Peerrelationsin childhood - Cardiff Universitypsych.cf.ac.uk/home2/hay/hay2004.pdf · Peerrelationsin childhood Dale F. Hay, ... than entering into ultimately unsatisfying debates

Peer relations in childhood

Dale F. Hay, Alexandra Payne, and Andrea ChadwickSchool of Psychology, Cardiff University, Wales

We present a developmental model that describes normal peer relations and highlights processes thatunderlie the emergence of problems with peers in childhood. We propose that children’s relationshipswith peers begin in the first years of life, with stable individual differences and preferences for particularpeers emerging by three years of age. Social skills that facilitate peer relationships consolidate in thepreschool years, during which time peer groups become structured with respect to friendship groups,gender, and dominance relations; some children begin to be rejected by their peers. In later childhoodsome children develop entrenched problems with peer relationships, in terms of loneliness, bullying andvictimisation. Underlying cognitive and emotional processes that facilitate successful peer relationshipsat all ages are identified, and the extent to which peer relations play a causal role in the genesis ofdisorder is evaluated. A review of the evidence suggests that, rather than a simple pathway fromproblematic peer relations to disorder, there is a reciprocal relationship between children’s problemswith peers and their psychological problems from infancy to adolescence.

‘ The chicken came first, ‘cause who else laid the egg?’’‘No, ‘cause a long time ago, you know what, an egg, itsprouted out of the ground.’‘Uh-uh, the chicken came first, the chicken came first!’(Two nursery school classmates, recorded in Caplan,1986)

The aim of this review is to propose a develop-mental model that both describes the normal courseof children’s relations with peers and draws atten-tion to cognitive and emotional problems that makeit difficult for some children to cope with peers. Thereis no doubt that peers play important, and some-times critical, roles in children’s lives, as attested toby the very many reviews on this topic (e.g., Deater-Deckard, 2001; Rubin, Bukowski, & Parker, 1998).The study of peer relations is particularly germanefor those who are interested in childhood disorders.There is bi-directional influence between peer re-lationships and children’s disorders, in a mannerthat evokes the well-known conundrum quotedabove. Children’s problems with peers may contrib-ute to the genesis of disorder, in particular anxiety,depression and conduct disorder; conversely, chil-dren with disorders may find themselves at oddswith their peers from the very first years of life.

We shall argue that the relationship between peerrelations and children’s risk for disorder is bestunderstood by taking the long view, and examiningthe early development of the ability to relate to peers,beginning in the first months of life. In this paper,the developmental model that we propose charts thebeginnings of peer interaction in infancy, describesthe consolidation of important skills in the childhoodyears and finally draws attention to some entrenched

problems that arise when children spend more oftheir time in organised peer groups. With respect toeach phase of the proposed developmental pathwayto problematic peer relationships (i.e., beginnings of

interest in peers, consolidation of skills and emer-gence of problematic relationships), we note evidencefor: dyadic relationships and group relations; stableand coherent individual differences; and underlying

cognitive and emotional processes. Finally, we con-sider whether peer relations play a causal role in thegenesis of disorder and whether, conversely, rela-tionships with peers serve protective functions indevelopment.

We believe that the study of children’s relation-ships, with peers as well as with family members andother adults, is best done in appreciation of the factthat children’s lives are bound up with complex so-cial networks (see also Hay & Nash, 2002). Ratherthan entering into ultimately unsatisfying debatesabout the differential contribution of parents andpeers to children’s development, it is important tostrive to understand the interplay of different typesof relationships within children’s social networks.Hence the developmental progressions in children’sabilities to relate to peers that we describe must beseen in the context of the other important relation-ships in children’s lives.

