Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We use an online game with randomized treatments to study gender differences in the impacts of competition and subjective feedback. 5191 participants were randomly selected into 8 groups: players either saw a Top 10 leaderboard or not (competition), and within these, they received no subjective feedback, supportive feedback, rewarding feedback, or "trash talk" (feedback type). Seeing a leaderboard increases the persistence (number of games played) of all players, but only increases the performance (score) of male players. When the leaderboard is combined with supportive feedback, the performance of female players increases as well. This points to important heterogeneities by feedback type and individual characteristics and suggests that personalized feedback may be key for decreasing gender gaps, particularly in competitive settings such as STEM fields.