The Psychological Effect of a Math Signal

This paper tests whether barely obtaining a pass score in at least one of two midterm tests has an effect on subsequent achievement in a Math course. To estimate the effect, we created a novel dataset by linking administrative and survey data on students at a medium size Spanish University and used a regression discontinuity design in which the cutoff is 5, i.e. the pass score in the national grading system. Although obtaining a score just equal to or barely greater than 5 in midterm tests has no immediate consequence for students, it may have a psychological effect by acting as a (de)motivating signal to study and pass the course, with the sign of the effect being unclear ex-ante. We find that obtaining a pass score in at least one midterm has a positive effect on the final exam score. The result seems to be explained by students' study strategy, i.e. the ability to obtain a value in the final exam score that when averaged along with the midterm tests scores leads to an overall passing score in the course. Overall, our results suggest that partly unexplored psychological mechanisms may help us deepen our understanding of the determinants of achievement in higher education.