Nevertheless, for present purposes, to limit thescope of this review, there are good reasons toexamine peer relations in their own right. Most hu-man infants are born into a world that contains otherinfants. Many are cared for in group settings, andmost others encounter peers in the course of theirdaily lives, as their parents and siblings spend timewith other parents and children. As children growolder, their social networks become increasinglycomplex, containing friends and acquaintances un-known to their family members. Therefore, to be ableto predict which children will find it most difficult torelate to peers, it is important to examine peer rela-tionships in longitudinal perspective, and to seekantecedents to problems in the first years of life. With

Journal of Child Psychology and Psychiatry 45:1 (2004), pp 84–108

� Association for Child Psychology and Psychiatry, 2004.Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA

its focus on the early determinants of problematicpeer relationships, our review complements the re-cent Annotation on peer relations for this journal(Deater-Deckard, 2001). The model proposed herefocuses primarily on development from early infancyto late childhood. We do not describe adolescent peerrelations in great detail, but do examine adolescentoutcomes when evaluating the contribution of peersto the genesis of disorder.

Because of our primary focus on links betweenpeer relations and childhood disorder, we have re-viewed the relevant literature in developmental psy-chology and child psychiatry. Much of this researchpertains to the lives of children in Western industri-alised nations. We have cited cross-cultural evidencewhere possible. However, we have not conducted athorough review of the ethnographic literature onchildren’s relationships with peers in many differentcultures (for further consideration of the issues thatarise when making comparisons of children’s socialexperiences across cultures, see Harkness, 2002;Shweder et al., 1998).

Beginnings

The study of the beginnings of peer relations hasebbed and flowed with fashion in developmental re-search, and many important findings date frombursts of activity in the 1930s and 1970s. Here wedraw on some of those classic studies, as well asmore recent observations, to sketch the developmentof the capacity to relate to one’s agemates.

Emotional reactions

For infants in the industrialised West, life often be-gins in hospital, and it has long been thought thatnewborn infants influence their peers in the hospitalnursery by making them cry (Sagi & Hoffman, 1976;Simner, 1971). Thus, the matching of negativeemotion with another infant constitutes the firstform of peer interaction that is easily documented.Contagious crying is not coincidental (Hay, Nash, &Pedersen, 1981), but little is known about individualdifferences in sensitivity to the emotions of a peer. Itis possible that variations in responsiveness toyoung peers may relate to general individual differ-ences in emotionality and behavioural inhibition,and indeed to early emerging differences in neuro-endocrine functioning (e.g., Lewis & Ramsay, 1995;Gunnar & Donzella, 2002).

Communicative gestures

By the second half of the first year of life, infantssmile at, reach toward and touch other infants(Hay, Nash, & Pedersen, 1983; Vandell, Wilson, &Buchanan, 1980). These behaviours appear at aboutthe same time or shortly after they first appear with

adult caregivers, and so competence within bothtypes of relationships develops in parallel (Hay,1985; Vandell, 1980). An important development inparent–infant relationships is the capacity to engagein interactions that have a topic, namely, where in-fant and parent interact with reference to a partic-ular object (Bakeman & Adamson, 1984).1 Such‘topic-related’ interaction between infant peers ischaracteristic of the second year of life, when infantsdirect their peers’ attention to toys, food and otherobjects by pointing, holding up, or offering suchobjects to their peers (e.g., Eckerman, Davis, &Didow, 1989).

Dyadic interactions

Contingency. Contingent responsiveness is an im-portant dimension of infants’ interactions withcaregivers and other adults, which have been de-scribed as dialogic in nature, with a to-and-fro be-tween infant and adult that resembles a turn-takingsequence. Contingent peer interactions can also beobserved as early as 6 months of age, in terms ofsequences in which infants touch each other or toysheld by their peers; the rate of such contingent se-quences distinguishes particular pairs of babiesfrom other dyads (Hay et al., 1983). Infants are bet-ter able to sustain interactions with peers if toysare not present (Hay et al., 1983; Jacobson, 1981;Vandell et al., 1980).

Cooperation, sharing and responses to peers’distress. Although young children were longthought to be incapable of engaging in cooperativeplay, one-year-olds have been observed to engage incooperative games with their peers, taking part insustained interactions that entail mutual engage-ment, repetition of key actions, alternating of turns,and a playful, ‘non-literal’ quality (Ross, 1982).Across a number of different cultures the joyful,exuberant quality of toddlers’ cooperative play hasbeen noted (Løkken, 2000).

