Centre for Development Economics

Department of Economics
Delhi School of Economics


SEMINAR ANNOUNCEMENT

Performance Observability and Feedback Avoidance among Students – Evidence from a Large-Scale Field Experiment on an EdTech Platform 

 

Presented by

Garima Agarwal 

Azim Premji University, Bangalore

with
Gerhard Riener

Date & Time
Thursday, November 5, 2026 | 3:00 PM

Venue
Amex Room

Abstract

Many online learning platforms use public leaderboards to motivate students, a default that applies to everyone. Where gender stereotypes make public failure costly, this default may change who chooses to participate. We test this in a pre-registered field experiment on a large Indian learning platform (19,932 student-wave observations), independently randomising whether a student’s result is publicly visible and whether the task is announced as difficult. On average, public visibility does not change how students respond to difficulty, and this null is precisely estimated. The average conceals a domain-specific pattern: when hard tasks are public, students participate less in gender-incongruent subjects (girls in mathematics, boys in English), and the shift is concentrated among girls. A joint test rejects that the effect is the same across domains. Girls respond to task domain, but less to positive performance signals, suggesting that gendered avoidance patterns do not naturally resolve with repeated exposure. Public-by-default leaderboards may therefore leave aggregate participation unchanged while discouraging girls from the domains where their participation matters most for later education and labour-market outcomes.

All are cordially invited.

 


← Back to Seminars