What pre-exam and post-exam quizzes can tell us about test construction

We describe a study of an instructor´s ability to write short problems that are effective measures of learning, using the traditional exam which uses formal problems as the standard for comparison. Pre-exam and post-exam quizzes containing short problems were administered in addition to the formal exam. Post-exam quizzes were written after the formal exam was graded, allowing the instructor some insight into the strengths and weaknesses of the students´ knowledge, while pre-exam quizzes were written with no a priori knowledge of student capabilities. Correlations and regression analysis were used to compare the pre-exam and post-exam quiz scores with the standard, the formal exam. The results of the analyses suggest that short problems must be carefully written if they are to reflect learning as measured by traditional problem solving exams which use longer problems to measure learning