A positive climate for social and moral growth is one that fosters peer interaction (Nucci, 2001). In middle childhood, 30% of a child’s social interactions involve peers, compared to 10% in early childhood (Rubin, Bukowski, & Parker, 1998). Children’s behavior in the peer group has proven to be a stable indicator of their social competence (Hartup, 1996; Zeller, Vannatta, Schaffer, & Noll, 2003). School-agers not only construct understandings of others but must also interact competently with their peers and sustain friendships over time. In addition, children’s concerns about acceptance in the peer group often rise during middle childhood. Many peer relations researchers, not surprisingly, have focused on the study of friendships among school-age children (for a review, see Rubin, Bukowski, & Parker, 1998).

Friendships are important for social development. Friendship processes are linked to social developmental outcomes, also called social provisions. In a classic text called The Interpersonal Theory of Psychiatry, Harry Stack Sullivan (1953/1981) argued that friends fulfill social needs, called communal needs, such as companionship, acceptance, and intimacy (Buhrmester, 1996). In many ways, this formulation is similar to Maslow’s need for belonging. Furthermore, communal needs can be distinguished from other human needs:

  • Communal Needs.Interpersonal needs for affection, nurturance, enjoyment, support, companionship, intimacy, and sexual fulfillment
  • Survival Needs.Physical needs for safety, food, shelter, and health
  • Agentic Needs.Individual needs for competency, achievement, status, power, approval, autonomy, identity and self-esteem (see Buhrmester, 1996)

The social concerns of school-age children often focus on the communal needs of acceptance by peers and avoidance of rejection.
Alesha’s need for continued acceptance by her best friends from elementary school—Jennifer and Kelly—may prove stronger than her need for Mandy’s companionship. When confronted with Mandy’s offer to get high, Alesha seems genuinely fearful of rejection by her peer group. Will they still like her if she started hanging out with Mandy? Why should they continue to be friends if she’s becoming one of the “druggies”?


Middle childhood brings about marked changes in the understanding of friendship. In early childhood, friendship is usually associated with sharing a current activity, whereas in middle childhood children begin to recognize that friendships can last over time (Parker & Seal, 1996). Sullivan (1953/1981) referred to friends in the preadolescent years between ages 8 and 10 as chums. School-age children usually indicate an increasing appreciation of each others’ feelings and intentions, brought about by advances in their social perspective-taking ability (Rubin, Bukowski, & Parker, 1998).

Contemporary researchers have typically investigated school-age children’s friendship concepts by asking them what they expect from a best friend. Friendship stages in middle childhood have been found to parallel the stages of social understanding. Children who are social-informational perspective takers and unilateral negotiators will likely think of friendships as one-way, that is, they may think about what a friend could do for them, or vice versa, but not what they reciprocally could do for each other (Selman, 1980). For example, in the hypothetical friendship story told previously, it is likely that a child at this level would suggest that Kathy take the “better” offer. Some researchers have termed this period a “reward-cost” stage, in which children around 7 or 8 believe that friends are rewarding to be with compared to nonfriends (Bigelow, 1977; see also Robin, Bukowski, & Parker, 1998). Many school-age children have best friends and are satisfied with those friendships (Parker & Asher, 1993).

By ages 10 or 11, however, most school-age children demonstrate “normative” friendships, in which they recognize that friends are supposed to be loyal to each other (Bigelow, 1977). Recall that these children are likely to be self-reflective perspective takers and cooperative or compromising negotiators. They conceptualize friendships as reciprocal, although it is still highly unlikely that these relationships will survive difficult arguments or negative events (Selman, 1980). For this reason, school-agers have often been called “fairweather” friends because their relationships may not weather stormy periods (Rubin, Bukowski, & Parker, 1998). In addition, when boys in one study terminated a friendship because of conflict, the typical length of time for working it through and renewing the friendship was one day, but for girls it took about two weeks. One explanation for this sex difference is that triads (like Alesha, Jennifer, and Kelly) are more common in the friendships of school-age girls than boys, causing one member of the group to feel left out (Azmitia, Kamprath, & Linnet, 1998).

