| Abstract: | Two experiments (total N = 195) examined third-party inferences about friendship and shared preferences from reported dyadic information. Four-year-olds (N = 48) inferred friendship between individuals based on reports of prosocial behavior, and similarity, but not based on arbitrary links. Children privileged prosocial behavior over similarity when asked to adjudicate between the two. Adults (N = 120) were tested online and showed the same overall pattern of inferences. Furthermore, 4-year-olds (N = 27) expected individuals who had engaged in prosocial behavior to be playmates as well as friends, and to share preferences for novel games, but not novel foods. These findings shed crucial light on preschoolers' third-party friendship inferences, and add to our knowledge of their concept of friendship. |
