A careful observer and listener can find young children exploring peer relations in all sorts of public spaces from playgrounds to grocery stores. Children barely able to toddle play peek-a-boo, run-and-chase, or simple pretend games like "I am pulling you in my wagon. We are going to the store." By 3 or 4 years of age, play becomes sophisticated with elaborate scripts and costumes. Children can say and act out. "We are going camping, and pretend that there is a bear, and you are the baby in your pajamas." "And I climb out of the tent and try to give the bear a cookie."
In this chapter we argue that, it is only within groups of peers that children develop both social interaction skills particular to peer interaction and construct social relationships particular to peers—friendships. The social interactions and relationships of the children within the group become the basis of that peer group's shared understandings and practices and a base for individual children's development. Through experiences within many different and at times overlapping in membership peer groups, children internalize representations of social relationships and of practices within peer groups that we assume influence their individual orientations to the social world as older children, adolescents, and adults.
Our approach to examining peer social interactions and relationships comes from a theoretical perspective best described as relationship development nested within socio-cultural contexts or cultural communities. In this framework a cultural community is defined as a grouping of people who participate in a shared set of practices and traditions (Rogoff, 2003). Peer groups are a kind of cultural community (Howes & Ritchie, 2002; Rogoff, 2003). As cultural communities peer groups construct shared understanding and meanings in forms that include shared scripts for pretend play, games, and conversations, knowledge of who hangs out with whom, who can and can not be trusted to gossip without hurting other people's feelings, and generally ways to behave within the group. These shared understandings and meanings are the practices of the peer community. Peer communities have shared histories as well as practices. Children within a peer group that lasts over time can remind each other of events that have meaning only within the context of the peer group, for example, "Remember when Sylvie was so mean to Nancy, and then they stopped being friends."
Much of the research on peer relations has been conducted as if all peer cultural communities were universally similar and thus developmental patterns were identical regardless of the characteristics of peer groups. We want to argue, in this chapter, that placing the development of peer interactions and friendships within cultural communities will help in resolving some of the contradictions that have persisted within the empirical literature on the topic. Take for example such a basic question of whether infants and toddlers engage in peer interaction. As we will discuss, research since the 1970s has documented that if 1-year-olds have regular playmates they engage in relatively sophisticated patterns of interactions (Lee, 1973; Rubenstein & Howes, 1976). However, if such young children are observed with unfamiliar peers they are fairly unskilled at interaction (Eckerman, Whatley, & Kutz, 1975). Many parents and teachers believe that infants and toddlers receive no benefit from peer interaction. Thus young children do not have regular playmates unless their families engage in practices that create baby peer groups such as enrolling the baby in child care, getting together regularly with friends who have same-age children, regularly attending a play group, or having regular extended family gatherings with age-mate children. Therefore in this example, the age of development of a particular social skill depends on belonging or not to a same-age regularly meeting peer group, and whether the child belongs to such a peer group depends on parent practices and values.
For another example, consider that children are in peer groups as diverse as those informally formed by living in a neighborhood, or by being part of a group of families that spends holidays or vacations together, or as formal as all the children in a child care arrangement, a school classroom, or a Scout Troop, or 4-H club, or religious club (Bryant, 1985; Ladd & Price, 1987; Medrich, Roizen, Rubin, & Buckley, 1982). The practices or ways to do things within each of these peer groups are particular to the peer group. For example, cross-gender or cross-age friendships may flourish in informal neighborhood peer groups and be actively discouraged within classroom based peer groups. Peer groups in school may facilitate cross-ethnic, cross-religious, or cross-class friendships forbidden within informal setting peer groups. Since children simultaneously hold membership in overlapping peer groups, they may have social practices that are competent in one peer group and incompetent in another. For example, ways to do things in one group (e.g., making in-group references and jokes), may be ways of not doing things in another group (e.g. the same references are not understood and are seen as denoting someone who does not belong). Such bridging of these different cultural communities may be tied to adaptive competence in children, a keen understanding and awareness of the significance of social context (Coll et al., 1996).