Other early forms of prosocial behaviour (sharing,helping, and comforting peers who are in distress)emerge around the first birthday. Some investigatorsreport an increase in prosocial behaviour over thesecond year (Zahn-Waxler, Radke-Yarrow, Wagner,& Chapman, 1992), but studies of one- and two-year-olds with their peers show no clear trends (Hay,Caplan, Castle, & Stimson, 1991; Hay, Castle,Davies, Demetriou, & Stimson, 1999); indeed, one-year-olds are significantly more likely than two-year-olds to share in response to peers’ requests forobjects (Hay et al., 1991). Interactions involving the

1 Some investigators refer to object-focused interaction as ‘tri-

adic’; however, in the context of peer relations, this terminology

is confusing. In this paper the term ‘triadic interaction’ refers to

exchanges amongst three individuals.

Peer relations in childhood 85

giving of objects decline over time (Eckerman et al.,1989).

Conflictand thebeginningsofaggression. Althoughone-year-olds share and sustain cooperative inter-action, they also engage in conflict with peers andsometimes use force to pursue their aims. Mostconflicts between young peers concern the posses-sion of toys and other objects (Shantz, 1987); an-other common source of problem is one infant’sviolation of another’s personal space (Hay & Ross,1982). Although early writers characterised infants’conflicts as ‘socially blind’ pursuits of interest inparticular toys (Maudry & Nekula, 1939), conflictsbetween young peers are characterised by commun-icative gestures, and signs of social influence (Hay &Ross, 1982).

Two ways in which toddlers direct force againsttheir peers – hitting them and wresting toys fromtheir grasp – may be precursors to later forms ofpersonal and instrumental aggression, respectively.The instrumental grabbing of toys, which is morecommon, declines over time, becoming less frequentthan spontaneous sharing; in contrast, hittingshows no significant decline over time, remaining atstable, low rates (Hay, Castle, & Davies, 2000).Younger toddlers show more frequent, brief aggres-sion, whereas older toddlers show fewer, but moresustained, aggressive behaviours (Fagot & Hagan,1985). These precursors to aggression peak around30 months of age (Hay, in press; Tremblay, 2001).

Adult behaviour affects toddlers’ responses toconflict. Mothers frequently intervene in their chil-dren’s conflicts with peers, usually supporting thepeers rather than their own children; mothers ofdaughters are especially likely to urge yielding topeers (Ross, Tesla, Kenyon, & Lollis, 1990). However,the presence of mothers in playgroup or domesticsettings can inhibit positive interaction betweenpeers (Field, 1979; Rubenstein, Howes, & Pedersen,1982).

Dyadic preferences and the beginningsof friendship

Infants do not respond in an undifferentiated wayto other infants. In general, infants are more likelyto respond positively to unfamiliar peers than tounfamiliar adults (Brooks & Lewis, 1976). Further-more, infants interact in different ways with familiaras opposed to unfamiliar peers (Stefani & Camaioni,1983; Young & Lewis, 1979). Infants and toddlerswho spend their daily lives together develop clearpreferences for certain companions, evident in stud-ies in day-care centres (Howes & Phillipsen, 1992)and in kibbutz toddler houses (Zaslow, 1980). Tod-dlers who are cared for in mixed age groups some-times prefer to play with older peers rather than withagemates (Rothstein-Fisch & Howes, 1988). Pre-ferences for peers of one’s own sex also emerge in the

toddler years, particularly evident on the part ofolder girls (Howes & Phillipsen, 1992; Hay et al.,1999). However, preferences for particular peers arealso related to each toddler’s behavioural traits(Howes & Phillipsen, 1992).

