Similarity over Status in Partner Selection under Competition

This repository contains the data, analysis scripts, and experimental code for "Similarity over Status in Partner Selection under Competition." The paper reports two pre-registered experiments examining whether gender homophily persists when it conflicts with strategic incentives in competitive partner selection.

Study 1 (N = 961) tests partner selection when participants must infer performance from gender cues alone, with a separate pre-registered belief-elicitation task (N = 598) validating that the arithmetic reasoning task evokes gendered performance expectations. Study 2 (N = 997) introduces explicit performance ratings for potential partners, creating direct tension between homophily and strategic optimization.

Across both studies, gender homophily emerged as the dominant pattern in competitive partner selection. In Study 1, participants most frequently selected same-gender partners as both competitors and teammates. In Study 2, same-gender preferences persisted even when explicit ratings identified demonstrably better opposite-gender alternatives. Female participants exhibited stronger same-gender preferences than males across both studies.

Repository Contents
Data files:
S1_data.csv — Study 1: partner selection data, N = 961
S2_v1.csv, S2_v2.csv, S2_v3.csv — Belief-elicitation validation for Study 1: three framing conditions (baseline, individual competition, team competition), combined N = 598
S3_data.csv — Study 2: partner selection with explicit performance ratings, N = 997

Analysis scripts:
S1_Analysis Code.R — analyses for Study 1
S2_Analysis Code.R — analyses for the belief-elicitation validation
S3_Analysis Code2.R — analyses for Study 2