LEARNING TO INTERACT: INFANTS AND TODDLERS
So how do children become a group? How do they figure out how to engage with each other? How do they form friendships with some children and not others? When children only entered same-age peer groups as preschoolers the answer was that children first learned social skills from their parents or perhaps from older siblings and cousins and then applied them to peers at some later time. When children enter same-age peer groups as infants the answer is not as clear. From the 1920s and 1930s until the early 1970s, the study of peer relations in the United States was dominated by early descriptive research about the ways in which individuals' social skills developed. Pioneer researchers in the 1930s (Buhler, 1930; Maudry & Nekula, 1939), just as all researchers in all periods, based their conclusions on the cultural communities that were available to them to study. Because the earliest researchers studied naturally occurring peer groups of infants and toddlers (milk distribution centers in New York City's Central Park and orphanages), they concluded that infants and toddlers engage in games and other early forms of peer interaction (Buhler, 1930; Maudry & Nekula, 1939). Partens (1932, 1933), a researcher frequently cited in today's text books, studied children in nursery schools willing to enroll only those children who were 2 years 9 months of age and toilet trained. She argued that all children progressed through a set of social participation categories from solitary play, through parallel play, to true social participation in the form of cooperative play. The 2-, almost 3-year-olds were the solitary players, the 3-year-olds the parallel players, and the older children the truly social children.
In retrospect, we can see that the development captured in Partens' theory was closely linked to the context in which the observations typically took place: nursery schools. Nursery schools were part of a social movement that was based on a particular belief system. Within this belief system, it was assumed that the ideal child rearing environment for preschoolers consisted of a "stay-at-home mother" and, two or three mornings a week, a socialization experience with children of similar backgrounds. The ideal child rearing environment for younger children did not include peers. According to Reed (1950), "The nursery school is a place where young children learn as they play and as they share experiences with other children" (p. 3).
Living with a group of equals is a significantly different experience for a child from that of being a member of a family group. As students in the nursery school laboratory we can add to our understanding of what relationships mean to people as we observe the children in their living group. (Reed, 1950, p. 332)
The first and most influential nursery schools were physically located in colleges and universities and served as laboratories for the study of child development and for training professional teachers. Because research in peer relations and teacher training was confounded with a particular social organization, for 50 years, early childhood teachers in training and parents seeking advice were told that 2- and 3-year-olds were solitary or parallel players, while 5-year-olds were cooperative players.
Current research and theory on peer interaction supports an alternative perspective. The forms of play identified by Partens (1932) are not an invariant sequence based on age but, instead, are categories of play which tend to be used by children of all ages as the play they engage in with peers gradually shows more signs of structure, cognitive ability, and communicative ability (Bakeman & Brownlee, 1980). When research begins with a different set of assumptions and takes place in a different context, children as young as 10 to 12 months of age can be seen engaging at least one peer in cooperative play, which is more social than expected based on early theories (Howes, 1988b).
This alternative perspective was established because of two sociocultural changes: (a) research on peer relations moved out of laboratory schools, and (b) a social movement changed our beliefs about appropriate child rearing environments. In the 1970s, the women's movement and changes in the structuring of the United States economy led to an influx of mothers of very young children into the workforce. Part-day nursery schools could not accommodate the child care needs of these women. There was a dramatic increase in full-day child care centers that served infants and toddlers, instead of just preschoolers. These social changes changed the social context for constructing peer interactions, peer relationships, and peer social networks. In the social context of child care, peer interactions and relationships developed within long periods of "everyday life events" rather than within relatively brief "socialization experiences." Children in child care centers were with their peers every day and stayed in child care from breakfast time, through nap time, until the end of the day. The nursery school experience might be compared to a date, while the child care experience is more like living together.
Beyond structural changes in the social context of peer encounters, the women's movement changed the context in which the researchers were formulating their questions. They began to ask questions about the function of peers and peer relationships. Rather than simply being part of enrichment experiences, peers potentially could function to provide the child with experiences of social support, trust, and intimacy in the absence of the child's mother. Children who grew up together sharing the common resources of the child care center might engage in close rather than conflictual interaction. Cross-sex or cross-ethnic peers who became friends in a different environment than a traditional nuclear family might form different kinds of relationships. In the context of these questions, research on developmental changes in the complexity of peer interaction structure began.