Individual differences in early competencewith peers

Consistent individual differences have been found inthe quality of play with peers during the first years oflife (Howes, 1980). Variations amongst individualinfants and toddlers are predicted by dimensions ofthe broader social environment, including experi-ence with siblings (Vandell, Wilson, & Whalen, 1981;Vandell & Wilson, 1987), attachment figures (East-erbrooks & Lamb, 1979; Pastor, 1981; but see alsoHowes, Matheson, & Hamilton, 1994), and othercaregivers (Howes et al., 1994).

Is this individual variation in early peer relationsat all consequential for children’s later development,and for their risk for disorder? There is reason tobelieve that stable individual differences in compet-ence with peers emerge in the first years of life. Forexample, longitudinal assessments of a group of 55children seen at intervals between 1 and 9 years ofage indicated that toddlers’ competent play withpeers predicted later competence as preschoolersand 9-year-olds (Howes & Phillipsen, 1998). Thosetoddlers who were capable of engaging in complexplay with peers were more prosocial in the preschoolsetting and less likely to show either aggressive orwithdrawn behaviour as 9-year-olds.

It is of particular interest to note that individualdifferences in precursors to externalising and inter-nalising problems – and consequent risk for dis-ruptive behaviour disorder and emotional disorder,respectively – may be discerned in the first years oflife. Aggression in the toddler years is often thoughtto be normative. However, closer inspection of theliterature on peer relations shows that, althoughmany toddlers may use force occasionally, the fre-quent, proactive and dysregulated use of aggressionin early life is a minority phenomenon that may in-dex risk for later externalising problems (Hay, inpress). Stable individual differences in aggressionand externalising problems emerge in the toddleryears (Fagot & Leve, 1998; Hay et al., 2000; Keenan& Wakschlag, 2002;1 Rubin, Bukowski, & Parker,1998; Rubin, Burgess, Dwyer, & Hastings, 2003;Shaw, Keenan, & Vondra, 1994). In one studygreater long-term stability was found for boys(Cummings, Iannotti, & Zahn-Waxler, 1989) butother studies have noted stability in girls’ aggression(Hay et al., 2000; Rubin et al., 2003).

With respect to internalising tendencies, there isevidence that socially inhibited behaviour in pre-school settings is also predicted by earlier patterns ofinhibition with peers. Toddlers who show beha-vioural inhibition in the presence of peers are likely

86 Dale F. Hay, Alexandra Payne, and Andrea Chadwick

to continue to be shy 4-year-olds (Rubin, Burgess, &Hastings, 2002).

Underlying skills

The foregoing evidence for stable individual differ-ences in competence with peers in the first threeyears of life suggests that attempts to prevent peerproblems from developing may need to begin ratherearlier than initially thought. An infant’s ability toengage successfully in interaction with peers restson a number of achievements in the first years of life,whereby newly emerging cognitive skills and self-regulatory abilities equip young humans to engagein and sustain social interaction with a number ofdifferent partners. We propose that the followingprocesses underlie harmonious interaction betweenyoung peers (see Figure 1). Deficits in any or all ofthese skills may interfere with successful peerrelations.

Joint attention. Social interaction depends on theability to coordinate attention with another person;this ability develops with adult partners over the lasthalf of the first year of life (e.g., Butterworth, 2001;Striano & Rochat, 1999). Joint attention in the sec-ond year predicts theory of mind skills two yearslater (Charman et al., 2001). Initiating bids for otherpeople’s attention and responding to the bids ofothers are both associated with patterns of frontallobe activation and deactivation (Mundy, Card, &Fox, 2000).

The regulation of gaze and the use of commu-nicative gestures are important components of earlypeer interaction, particularly when more than one

peer might be present (Ishikawa, 2003). Gesturesthat serve to regulate another person’s attention –e.g., pointing to or holding up objects for another tosee – are incorporated into peer interaction as earlyas 12 months of age (Hay et al., 1991); thus suc-cessful peer interaction requires being able to readas well as produce such signals. Infants and toddlerswith deficits in joint attention skills may be illequipped to sustain bouts of nonverbal interactionwith their peers. Peers, unlike parents, cannot takeinto account the developmental level of their part-ners and scaffold the interaction accordingly.

