Social Conformity or Complementarity: Peer Effects in Online Learning

Activity: Talk or presentationPresentation at conference/workshop/seminar


We study how revealing peers’ past learning behaviors in a randomized controlled trial (RCT) affects individuals’ subsequent online learning participation and performance in a large online teacher training program. Treatment groups receive text messages specifying tenured or untenured peers’ historical performance. Applying a difference-in-differences (DID) model, we find the treatment effects on in-class participation behaviors and after-class learning outcomes are mainly driven by knowing peers’ pre-treatment attendance duration rather than participation propensity. Moreover, we utilize the non-competitive relationships between tenured and non-tenured teachers in China’s education system to separate social conformity from complementarity sources of peer effects. Additionally, underperforming trainees increase attendance and stay longer in the post-treatment lectures, while outperforming trainees reduce the effort in participation regardless of tenure status. However, tenured trainees exert more effort in homework submission irrespective of their relative historical performance, while non-tenured trainees’ homework submission and performance are unchanged. Other factors like teaching loads, school type, college major, and familiarity with work location also impact the responsiveness to peer’s nudge information. Our findings have important practical implications for motivating participation and improving online learning outcomes.
Period18 Nov 202320 Nov 2023
Event titleSouthern Economic Association 93rd Annual Meeting
Event typeConference
LocationNew Orleans, United States, LouisianaShow on map
Degree of RecognitionInternational


  • Online Learning
  • Field Experiment
  • Peer Effects