 12671 expresses a number of the concerns well when he writes:One major cause for the growth of the SFE (student-faculty evaluation) controversy is the trend toward formal, quantitative use of the results of the evaluations in determinations of faculty promotions and salaries. There are at least three reasons for this trend: (i) SFE provides documented, precise, numerical evaluation of instructors; (ii) it tends, thereby, t o relieve academic administrators from the responsibility of exercising judgment about teaching performance and ability; and (iii) it tends to constitute proof that something, indeed, is being done t o improve teaching.Zelby[S, p. 12671 further adds:Until quite recently, the primary purpose of SFE was to provide information that could be helpful t o students and teachers alike. The current tendency to use these evaluations in a quantitative and formal sense changes the complexion of SFE from that of a helpful collection of information t o a device that could become detrimental t o education.After reporting data which indicates that a college instructor can intentionally adapt his teaching technique to obtain a good or bad student evaluation, Zelby[S, p. 12671 points out that "teaching for a good SFE may not be consistent with the best educational practices and may not challenge them (students) to go where they have never been." While there are a number of research issues buried in Zelby's concerns, at least one issue seems critical: namely, that of the validity of the student rating instruments used.The question of validity, or the extent t o which a student evaluation instrument measures what it purports to measure (i.e., college teaching effectiveness) is closely tied t o the question of instrument reliability since reliability affects validity. McKeachie [6] has reported significant reliability coefficients for a number of student rating instruments.
493Science Education, 59(4): 493498 (1975)
