This work analyzes and compares the efficiency of alternative evaluation methods for the continuous assessment of pre-graduate students of Microeconomics. The efficiency is measured as the performance of students in the final exam. The study also analyzes the correlation between the performance in the final exam and in the continuous assessment exercises. The conclusions aim at the classification of the alternative methods in terms of efficiency, cost and predictability. We estimate a logit model with three alternative dependent variables: a dichotomous variable that takes a value of one if the student takes the final exam as scheduled, a similar variable that takes the value of one when the student pass the exam or the score in the final exam. In addition to the dummy variable that captures gender (1 for males) and the student´s score in the Continuous Assessment, the set of control includes also a dummy for each of the Evaluation Methods under consideration and their iteration term with the Continuous Assessment score. The interpretation of the size of the coefficients in a logit model is problematic. The presence of the iteration terms makes it even more cumbersome. However, the sign and significance levels yield enough information to evaluate the relative performance of the Evaluation Methods relative to the baseline, which is the Two-Exams method. The iteration terms estimated are, in general, negative and statistically significant. This is because the impact of the Continuous Assessment scores is not identical among the EMs. The methods used for the continuous evaluation during the course are key determinants of the performance at the end of the course.
Buscar en Dimension