Working Paper: CEPR ID: DP13541
Authors: Nagore Iriberri; Pedro Reybiel
Abstract: We study gender differences in willingness to guess using approximately 10,000 multiple-choice math tests, where for half of the questions, both wrong answers and omitted questions are scored 0, and for the other half, wrong answers are scored 0 but omitted questions are scored +1. Using a within-participant regression analysis, we find that female participants leave significantly more omitted questions than males when there is a reward for omitted questions. This gender difference, which is stronger among high ability and older participants, hurts female performance as measured by the final score and position in the ranking. In a subsequent survey, female participants showed lower levels of confidence and higher risk aversion, which may explain this differential behavior. When both are considered, risk aversion is the main factor explaining the gender differential in the willingness to guess. A scoring rule that is gender neutral must use non-differential scoring between wrong answers and omitted questions.
Keywords: gender differences; willingness to guess; risk preferences; confidence; perceived ability in math; natural field experiment
JEL Codes: C93; D81; I20; J16
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| gender (J16) | willingness to guess (D80) |
| willingness to guess (D80) | performance (D29) |
| gender (J16) | omitted questions (C34) |
| omitted questions (C34) | performance (D29) |
| risk aversion (D81) | willingness to guess (D80) |
| scoring rules (C52) | performance disparities (I24) |