Emotion regulation. Joint attention between an in-fant and another person also has a positive affectivecomponent that contributes to pleasurable socialinteraction (Mundy, Kasari, & Sigman, 1992). Ingeneral, the ability to regulate one’s emotion, inparticular negative reactions to momentary frustra-tion, is an important skill called upon by the de-mands of peer interaction. It has been reported thatinfants generally show lower levels of affect withpeers than with their mothers (Adamson & Bake-man, 1985); however, challenging situations pro-duce a broader range of individual differences inemotion regulation (Miller, McDonough, Rosenblum,& Sameroff, 2002), and conflict with peers may beone such emotionally charged setting. It should benoted that self-regulatory behaviours are more suc-cessful in reducing anger than in reducing fear (Buss& Goldsmith, 1998), which has implications fortoddlers who are aggressive as opposed to fearful oftheir peers. Interventions that focus on emotionregulation strategies may be more successful withthe former than the latter.

Differences in the ability to regulate emotion harkback to excessive crying in very early infancy (e.g.,Stifter & Spinrad, 2002) and are also linked to at-tachment security (Diener, Mangelsdorf, McHale, &Frosch, 2002). Thus problems in emotion regulationmay mediate some links between parent–child andpeer relations.

Inhibitory control. Although a generally inhibitedtemperament in infancy predicts later shyness withpeers (Rubin, Burgess, & Coplan, 2002), inhibitoryprocesses are required for successful social interac-tion. Inhibitory control is an important dimension ofexecutive function, applied in response to novelchallenges (Rabbitt, 1997). When meeting newacquaintances, infants must inhibit impulses to ex-plore their peers as they would objects; intrusions onpeers’ personal space provoke conflict (Caplan&Hay,1991; Hay & Ross, 1982). To the extent that ex-ploratory behaviour is primarily directed to novelstimuli, this type of exploration of peers may natu-rally decrease upon extended acquaintance. Thoseinfants and toddlers who continue to use physicalexploration with peers rather than more distal, com-municative gestures or words may be disadvantaged.

Joint

Attention

Emotional

Regulation

Inhibitory

Control

Imitation

Causal

Understanding

Language

Harmonious

Interactions with Peers

Figure 1 Very early competence with peers is likely tobe supported by emotional, cognitive and behaviouralskills, as depicted in the hypothesised developmentalmodel


Infants must also suppress their tendencies toseize toys held by their peers, a behaviour that isalready present at 6 months of age, especially in boys(Hay et al., 1983). Attempts to grab toys from peersdeclines over time (Hay et al., 2000); children whocannot inhibit the impulse to grab may be reacted toquite negatively by their peers.

Imitation. Imitation is an important means wherebyinfants and toddlers succeed in interacting harmo-niously with their peers (Eckerman et al., 1989).Matching the peer’s behaviour is an importantstrategy for interaction amongst younger toddlers;this strategy emerges around the same time in verydifferent cultural contexts (Eckerman & Whitehead,1999). With age, interactions are more likely to becomplementary, rather than precisely imitative (Ca-maioni, Baumgartner, & Perucchini, 1991). How-ever, both imitative matching of another’s behaviourand complementary activities play an important partin maintaining cooperative games (Ross, 1982).

Causal understanding. Successful social interac-tion depends on a mutual understanding of eachparticipant’s status as an active and intentionalagent. For young peers, therefore, an important taskis the appreciation that another infant may be theagent of one’s own experiences, be they good or bad.Conversely, it is important to appreciate whenunpleasant events occur by accident, and are notintentionally caused by one’s peers. This level ofsocial understanding is underpinned by more gen-eral comprehension of cause-and-effect relation-ships in the world.

Current research on infants’ understanding ofcausality in the physical world suggests thatunderstanding of cause-and-effect relations devel-ops over infancy, with important strides being madein the last half of the first year of life (Oakes, 1994;Desrochers, Ricard, Decarie, & Allard, 1994).Younger infants are sensitive to causal relationshipsin some tasks but not others (Belanger & Desro-chers, 2001; Oakes, 1994).

