PRECLINICAL COURSE-EVALUATION METHODS AT US AND CANADIAN MEDICAL-SCHOOLS

Background. Despite the apparent prevalence of the use of course evalu ations by medical schools, course and curriculum evaluation have gone relatively unnoticed in the medical education literature. Method. In t he fall of 1993, a 17-item questionnaire was mailed to all 141 U.S. an d Canadian medical schools to elicit information concerning any course -evaluation systems in place in the schools' preclinical curricula. Re sults. A total of 101 schools (72%) returned usable questionnaires. Of these, 79 reported having a centralized course-evaluation system and 56 used oversight committees consisting of administrators, faculty, an d students. Beyond the use of written questionnaires (reported by 100 schools), course evaluation practices varied widely. Eighteen schools reported that questionnaire content was the same across courses, while 56 used a common core of items with modifications for specific course s. The frequency and timing of questionnaire distribution varied from once at the final examination to weekly during the course. Summarized course-evaluation results were made available to departments (91 schoo ls), the oversight committee (50 schools), the administration (47 scho ols), and students (13 schools). The feature most frequently cited as being a positive aspect of the course-evaluation process was student i nvolvement (23 schools). Most frequently cited as areas of concern wer e low response rates to questionnaires (20 schools) and the need for s impler, more reliable methods of data collection (13 schools). Seventy schools reported specific types of changes that had resulted from cou rse evaluation. Conclusion. The results confirm that course evaluation via student questionnaires is ubiquitous in North American medical sc hools. Most schools used centralized systems, but individual schools h ad developed their own combinations of technique and organization. Thi s lack of uniformity may be due to the sensitivity of evaluation proce sses to local contexts, but it may also be due to the lack of literatu re on the subject.