Casual observations of infants and toddlers within full day child care settings suggested that they were not behaving according to Partens' (1932) observations: Their play was more interesting and complex than that of an onlooker or parallel player. Faced with this discrepancy, the first author conducted a qualitative study of two sets of infants from their first encounters in child care through the end of their first year together. Based on detailed narrative observations, both sets of children appeared to form friendships (Howes, 1981). This led to a program of research designed to understand how infants and toddlers constructed their play encounters within peer groups within child care centers.
To answer these new questions about preverbal children engaged with each other, we needed new tools for looking at peer interaction. In 1972, Blurton-Jones (1972) published an influential collection of observational studies of peer behaviors using ethological methods, with careful attention to the description and functional meaning of behaviors. Ethology had its roots in biology and the study of animal behavior. It provided a method for close, detailed observations of behavior. Subsequently, a generation of researchers including Eckerman, Davis, and Didow (1989), Hay (1985), Howes (1988b), Ross and Goldman (1976), and Vandell, Wilson, and Buchanan (1980) applied ethological methods to the task of identifying and describing changing structures underlying peer interaction.
One descriptive system that emerged from this research was the Peer Play Scale (Howes, 1980, 1988b; Howes & Matheson, 1992). Because each of the points of the scale represent increases in the complexity of play, we will use the scale to frame our discussion of how young children participate in constructing social structures within peer groups. There are several assumptions underlying the peer play scale. One assumption is that a necessary condition for children to be considered friends is that adult observers can infer from their behavior that each child understood the other to be a social actor and that social actions between partners could be coordinated and communicated (Howes, 1983, 1996). Therefore, as a starting place, research had to establish that children behaved as if they had these understandings. A second assumption is that later development in social play could occur only as the child increasingly understood the role of the other, incorporated symbolic play, and communicated shared meaning (Howes, Unger, & Seidner, 1989). These assumptions are the bases for the behaviors that are captured in the Peer Play Scale. Initially, children are expected to show signs of each of these three components but, eventually, they are expected to use them fluidly and communicate negotiations with each other about their play.
There are two key aspects of early play that presuppose such social understanding: mutual social awareness and coordination of action. Together, these two markers represent the necessary components for what Howes (1980) called complementary and reciprocal play. Specifically, each play partner's actions reverse the actions of the other. A child chases his or her partner, then is chased. One child peeks at his or her partner, the partner says boo, and then peeks back. Research in cognitive and communicative development (Howes, 1988b) suggests that the representational underpinnings of these understandings are present in children as young as the toddler developmental period. Naturalistic observations (Howes, 1980, 1988b) established that toddler-age children constructing their peer interactions within full-time child care centers were indeed engaging in complementary and reciprocal play.
CONSTRUCTING SHARED MEANINGS: TODDLERS AND PRESCHOOLERS
The next developmental step is to incorporate symbols into shared play. Children in peer group settings first begin to use symbols or to play pretend alone or with a competent adult player (Howes, 1985b; Howes & Matheson, 1992; Howes, with Unger, & Matheson, 1992), but symbolic play soon enters the realm of peer play. Pretend play with a partner requires both that the child manipulate symbolic transformations and communicate the resulting symbolic meaning to a partner. We see the simplest form of social pretend play, called cooperative social pretend play in Peer Play Scale terminology, among toddlers in full-time child care centers. Note that once more we are placing a developmental event within a particular socio-cultural context. Central to this level of structure of interaction is that children enact nonliteral role exchanges (Howes, 1985b; Howes et al., 1989). Play partners integrate their pretend actions by using a familiar pretend theme or script such as a tea party. Similarly to complementary and reciprocal play, cooperative social pretend play requires that children reverse the actions of the other but, in this form of play, the actions are nonliteral or symbolic. The actions of the children presuppose that each partner understands that each player may engage in the symbolic behaviors. The children are able to share understanding about the symbolic meaning of their play, but this is communicated through the implicit script of the play rather than explicit talk about the play.
