Most children and adolescents have a “best friend” and several “good friends”. Friendship networks remain relatively small during the preschool years, with the number of mutual friends averaging 0.88 for boys and 0.63 for girls. The essentials of friendship are reciprocity and commitment between individuals who see themselves more or less as equals. Children’s friendships can also be identified by asking children to name other children who are “especially liked”. Close relationships have beginnings, middles, and ends. Relationship dynamics vary accordingly. Friendships terminate when common ground dissolves and children cease to be behaviourally interdependent. Mutual attraction ensures continuing interaction between the children involved. Friendships are cognitive and emotional resources furnishing the child with maximum capacity for problem-solving as well as security for striking out into new territory, meeting new people, and tackling new problems. Observations in child care centres show considerable differentiation in the social interaction occurring among infants and toddlers.

This chapter summarizes the emergence and prominent features of a science of psychological development. Pioneering researchers established laboratories in Europe in the mid-nineteenth century to examine the significance of successive changes in the organism with the passage of time. American psychologists, many of whom had studied in the European laboratories, subsequently inaugurated similar efforts in the United States. Scientific theories and methods in the fledgling field were fostered by developments in experimental psychology, but also in physiology, embryology, ethology, and sociology. Moreover, organized efforts to provide information about development to parents, educators, and public policy specialists further propagated support for developmental science. The evolution of the field in its first century has provided a substantial platform for future developmental research.

The early developmental antecedents of individual differences in children's social functioning with peers in third grade were examined using longitudinal data from the large-scale NICHD Study of Early Child Care. In a sample of 1364 children, with family and child factors controlled, the frequency of positive and negative peer interactions in child care between 24 and 54 months and the number of hours spent in child-care peer groups of different sizes (alone; dyad; small; medium; large), predicted third graders' peer competence at three levels of analysis: individual social skills, dyadic friendships, and peer group acceptance. Children who had more positive experiences with peers in child care had better social and communicative skills with peers in third grade, were more sociable and cooperative and less aggressive, had more close friends, and were more accepted and popular. Children with more frequent negative experiences with peers in child care were more aggressive in third grade, had lower social and communicative skills, and reported having fewer friends. When children spent more time in small-sized peer groups in child care (4 or fewer children at 24 months of age, up to 7 or fewer at 54 months), they were more sociable and cooperative in third grade, but their teachers rated them as more aggressive suggesting that such children may be more socially outgoing and active both positively and negatively. Like those who spent more time in small peer groups, children who spent more hours in medium-sized groups received higher ratings for peer aggression by their third grade teachers. Children who spent more time with one other child in child care or in small peer groups had fewer classroom friends in third grade as reported by the teacher but not according to maternal report or self report. There were no significant associations between the amount of time children spent in large child-care based peer groups and third grade peer social competence.

Peer interaction is examined as a source of behavior change in children and adolescents. The dynamics of peer influence are discussed in terms of six issues: characteristics of the influence source; characteristics of the child or adolescent being influenced; the relationship existing between influence source and the individual being influenced; developmental change; processes underlying peer influence; and constraints deriving from the behavior domain being examined. Research shows that contact with peers has complex effects on individual behavior, and numerous moderators and mediators must be considered in order to account for them. The discussion is linked to the papers included in this special issue.

Three changes have occurred in research on the development of aggression. First, both theoretical and empirical attention have shifted from the aggressive act, on one hand, to the development of aggressive individuals, on the other, that is, from aggressive events and the situations that instigate them to aggressive children assessed within ecological systems. Investigators no longer examine the effects of exposure to aggressive models (Bandura & Walters, 1963), for example, but rather the developmental consequences of reciprocities in social interaction involving individuals, bilateral and transactional processes in family and peer interaction, gene–environment permutations, and contextual components (including exposure to aggression and violence in the community). A second change is the emerging synthesis between research on aggression and research on antisocial behavior (Tremblay, 2000). Everyone knows that aggression and antisocial behavior are overlapping but not identical constructs. Nevertheless, because aggressive behavior is relatively stable as an individual characteristic, and because stable aggression predicts antisocial behavior during adolescence and adulthood for males at least (Moffitt, Caspi, Rutter, & Silva, 2001), it no longer seems wise to separate these terms in thinking about social development and adaptation. Third, a trend toward more developmentally oriented studies of aggression has emerged. What was known about aggression in childhood and adolescence in the late 1960s was quite nondevelopmental: The relevant work dealt with contingencies in observation or social interaction that increased aggression in the child, usually in constrained situations and with little attention to age differences. Normative studies encompassed only a few years (in early

Relationships based on aversion and antipathy may turn out to have major developmental implications, but more refined analysis is needed to support this assertion.

Friendships are transformed from equality-based exchange relationships during early childhood into need-based communal relationships during adolescence.

Mutual antipathies (when two children or adolescents dislike one another) were studied among 2,348 school-age children and 2,768 adolescents to determine incidence, gender and age differences, and implications for social adjustment. The children were more frequently involved than were the adolescents in same-sex antipathies but not mixed-sex ones. Boys were involved more frequently than were girls in same-sex antipathies, but involvement in mixed-sex antipathies was comparable for the two genders. With peer rejection scores used as a covariate, same-sex antipathies were associated with antisocial behavior and social withdrawal among children and adolescents of both genders and, in addition, to emotionality and lack of friendship support among adolescents. Mixed-sex antipathies were related to social adjustment depending on gender: these antipathies were related to antisocial and bullying behavior in boys but not girls; and to nonaggressiveness, victimization, lesser cooperation, shyness, and depression in girls but not boys. Mutual antipathies thus appear to be concomitants of adaptational risk in both childhood and adolescence.

Three conceptually distinct dimensions of classroom social position (number of mutual friendships, social network centrality, and sociometric status) were examined in relation to each other and to peer‐nominated behavioral reputation among 205 7‐ and 8‐year old children. There were moderate correlations in children’s standing across the three dimensions, but categorical analyses underscored the limits to these associations (e.g., 39% of Rejected children had at least one mutual friendship; 31% of Popular children did not). Each dimension was associated with a distinct profile of peer‐nominated social behavior and, in multiple regression analyses, accounted for unique variance in peer‐nominated behaviors. Number of friendships was uniquely associated with prosocial skills; network centrality was uniquely associated with both prosocial and antisocial behavioral styles; and being disliked was uniquely associated with the full range of social behaviors. Results provide empirical validation for the conceptual distinctions among number of reciprocated friendships, social network centrality and being liked or disliked.