By the end of middle childhood, most children are capable of mutual role-taking and collaborative negotiation, as we have discussed. Friends at this stage expect each other to provide companionship, help, protection, and support (Azmitia, Kamprath, & Linnet, 1998). For them, friendships are becoming intimate, characterized by an enduring sense of trust in each other. For example, between third and sixth grades, the expectation that friends would keep secrets rose from 25% to 72% among girls, although boys had no such similar expectation until the end of sixth grade (Azmitia, Kamprath, & Linnet, 1998). These maturing notions of friendship translate into social interaction with peers during middle childhood. In addition, peer acceptance influences school-age friendship dynamics. Overall, the ability to form close, intimate friendships becomes increasingly important during early adolescence (Buhrmester, 1990).

Comparing Friendship and Peer Relations

In middle childhood, maturing peer relations are dependent on children's developing understanding of friendship.

Stage Peer Relations Friendship
Stage 0 Physcial connections Momentary physical interaction
Stage 1 Unilateral relations One-way assistance
Stage 2 Bilateral partnerships Fairweather cooperation
Stage 3 Homogenous community Intimate and mutual sharing
Stage 4 Pluralistic organization Autonomous interdependence

Source: Based on "Four Domains, Five Stages: A Summary of Portrait of Interpersonal Understanding, " by R.S. Selman, In The Growth of Interpersonal Understanding Developmental and Clinical Implications ((pp136 - 147), 1980 New York: Academic.


During middle childhood, some children seem to have many friends and others only a few. However, a central concern of many—if not all—school-agers is popularity. Popularity (also called social status) has been operationally defined by a majority of peer interaction researchers as the number of individuals who name an individual target child as “liked” or “disliked” or as a “friend” or “best friend” (Newcomb, Bukowski, & Pattee, 1993). Children with the most “liked” nominations are considered popular, whereas those with the most “disliked” nominations are considered rejected. Children with few or no nominations are often termed neglected. Children are considered controversial if they are both nominated frequently by some and actively disliked by others.

Boys’ social status tends to be based on social dominance, athletic ability, coolness, and toughness, whereas girls’ status depends more on family background, socioeconomic status, and physical appearance (McHale, Dariotis, & Kauh, 2003). Generally, school-age children with diverse social status classifications differ in behavior and characteristics (Newcomb, Bukowski, & Patee, 1993; Wentzel, 2003):

  • Popular Children.Tend to exhibit higher levels of positive social behavior and cognitive ability and lower levels of aggression and withdrawal than average children
  • Rejected Children.Tend to exhibit just the opposite pattern—more aggressive and withdrawn and less sociable and cognitively skilled than average children
  • Neglected Children.Tend to exhibit less social interaction and disruptive behavior but more withdrawal than average children
  • Controversial Children.Tend to be less compliant and more aggressive than average children

In addition to totaling the number of friendship nominations a particular child may receive, researchers may examine the nominations to see if they are reciprocal. Peer relations researchers often distinguish peer acceptance (i.e., the number of “liked” ratings children receive) from friendship (i.e., the number of reciprocated “friend” choices) (Asher, Parker, & Walker, 1996). The concept of peer acceptance differs from friendship because it refers to children’s relationships within a group rather than the quality of children’s dyadic relationships with individual peers. For example, peer relations researchers sometimes observe classroom or playground social interactions and chart peer relationships based on frequencies of actual contacts between acquaintances (i.e., peers children “know” but with whom they have no close reciprocated ties) (Ladd & Kochenderfer, 1996). Either method (i.e., ratings or observations) is called a sociometric classification of group acceptance and can be summarized in a visual diagram called a sociogram.

Peer acceptance may influence friendships by determining the amount of choice that children have in friends (Azmitia, Kamprath, & Linnet, 1998). Recall that in the transition to middle school, peer group sizes may increase as children attend several different classes in a typical day. For both boys and girls, the anticipated transition produced feelings of anxiety (Pratt & George, 2005). When patterns of attraction to peers were examined across the transition to middle school, researchers found that attractions to aggressive peers and to children who stand out in the peer group increase with age, especially attraction to aggressive boys among girls (Bukowski, Sippola, & Newcomb, 2000).

Girls experience less stability in the number of reciprocated friendships across the school transition than do boys, although they have similar numbers of friends overall. Girls also are more likely than boys to form new reciprocated friendships with previously unfamiliar peers, especially if they had attended relatively small elementary schools (Hardy, Bukowski, & Sippola, 2002). In other words, girls may experience more changes in friends than boys after the middle school transition.

Recall that Alesha and Mandy had just met in middle-school choir and quickly formed a new friendship, as is typical of many middle-school girls.