For example, when a toddler offers a cup to a partner who is holding a pitcher, the child is engaged in a very simple form of social pretend play compared to the preschool-age child who discusses the play script, sets the table, brings festively dressed toy bears to the table, and gives the bears a tea party. Nonetheless, toddler-age children are beginning to understand the role of the partner in constructing social sequences.
Despite the new skills incorporated into cooperative social pretend play, it remains a pale imitation of the well-developed fantasy play of older children that we label complex social pretend play. Toddlers have only just begun to transform symbols so their transformations are not fluid and may be only partially developed. By preschool age, children's symbolic, linguistic, and communicative development permits meta-communication about social pretend play. Children can plan and negotiate the sequences of symbolic actions with fluidity, modify the script as it progresses, and step out of the pretend frame to correct the actions or script. These behaviors, such as those seen in a tea party for the toy bears, indicate the most structurally complex play captured in the Peer Play Scale. This play form is labeled complex social pretend play.
The Peer Play Scale is based on a set of measurement assumptions: the play forms develop in the predicted sequence and children develop particular play forms before or during particular age intervals predicted by theories of cognitive and communicative development (Howes, 1987; Howes, with Unger et al., 1992). Longitudinal (Howes & Matheson, 1992) and cross-sectional studies (Howes, 1980, 1985b) which focused on validating the Peer Play Scale supported these two assumptions. For example, in a longitudinal study of 48 children all in full-time child care centers and observed at 6-month intervals, we calculated each child's highest level of the Peer Play Scale during each observation period (Howes & Matheson, 1992). More than half of the children had engaged in complementary and reciprocal peer play by 13 to 15 months and nearly all by 19 to 23 months. More than half of the children had engaged in cooperative social pretend play by 30 to 35 months, and nearly half of the children had engaged in complex social pretend by 42 to 47 months. Seventy-four percent of the children followed the predicted sequence for emergence of play forms: complementary and reciprocal play, cooperative social pretend play, and complex social pretend play. Furthermore, children who showed earlier emergence of complementary and reciprocal play also showed earlier emergence of cooperative and complex social pretend play.
As a measure of structural complexity, the Peer Play Scale makes no distinctions among positive, aggressive, or agonistic (instrumental aggression such as toy taking) social bids. A structurally complex interaction could reflect prosocial behavior, a conflict, or any number of social styles. As we discuss below, research examining associations between the structure and content of peer interaction appeared later in the 1980s and 1990s.
Perhaps because it captures structural complexity rather than the content of peer play, the Peer Play Scale has been successfully used in cultures other than the United States. Farver has used the Peer Play Scale to describe and examine peer interaction in such diverse cultures and ethnic groups as Mexican children (Farver, 1992), Latino and African American Head Start children in Los Angeles (Farver & Frosch, 1996; Farver, 1996), Indonesian children (Farver & Howes, 1988; Farver & Wimbarti, 1995), and Korean American children (Farver, 2000; Farver & Shin, 1997; Farver, Kim, & Lee, 1995). Across studies, children's play was represented at each structural play level and play forms emerged at similar ages. When differences emerged, they were in the frequency of play forms rather than in different play forms or a different sequence of play. Farver (1996) suggested that the sociocultural context influences the style or frequencies of peer play. In particular the types of themes in pretend play appear rooted within children's particular cultural communities (Goncu, 2002). Whether children play at wrapping the "babies" in shawls and placing them on their backs or enacting the latest TV superhero's antics is dependent on the practices of daily life within their cultural community.
Gender also appears to influence the style rather than the structure of peer play. There are well-established differences in the content of the play of boys and girls (Fabes, Hanish, & Martin, 2003; Maccoby, 1988, 1990; Martin & Fabes, 2001). However, consistent with the lack of cultural and ethnic differences in the structure of peer play, there appear to be few differences in the complexity of the structure of children's peer play (Howes, 1980, 1988b; Howes & Matheson, 1992; Howes & Wu, 1990). Girls and boys of the same age engage in structurally similar play when the content differs. For example, both a game of mother, sister, and baby among girls and a game of the day the tigers ate the village among boys are very likely to be rated as complex social pretend play.
