However
Data were collected after results were known (selection bias).
Several post FCIs were administered as exams (pre-test as diagnostic).
Data were not always matched.
Differences among different interactive curricula were not examined.
Students’ understanding of specific concepts was not probed.