We found significant cross-level interactions between participant gender and T1 gendered
classroom-level aggression scores for both forms of aggression at T2 that indicated that the children were more influenced by their same-gender than by their opposite-gender peers.