A Comparison of the Relative Efficiency and Validity of Tailored Tests and Conventional Quizzes

Two types of computer-administered unit quizzes in a systematic pathology course for second-year medical students were compared. Quizzes composed of questions selected on the basis of a student's ability (tailored or adaptive testing) had higher correlations with the final examination than did quizzes composed of questions randomly selectedfrom topic areas (conventional testing). Further, tailored tests obtained optimum correlations with 25% fewer test items.