Of particular relevance to peer relations is thechallenge of perceiving cause-and-effect in socialinteraction, which constitutes one of the first earliestforms of social understanding. Even quite younginfants are sensitive to human-like motion that adultobservers interpret as causal (Rochat, Morgan, &Carpenter, 1997). Infants also show sensitivity tophysical displays that suggest human beings areapproaching (Schlottman & Surian, 1999), and tonegative vs. positive social interactions (Premack &Premack, 1997). However, only somewhat oldertoddlers appear to comprehend more complex causalsequences, such as those found in complex bouts ofsocial interaction (Cohen, Rundell, Spellman, &Cashon, 1999). It is not yet known whether there arestable individual differences in causal under-standing in the toddler years, nor whether these

impact on actual interaction with peers. However,causal understanding is associated with two sets ofskills used in social contexts, infants’ intentionalcommunications (Harding & Golinkoff, 1979) andsocial referencing (Desrochers et al., 1994).

Toddlers also show an understanding of theirpeers as intentional agents in the course of conflictsover toys, when one toddler reaches for or gesturestoward a particular object that is in the possession ofthe peer. In response to such gestures, peers areoften likely to withdraw the object, or move awayfrom the peer (Hay et al., 2000), or, less frequently,give it to the peer (Hay et al., 1991, 1999). The for-mer, suspicious responses to peers’ indications ofinterest in an object are correlated with aggression,especially proactive aggression (Hay et al., 2000),whereas the latter, more generous behaviour ispositively correlated with sensitivity to peers’ dis-tress (Hay et al., 1999). Thus an understanding ofthe intentions of peers as social agents may beviewed as an early step in the development of a the-ory of mind, and is correlated with both aggressionand prosocial behaviour.

Language. As soon as toddlers acquire some words,they use them in interaction with their peers, al-though nonverbal means of interacting remain moreimportant for some time (Eckerman et al., 1989).Speech to peers is used both as a means of expres-sing one’s desires and pursuing one’s aims in con-flict, without using overt aggression, and in attemptsto resolve conflict (Caplan et al., 1991). Toddlers’comments concerning the possession and use ofobjects, and the display of manners, are related toboth aggression and sharing (Hay, 2000). In a largesample of 19-month-old twins, expressive vocabu-lary was significantly and negatively related tophysical aggression (Dionne, Tremblay, Boivin,Laplante, & Perusse, 2003).

Implications for children with developmentaldisorders

The foregoing consideration of the various skills thatunderlie successful interaction with peers suggeststhat young children with developmental disordersmay be at a particular disadvantage. For example,for children with disorders in the autistic spectrum,impairment of joint attention skills (Charman et al.,1997) and deficits in imitation (Rogers, Hepburn,Stackhouse, & Wehner, 2003) might hinder adjust-ment to peers in group settings. It is also possiblethat language delay might impede peer relationships(Dionne et al., 2003). Thus some of the problemswith peer relations seen in mainstreamed preschoolclassrooms (e.g., Guralnick, Connor, Hammond,Gottman, & Kinnish, 1996; Guralnick et al., 1998)may partly derive from deficits in skills typicallyacquired in the course of the first years of life.This possibility deserves further investigation.


Conversely, comparisons of children with develop-mental disorders with typically developing childrenmay draw attention to the processes that underliesuccessful peer interaction.

Consolidation of social skills and group relations

During the next few years of life, most childrenconsolidate their social skills and form pleasurablerelationships with other children. In the industrialWest, this often takes place in the context of organ-ised groups, in nursery schools, day-care centresand recreational activities. New ways of relating topeers come to prominence in the preschool years. Aschildren spend more time in groups, their experiencecomes to be organised with respect to particularfriendships, gender-related preferences, and dom-inance hierarchies. It is against this background ofincreasing social competence and more rigid groupstructure that some children’s difficulties in dealingwith their peers become apparent.

