Student Ratings as Criteria for Effective Teaching

Within each of three Liberal Arts curriculum areas, this study examined intercorrelations among college instructors' self ratings and, in addition, various correlates of those ratings. Item pairs that correlated substantiallywere: "clear" and "approachable," "got students interested" and "how much students learned," and "clear" and "overall teaching ability." Linear composites of specific self ratings significantly predicted general self ratings, and linear composites of instructor characteristics predicted both specific and general self ratings, but the multiple correlations were low. The only instructor characteristics that related substantially to self ratings in all curriculum areas were motivation-related, i.e., their own enjoyment of teaching and liking for the subject matter seem to play an important part in instructors 1 definitions of good teaching and good learning.Among measures that might be used in the evaluation of college teaching are student ratings, peer ratings, and self ratings. A substantial and growing body of literature exists on student ratings (See reviews by Costin, Greenough, & Menges, 1971; and Doyle, 1975), but relatively little research has so far been published on peer or self ratings. Centra (1973) found peer ratings less reliable and more lenient than student ratings (Note 1) and self ratings sometimes lower than, sometimes the same as, but usually higher than student ratings. In the latter study he also noted a good deal of similarity (rank correlation of .77) in the way students and instructors rank ordered items, which suggests that instructors tend to be quite aware of their own relative strengths and weaknesses. Finally, he noted a curriculum effect, in which different patterns of self vs. student discrepancies occurred across broad subject-matter areas.Research on college instructors' self ratings, then, seems to have dealt almost exclusively with the comparison of self ratings to student ratings. No study could be located that looked for factors that define, correlate with, or predict self ratings. Because self evaluation is an important part of instructional diagnosis

Section: Discussionsupporting

confidence: 87%

Section: Instrumentmentioning

confidence: 99%

Self Ratings of College Instruction

Doyle

Webber

1978

“…Hofman and Kremer (1980) concluded that course evaluations were related to student and instructor attitudes toward higher 1 Discussions of TRF validity are complicated by the use of alternate validity models and languages (Abrami, Leventhal, & Perry, 1977). For example, compare similarities in the definition of discriminant validity with Doyle and Whitely's (1974) definition of TRF generalizability: "the extent to which sampling of student opinions can be done without regard to student characteristics" (p. 270). Furthermore, finding a sizeable relationship between student characteristics or student/teacher attitude similarity and ratings may not be considered by some as a limitation or source of bias in ratings (Feldman, 1977).…”

mentioning

confidence: 99%

Does the Attitude Similarity of College Professors and Their Students Produce “Bias” in Course Evaluations?

Abrami¹,

Mizener

1983

This study examined the relationship between student/teacher attitude similarity and teacher rating form (TRF) scores. Students (N = 345) completed a 24-item attitude scale, once for themselves and once for their instructors, and a 28-item TRF. The results indicated that TRF variance explained by total similarity was reduced by almost half when teacher effects were removed. Similarity of attitudes relevant to higher education was related to teacher ratings; similarity of irrelevant attitudes was not. The component scores (own attitudes, perceived attitudes) accounted for more TRF variability than assumed similarity. Of the component scores, only perceived attitudes accounted for a significant amount of TRF variance independent of other effects. We conclude that the unique variance explained by student/teacher attitude similarity is of minimum importance when evaluations are used to make gross distinctions (e.g., poor, acceptable, outstanding) among instructors.

“…Identifying generalizable dimensions of teaching from these items is important for effectively summarizing evaluations to instructors as well as theoretically important in the scientific study of teaching. Although such dimensions may relate differentially to the goals of teaching, such as student learning (Doyle & Whitely, 1974), they provide parsimony by reducing the number of variables needed to describe teaching behaviors. Consequently, quite a number of studies have attempted to organize rating items into a set of independent dimensions through factor analysis.…”

mentioning

confidence: 99%

Implicit Theories in Student Ratings

Whitely

Doyle²

1976

Self Cite

Identifying generalizable dimensions of teaching from student ratings data has been both practically and theoretically important in the study of teaching. However, an examination of the methods employed reveals that many studies may have identified dimensions which were implicit in the student raters rather than the instructors. The current study compared various methods of identifying teaching factors and found substantive similarity of the dimensions defined from students' implicit theories and various types of correlational data. The results are interpreted both with respect to the rating process and to the nature of the factors identified in previous student ratings research.