The present study examined to what extent the number of friends and their social and personal characteristics were related to peer victimization in adolescence. Participants were 2,180 adolescents (1,143 girls), aged 11-18 (M = 14.2), who were classified as victims, bullyvictims, or non-involved (i.e., adolescents who neither bullied others nor were victimized by others). Three types of friends were distinguished: reciprocal friends, desired friends (who were unilaterally nominated by a target adolescent) and choosing friends (who unilaterally nominated a target adolescent). Between-group comparisons of the three types of friends showed that victims had fewer reciprocal and choosing friends than non-involved adolescents. Compared to bully-victims and non-involved adolescents, victims had reciprocal friends who were socially less well adjusted. No differences existed with respect to the characteristics of the desired friends. In general, victims' choosing friends scored less positive on the personal characteristics than bullyvictims' and non-involved adolescents' choosing friends. Within-group comparisons revealed that victims' reciprocal friends showed lower adjustment than victims' desired friends, but higher adjustment than their choosing friends.For bully-victims and non-involved adolescents, such differences between their three types of friends were largely absent. Our findings seem to suggest that victims' reciprocal friendships may not be totally default associations and that out of all possible friends, victims might tend to select those who score most positive on personal or social factors.

This is the first study to find evidence that different cognitive biases each predict unique variance in avoidance behavior. Furthermore, it is also the first study in which we found evidence for a relation between fear of spiders and size and distance indication. We showed that this bias is distinct from other cognitive biases.