Comparing students' feedback about clinical instruction with their performances.

Validation of students' feedback as a measure of teaching effectiveness has been problematic for courses teaching clinical skills. This is true in part because establishing a valid and reliable method of assessing students' mastery of clinical skills has been a stumbling block. Reported here is the correlation of students' performances on an objective structured clinical examination (OSCE) with previously and independently collected feedback from students. In 1987-88, 190 second-year medical students at the University of Minnesota Medical School--Minneapolis spent one fourth of a second-year clinical skills course on neurology randomly assigned to one of four teaching sites--hospitals A, B, C, and D. Following their rotations, 180 of the students completed usable feedback forms. The students were consistently and significantly more positive about the teaching at hospital A. At the end of the year, all 190 students were tested using an OSCE having 20 stations, four of which presented neurologic problems. The students who had the neurology course at hospital A performed better on all four neurology problems, and differences were statistically significant for two of the problems. Feedback in this case accurately reflected a more effective teaching program.