Maximum Points and Effort on the Test - Evidence from a Natural Field Experiment

Students often exert low levels of effort on the test, especially in low-stake tests which may paint an incorrect image of their true learning levels. Offering pecuniary and non-pecuniary incentives may be effective in increasing such effort but often not feasible. Increasing the maximum points assigned to the test may be a supposedly irrelevant factor that can frame it as more valued, and nudge students to exert more effort on the test. Using a natural field experiment, we randomly assign 1235 students to a test framed as either carrying maximum 20-points or 100-points, and study its impact in both high-stake (comprising non-risky and one risky question) and low-stake (comprising only non-risky questions) test settings. While we find a small and statistically insignificant average treatment effect in both settings, we find positive treatment effects (1) among lower ability students in a low-stake test, (2) among higher ability students in a high-stake test only when the question carries the risk of a negative penalty. We discuss increased mental effort as the mechanism driving these effects.

Working paper available on request