Student-Faculty Evaluation

Using Multiple Measures in the Interpretation of Student Ratings of Instruction

This study investigated a multiple measurement approach to validating the inter pretation of student ratings of a medical school course. Concerns of priority to students were isolated from the ratings and data on student opinion collected by an open-ended question and student faculty documents. An interpretation consonant with all data was derived. Influences on student opinion, including demographic, academic, and personal characteristics of students, teaching practices, and the instructional context were explored through a comparison of the present class with a previous class. Ratings of the course by the two classes differed in the absence of significant differences in the students or teaching practices. Divergences in the ratings were attributed to changes in the instruc tional context reported in the multiple measures of student opinion. It was concluded that a multiple measurement approach allows greater precision in the interpretation of student opinion than is afforded by ratings alone.

Validity of students' ratings of clinical instructors

The validity of students' rating of clinical instruction was correlated with students' performance in a clinical examination. One-way analysis of variance showed that the students' ratings discriminated among instructors, and that the instructors were a significant source of variance in their students' adjusted examination scores on a written clinical simulation. There was no significant correlation between students' performance in the examinations and their ratings of the clinical instructors.

Lack of agreement between subjective ratings of instructors and objective testing of knowledge acquisition in a urological continuing medical education course

Objective scores from multiple-choice questions before and after a postgraduate course were compared to subjective ratings of the instructors at a 3-day seminar. The objective mean scores after the course were significantly higher than the scores before the course (p less than 0.0001). There was no correlation between test results and subjective ratings of instructors.