However, consistent with a Vygotskian perspective, the complexity of the structure of play is influenced by the skill level of the play partner. If the sociocultural context of the peer group includes mixed age children such as in family child care homes, toddlers play more skillfully with their peers when their partner is somewhat older and presumably more skilled at play (Howes & Farver 1987a; Rothstein-Fisch & Howes, 1988). In contrast, when the mother is present as in parent cooperative preschools the presence of the child's mother appears to reduce the complexity of peer interaction (Smith & Howes, 1994). Similarly when adults engage with children in preschool settings, peer interaction is inhibited (Harper & McCluskey, 2003; Kontos & Keyes, 1999). Although adults are more skilled play partner than children, peer play in the presence of an adult may be less skillful because three partner interaction is more difficult than dyadic interaction or because adult-child play content is different from child-child play content (Howes, with Unger et al., 1992).
FORMING SOCIAL STRUCTURES WITHIN PEER GROUPS: I'LL BE YOUR FRIEND IF . . .
Young children's peer interaction is organized around networks of play partners (Howes, 1983, 1988b). Within any peer group, there are children who prefer to play with each other, and children who prefer not to play with each other. Some children have no problem finding and keeping play partners, while others have difficulty entering peer groups and sustaining play. Children as young as 3 years old who have experienced full time child care in stable peer groups can describe these patterns of peer acceptance within their classroom (Howes, 1988b). When shown pictures of all children in the classroom, they can identify by name each child in the classroom and reliably rate how much they would want that child as a friend. This picture sociometric procedure provides a description of social status.
In the late 1970s, a series of longitudinal studies were published suggesting that children's peer acceptance in middle childhood was associated with positive mental health outcomes in adulthood (Parker & Asher, 1987). These studies precipitated a period of intense research into children's sociometric status (e.g., Coie & Dodge, 1983; Ladd, 1983) and continues as an important area of research (Crick & Dodge, 1999; Stormshak et al., 1999). The term sociometric refers to ways of measuring peer acceptance and friendships. Sociometric nominations involve asking each child in a defined group to identify the children that he or she likes and the children that he or she dislikes as one means to determine peer acceptance. Children with reciprocated nominations of liking are usually considered friends. Children also can be assigned a sociometric rating based on the average of ratings provided by each of the children in the group. Social status groups (popular, rejected, neglected, and average) are formed using either nominations or ratings. Popular children have high liking and low disliking profiles. Rejected children have low liking and high disliking profiles. Neglected children have low liking and low disliking profiles. Average children fit none of the other profiles. Peer acceptance refers to high ratings or many positive and few negative nominations.
Studies of sociometric status found that children who were classified as popular or more socially accepted were more friendly and prosocial, and less likely than children who were not accepted to engage in aggressive behaviors with peers (Asher & Coie, 1990; Coie 1990). Although the original research was conducted in elementary school classrooms, similar findings have emerged in preschool settings (Walden, 1999). Furthermore, sociometric status tended to be stable over time (Asher & Coie, 1990). These findings based on sociometric measures raised two questions about the Peer Play Scale. First, if the Peer Play Scale represents socially competent behavior with peers, would children who engage in structurally complex play also have higher peer acceptance and engage in prosocial, nonaggressive interactions with peers? Second, is competent peer play stable over developmental periods? That is, does demonstrating marker behaviors of structurally complex peer interaction during an earlier developmental period predict engaging in marker behaviors of structurally complex peer interaction during a later developmental period?
Two longitudinal studies addressed the first question. In the first, 48 children were observed with their peers in their child care centers at 6-month intervals beginning at age 13 to 24 months (Howes & Matheson, 1992; Howes, Phillipsen, & Hamilton, 1993). Within developmental periods, independent observers rated children who engaged in more complex play as more prosocial and sociable, and less aggressive. Children tended to be stable over time in positive, gregarious, and aggressive behaviors. Children who engaged in more complex play at earlier developmental periods were observed and rated as more prosocial and sociable, and less aggressive and withdrawn, during subsequent periods.