Types of interactions in early childhood

Conversation. As children acquire language andfluent vocabularies, they spend more time simplytalking to their peers. Conversational competence isgreatest in familiar situations, where children candraw on reserves of ‘script knowledge’ for their con-versations with peers (Furman & Walden, 1990).Preschoolers use their conversational abilities formany different purposes; for example, they may seekinformation, argue, agree or disagree with peers, andtell stories (Kuentay, 2001; Pickert, 1985). Youngchildren take their conversations seriously: infor-mation received in conversation with peers maydistort preschoolers’ memories of actual events(Principe & Ceci, 2002).

Conversational competence is one factor that af-fects the extent to which children interact success-fully with, and are accepted by, their peers (Black &Hazen, 1990). In one longitudinal study of children’sconversations with their classmates, responsivenessto peers’ speech predicted social status a year later(Kemple, Speranza, & Hazen, 1992). Children whostutter sometimes meet with negative reactions fromtheir peers (Meyers, 1990; Davis, Howell, & Cooke,2002).

Cooperation. We have seen that the beginnings ofcooperation lie in infants’ and toddlers’ first attemptsto coordinate their activities with those of their peers.The ability to cooperate with others develops furtherin the preschool years. In situations where access toresources is restricted, some children may behavecooperatively and others competitively. Cooperationvs. competition is affected by gender and culture(e.g., Sparkes, 1991), and by attachment status(Kerns & Barth, 1995). The choice of competition or

cooperation may depend on the identity of the peers,with children cooperating more equitably withfriends (LaFreniere & Charlesworth, 1987). Pre-school children gradually learn to balance cooper-ative and competitive behaviour whilst dealing withtheir peers (LaFreniere, 1996). However, differentnorms may hold for the two sexes. Sociometricmeasures indicate that competitive boys, but notcompetitive girls, are liked by their same-sex peers(Sebanc, Pierce, Cheatham, & Gunnar, 2003).

Helpfulness, sympathy and kindness. Meta-analyses have shown that the capacity for prosocialinteractions involving helping, sympathising andconcern for others increases over the years of child-hood (Eisenberg & Fabes, 1998). Helping and shar-ing are the more common forms of prosocialbehaviour in preschool groups (e.g., Babcock, Har-tle, & Lamme, 1995). Although preschoolers onlyrarely intervene actively to alleviate the distress oftheir peers, they are very likely to pay close attentionto such distress and, when interviewed, can suggestconstructive ways of comforting or helping a dis-tressed peer (Caplan & Hay, 1989).

Gender differences in prosocial behaviour emergeduring the preschool years, although the extent ofthese depends on the mode of assessment (Eisenberg& Fabes, 1998) and on the social context, in par-ticular whether children are interacting with peers ofthe same or opposite sex (e.g., Burford, Foley, Roll-ins, & Rosario, 1996).

Conflict and aggression. Conflict over tangible re-sources decreases in frequency over the preschoolyears; instead, children begin to quarrel over socialissues, such as the violation of gender-role stereo-types, or the inclusion of particular children in groupactivities (Hay, 1984; Shantz, 1987). For example, ina study of 400 preschoolers, two-year-olds weresignificantly more likely to engage in conflict overresources than were three- and four-year-olds; theolder children were more likely to engage in disputesabout forms of play and ideas (Chen, Fein, & Tam,2001). Thus, the frequency of conflict does not de-crease over the preschool years, but changes occurin the content of disputes and children’s roles inresolving their conflicts (Chen et al., 2001). Aggres-sive strategies tend to bring an interaction to anabrupt end, whereas conciliatory attempts to resolveconflict lead to continued interaction following thedispute (Caplan et al., 1991; Laursen & Hartup,1989).

Early childhood is a time when gender differencesin conflict and aggression become quite marked.Girls and boys show similar approaches to conflict inthe toddler years (Caplan et al., 1991; Hay & Ross,1982; Hay et al., 2000); gender differences inaggression emerge more clearly between three andfour years of age (Keenan & Shaw, 1997; Loeber &Hay, 1997). Furthermore, girls and boys begin to


take qualitatively different approaches to conflict.Girls express more positive emotion than boys do inthe course of conflict (Garner, Robertson, & Smith,1997). When interviewed, preschool girls are morelikely than their male counterparts to advocatesocially acceptable approaches to conflict with peers(Hay, Zahn-Waxler, Cummings, & Iannotti, 1992;but see also Iskandar, Laursen, Finkelstein, &Fredrickson, 19955 ).