In a recent study, sixth-grade sociometric status (“liked-most” and “liked-least” scores) predicted eighth-grade school adjustment (Wentzel, 2003). Although the number and stability of friendships do not always contribute to adjustment in the transition to middle school, having supportive friends predicts increasing popularity (Berndt, 1989). A classroom activity called the “Class Play” is described in . This technique has proven useful for determining school-agers’ reputations in their peer groups.

Peer groups, like other organizations, have distinctive structures. In a peer group’s organization, some members are central while others are marginal, for example. In a study of fourth, fifth, and sixth graders, peers’ positions in the organization of same-sex peer groups were significantly related to their social status. Peer group members who were marginal in the peer group organization were more likely to be rejected and friendless, whereas those in a central position were likely to be popular and have at least one friend (Lease & Axelrod, 2001). Furthermore, rejected children on the margins of the group were perceived by their peers as “different” in their social behavior (i.e., more odd, inattentive, excluded, shy, or anxious) compared to rejected peers who were less marginal in the peer group organization (Lease, McFall, & Viken, 2003).

Peer networks

Researchers studying the formation of peer networks define a clique as a small group of close friends (for an ethnographic study of peer networks, see Adler & Adler, 1998). Generally, girls’ friendship networks are smaller (i.e., more intensive) than boys’ networks (i.e., more extensive). Longitudinal research, however, has shown that boys’ friendship networks are more likely to become interconnected over time compared to girls’ (Ladd, 1999). Interestingly, the size and density of adolescents’ networks has been related to parents’ friendship networks—especially mothers’—perhaps due to the influence of social learning (Parke & Buriel, 1998).

Unlike interaction-based cliques that are composed of friendship networks, crowds are reputation-based groups of children who are not necessarily friends but who share similar values, attitudes, and behaviors, such as “jocks” or “brains” (Prinstein & La Greca, 2002; see also Brown, Eicher, & Petrie, 1986). Peer crowds serve an important relational function when children enter a new school and interact with increasing numbers of peers. The crowd system may act as a social guide to help school-age children maintain peer relationships, meet possible friends, and—eventually—choose romantic partners. In adolescence, students who are more central in their peer networks feel more positively about school (Crosnoe, 2000). On the other hand, adolescents who affiliate with deviant crowds (e.g., “burnouts”) report higher levels of illegal behavior, alcohol and marijuana use, aggression, and risky sexual behavior (Prinstein & La Greca, 2002).

Will Alesha be able to build connections between her “clique” of friends from elementary school and a new “crowd” in middle school?

Peer and friendship networks are likely to form because of three factors, based on the principles of social learning theory discussed earlier (Hartup, 1996):

  1. Sociodemographics.  Children are likely to come into proximity because of age, socioeconomic status, ethnicity, etc.
  2. Social Selection.  Children construct relationships with others who are similar to themselves or to whom they are attracted.
  3. Mutual Socialization.  Children become increasingly similar to their friends as they interact.

Typically, the peer groups of school-age children are segregated by sex, although socially unskilled children who are rejected by their peers may be more likely to seek opposite-sex friends. Girls usually place a priority on interpersonal connections (i.e., communal needs), whereas boys place a higher priority on status concerns (i.e., agentic needs) (see Maccoby, 1990, 1998). For example, in a study of friendship quality, girls reported more frequent intimate and supportive interactions with female friends than boys did with male friends (Buhrmester, 1996). In other studies, boys were more likely than girls to express anger towards well-liked peers, perhaps due to concerns about competition. Girls were more likely to judge a friend’s misdeeds in terms of how these behaviors would affect their relationship (see Ladd, 1999).

Such sex differences are likely due to differential gender socialization experiences in family, school, and community contexts (discussed later in this chapter). For example, in a study of school-age children’s social networks on a school playground, ethnographic observers found that sex-typed play (i.e., all-boy or all-girl) interacted with stereotypical behaviors related to other social categories, such as age, race, and social class. Girls’ and boys’ hairstyles and clothing, for instance, amplified differences between male and female peer groups (Thorne, 1993, 1997a). In addition, the greater social prestige of early pubertal maturation for boys versus girls reproduced the dynamics of male dominance on the playground (Thorne, 1997b).

Another typical cross-sex interaction, heterosexual dating (or “going together”), is often an extension of peer networks in middle childhood rather than the intimate dating relationship characteristic of adolescents. For example, middle-school dating usually consists of groups going out together, often a combination of male and female cliques (Pellegrini, 2001).