Another longitudinal study followed 85 children from 12 months through adolescence. Children were observed with peers soon after entering group child care (Howes, Rodning, Galluzzo, & Myers, 1988) and again at 4 years of age (Howes, Hamilton, & Matheson, 1994). In addition, at these time period both preschool teachers and independent observers rated aggression and social withdrawal when the children were toddlers and, again, when they were preschoolers. When the children were 9 years of age, aggression, social withdrawal, and prosocial behaviors with peers were rated by elementary school teachers. Children who engaged in more complex play with peers as toddlers were more prosocial, engaged in more complex play, and were less withdrawn as preschoolers, and were less aggressive and withdrawn at age 9. Children who were either more aggressive or more withdrawn as preschoolers were more aggressive as 9-year-olds (Howes & Phillipsen, 1998). Children who were rated by teachers as having close friendships in preschool rated themselves as having a more positive friendship in middle childhood (Howes , Hamilton, & Phillipsen, 1998). Children who rated themselves as having a more positive friendship in middle childhood and who had a history of positive friendships as preschoolers rated themselves as having more positive friendships in early (Howes & Tonyan, 2000) and mid-adolescence (Howes & Aikins, 2002). The last two of these findings are important in that the peer relations literature has relatively few studies that follow children longitudinally from very early childhood into adolescence. They suggest that social competency in peer relations may be quite stable. Some of this stability may occur because once children develop early social skills with peers they continue to use them and become the type of older children who are comfortable and easy interacting with peers. On the other hand, the children who are shy or aggressive in early peer interactions have fewer and fewer opportunities to learn how to be comfortable and easy with peers as they become more isolated from positive peer interactions.
In a second more short term longitudinal study of stability in social skills with peers, 329 children between the ages of12 and 53 months participated for 1, 2, or 3 years in a study of the development of social interaction with peers (Howes, 1988b). Naturalistic observations of peer play were conducted on a yearly basis. Complementary and reciprocal play in the early toddler period (13-23 months) predicted cooperative social pretend play in the late toddler period (24-35 months). Cooperative social pretend play predicted sociometric ratings in the preschool period. An extension of this longitudinal study, limited to 45 children who remained in the same school from prekindergarten through third grade, found that sociometric ratings in prekindergarten and kindergarten were associated with sociometric ratings in third grade (Howes, 1990). That is, children who more often engaged in complex social pretend play in their first year in the peer group were rated as more popular with the peers in this group 3 years later.
This last finding brings us full circle back to the idea first stated in the introduction that peer groups become a form of cultural community with its own particular practices and traditions. When the peers in their peer group accept children and the peer group socializes prosocial behavior this can be a wonderful context for individual development. The complement of this statement may also be true if the social status structure of a peer groups comes to be part of its practices and traditions, children who are not well accepted have a difficult time.
The natural variation in familiarity and stability of peer partners provides an opportunity to examine socio-cultural variations in children's experiences in peer groups. An example of this natural variation is found in study discussed above (Howes, 1988b). In the first year of the study, children were recruited from child care classrooms with peer groups that remained intact for 1 year. In each subsequent year of the study, investigators both returned to the original peer groups and followed children who moved to different child care centers. Thus, in the second year of the study, the sample included four distinct groups of children: (a) the children who had been in the toddler room of a particular child care center the previous year and who had moved "up" to the preschool room in the same center; (b) the children who had been newly enrolled in the original child care centers to fill in preschool classroom peer groups (generally, group size increases with age); (c) the children who were in the sample in year 1, but had changed child care centers; and (d) the children who were in the same child care center classrooms as the children from year 1 who had moved to different child care centers. The results of this study suggested that toddler and preschool peer interaction is more skillful when children enroll in group settings at earlier ages, and when they remain with the same peer group for longer periods of time (Howes, 1988b).
That is, it appears to be advantageous for children to enter peer groups as younger children and to maintain a familiar and stable peer network. Some children who moved between child care centers actually moved with a peer group, as opposed to moving to a child care center with an entirely new peer group. Those children who were able to move with a peer group, and who therefore did not have to reestablish relationships, were more competent in peer interaction (Howes, 1988b).
Of course, some peer groups might promote more social competence than others. In general children are more socially competent within a peer group if the peer group is within a child care program, center or family child care that is more supportive of positive relationships (Howes & Matheson, 1992; Howes, Phillips, & Whitebook, 1992; Howes & Stewart, 1987). These findings point to the importance of adults' structuring a child care environment which supports the construction of positive and skillful peer interaction (Howes, 2000). In a recent study the social-emotional climate of the child care environment as well as children's individual social competence and teacher-child relationships predicted social competence with peers five years after children were in child care.
