On Monday NCID hosted a seminar from Professor Antonio Cabrales of University Carlos III and University College of London. While better known for his theoretical work in Economics, Cabrales presented his applied research on the effect of feedback on student performance at the university level. 

The essence of the project was to investigate whether informing students of their standing relative to their cohort average would improve or worsen their performance on exams, course, and grade-point average. Beginning in their second year, students in the treatment group were offered the opportunity to view in which decile of their cohort the are currently in, based on exam performance. They received this information twice a year, during their second, third, and fourth years.

The study had several significant conclusions after analysis. Perhaps the key finding was that the effect was not uniform across all students. The dominant effect was a negative, represented by an initial large drop in performance, followed by a gradual recovery. Cabrales suggested the following interpretation from a student's perspective, "It seems that if I had good news with respect to how I thought I was doing, that makes me slightly less likely to pass exams. Getting this information that I am doing well is making me relax a bit and I am more likely to fail." 

Secondary findings included that student evaluations are not a reliable index of teacher performance and that class size does not improve student performance, but has a significant effect on students' satisfaction.