RESUMEN
The influence of group norms is often tested via people's norm perceptions. Yet, people's group norm perceptions can be inaccurate, which raises the question to what extent effects of perceived norms indicate real group influence. The present study sought to obtain a better understanding of the value of group norm perceptions in social influence research. We used (longitudinal) data collected in 51 primary school classrooms in the Netherlands (Grades 3-6) and examined how majority of children's (N = 779; aged 7-13 years) perceptions of the anti-prejudice norm of their classroom peer group affected their ethnic outgroup attitudes both concurrently and over time. We divided these perceptions into a consensual and a unique part and investigated the moderating role of ingroup identification. Results showed concurrent effects of the consensual and unique norm perceptions, but a longitudinal effect of the consensual perceptions only. Classroom identification increased the concurrent effect of unique norm perceptions but decreased their longitudinal effect. Our findings indicate that norm perceptions can be important sources of actual group influence as long these perceptions are consensually shared, and that especially high identifiers rely less on their unique norm perceptions over time.