Feedback Spillovers across Unrelated Tasks

Abstract

In this paper we describe the results of a laboratory experiment involving 130 students of Lomonosov Moscow State University. The participants were asked to perform unrelated tasks involving different skills: a test of analytical skills and a capitals quiz. After the first task, participants in the treatment group are given feedback on whether they did better or worse than half of the other participants in their group. This feedback is uninformative about the new task, as success or failure in the analytical ability test is not related to whether the individual knows the countries and their capitals. It is revealed that students respond to negative feedback: participants from the group that received negative feedback significantly lower their probability of success in the next task. The peculiarity of the work is to conduct the experiment within 2 days in order to avoid the occurrence of the “hot hand error” described in previous studies. Statistical tests and regression analysis as well as bootstrap are used to test the hypotheses. A time-stable effect of uninformative relative to the new task feedback is obtained.

Keywords

• feedback
• experiment
• asymmetric belief updating
• confidence
• bootstrap