A recent report of the NICHD early child care research study highlights these concerns (ECCRN, 2003). The NICHD study reported that at ages 2 and 3 children who had experienced positive and responsive caregivers and the opportunity to engage with other children in child care were observed to be more positive and skillful in their peer play in child care but not in play with a friend in a laboratory (ECCRN, 2001). When the children were 4-1/2 years old and in kindergarten, teachers, caregivers, and mothers rated children with more time in child care as having more problematic relationships and behaviors with others (ECCRN, 2003). While overall measures of child care quality did not mediate this relationship between time in child care and problematic behavior this study did not include specific measures of adult structuring of peer relationships. As one of the commentaries on this study suggests perhaps it was not the amount of child care but amount of time in a kind of child care that fosters individualistic rather than cooperative interactions with others (Maccoby & Lewis, 2003).
How do adults structure a child care environment that promotes positive peer relations? Not it turns out by coaching children in social skills. Inspired by the more traditional literature on the influence of adults on children's peer relations, Howes and colleagues conducted another 3-year longitudinal study of the development of peer relations that included a focus on a large number of adult behaviors that could be expected to enhance peer interaction (Howes & Galluzzo, 1994). These behaviors included teaching toddlers turn-taking games and structuring their peer interactions to support these games, intervening to redirect agonistic encounters, and so forth. Therefore, as well as coding the structure and content of children's interactions with peers, we also noted any simultaneously occurring adult behaviors. The findings of this study suggested that adults working with groups of children rarely engaged in such behaviors and whether or not they did so had no association with the development of toddlers' and preschoolers' social competence with their peers.
A more promising line of research examined relations between children's attachment relationship quality with the primary caregivers in the child care setting and the development of children's social competence with peers (Howes, 1999). Adult caregiving behaviors directly influence children's attach ment relationships that, in turn, influence peer relations. More specifically, more securely attached children have caregivers who are rated as more sensitive and responsive to the children in their care (Howes, 1999). In turn, children who are more securely attached to their child care providers are more socially competent with their peers (Howes, 1997a, b; Howes et al., 1994; Howes, Matheson, & Hamilton, 1994). When children's attachment relationships with their child care provider are contrasted with their attachment relationships with their mothers, the child care attachment relationship is more powerful in predicting social competence with peers (Howes, Matheson, & Hamilton, 1994b). We suspect that this is because these adults are physically present with the children's earliest experiences within peer groups. The children can use the child care provider as a secure base as they explore interactions and relationships within the peer group.
Unfortunately, child care providers are not a stable presence in children's lives. In contrast, mothers usually remain constant. Changing child care providers appears to have implications for children's social competence with peers. In a 3-year longitudinal study, children who had the most changes in primary child care providers were most aggressive with peers (Howes & Hamilton, 1993). In this study, some of the caregiver changes had a positive effect in the short term: The child was able to construct a more secure relationship with the new provider than with the old and, simultaneously, became more skilled with peers. However, it appears that the cumulative effect of instability in child care caregivers is detrimental to the development of positive social competence with peers.
Broadening the study of peer relations to include a more careful examination of child-caregiver relationships has led to a particularly fruitful series of research studies. The evidence that children construct interactions and relationships with their peers from infancy onward highlights the importance of these early peer relationships. Likewise, the relationships that children form with the adults who supervise their encounters with peers influence social competence. It is noteworthy that both of these types of relationships, with peers as friends and with child care providers, are relationships outside of the family. While these results should in no way downplay the importance of family influences on child development, carefully examining alternative relationships underscores the value of social networks both inside and outside of the family.
Was this article helpful?
Parenting is a challenging task. As a single parent, how can you juggle work, parenting, and possibly college studies single handedly and still manage to be an ideal parent for your child? Read the 65-page eBook Single Parenting Becoming The Best Parent For Your Child to find out how. Loaded with tips, it can inspire, empower, and instruct you to successfully face the challenges of parenthood.