Student ratings of problem-based tutors at Harvard Medical School in 1
987-1988 indicated that tutors were seen by first year students as mos
t helpful in encouraging critical appraisal of information, questionin
g and probing the reasoning process, and balancing basic science and c
linical discussions. Narrative comments about the best tutors further
indicated that such tutors were most helpful in balancing student-dire
ctedness with guidance and sharing professional expertise without lect
uring. A concurrent survey of faculty tutors indicated that tutors rat
ed these same skills as important. Comparing students' ratings of help
fulness with tutors' ratings of importance, two discrepancies emerged.
Tutors rated student direction as more important than students rated
its helpfulness. On the other hand, tutors rated the ability to balanc
e clinical and basic science discussion as the least important skill w
hile students' ratings indicated that it was extremely helpful. These
disparities suggest points of discussion and skill building for facult
y development programming.