Pretend play. A striking feature of peer relations inthe early childhood years is the extent to whichinteractions between young peers reflect the life ofthe mind, and indeed a degree of intersubjectivitybetween children’s minds. Preschoolers spend con-siderable periods of their time in imaginative pre-tence, which is an important force in theconsolidation of early friendships (Gottman, 1983)and the persistence of those friendships into laterchildhood (Dunn, Cutting, & Fisher, 2002).

The ability to pretend that imaginary things arereal emerges in a rudimentary fashion in the secondyear of life (e.g., Howes & Farver, 1987; Musatti &Mayer, 1987), and increases in complexity thereafter(e.g., Howes & Matheson, 1992), partly as a functionof language development (Garvey & Kramer, 1989).Pretend play is seen across cultures (e.g., Farver,Kim, & Lee, 1995), although themes may vary acrosscultural settings. Socialisation practices in differentcultural groups appear to affect the frequency andcontent but not the complexity of pretend play(Farver, Kim, & Lee-Shim, 2000).

Both girls and boys show pretend play, thoughagain the topics and themes of the pretend episodesmay differ (Black, 1989). Pretend play is found to beless complex under conditions of socioeconomicdisadvantage (Doyle, Ceschin, Tessier, & Doehring,1991) and less adequate child care settings (Howes &Matheson, 1992). Because pretend play fostersfriendships, factors that disrupt it may generallyinterfere with harmonious peer relations.

Group relations in the preschool years

The major transition that takes place in the pre-school years concerns the ability to relate to peers ingroups, as opposed to one-to-one relationships.These groups may take the form of informal, oftenunsupervised, mixed-age play groups in villages andneighbourhoods in many cultures, or more formal,age-graded organisations in day nurseries and pre-school classrooms. When young children begin tospend time in groups, some children interact pre-ferentially with others, and so groups of childrenbecome structured in particular ways.

Interaction beyond the dyad. Infants’ interactionswith peers are usually dyadic, and it is not com-pletely clear when in development children becomeable to interact with more than one person at a time;

the first documentation of triadic interaction comesfrom studies of family groups (e.g., Hay, Vespo, &Zahn-Waxler, 1998; Ross et al., 1996). When two-year-olds are tested in groups of three, about a fifthof the interaction is actively triadic (Ishikawa, 2003).However, as young children spend more time innursery schools, day-care centres and organisedplaygroups, they begin to engage in more formalgroup activity. In these groups of young peers, socialstructures soon emerge. The following patterns canbe seen.

Friendships. The early preferences for particularpeers discerned as early as the second year of life(e.g., Howes & Phillipsen, 1992) flower intoreciprocated friendships in the preschool years.Although it was once thought that early friendshipsreduce to fleeting interactions around common toys,more recent research has confirmed the complexityand enduring quality of early friendships (e.g.,Gottman, 1983), some being maintained from in-fancy through early childhood (Howes & Phillipsen,1992). Preschool children’s cognitive understandingof friendship is attested to by the fact that they makedifferent judgements about other children’s trans-gressions, depending on whether or not they arefriends (Slomkowski & Killen, 1992). Thus any groupof preschool children becomes structured by thevarious friendships embedded within it.

Gender segregation. Initially young children formfriendships with same- and opposite-sex peers, butgradually come to prefer to interact with members ofthe same sex, a phenomenon referred to as gendersegregation. Maccoby (1998) has gone so far as todescribe girls and boys as living in two separateworlds through most of childhood, which she con-siders a critical influence on the gender differencesthat characterise adult life. The phenomenon is seenacross cultures (e.g., Hold-Cavell, Attili, & Schleidt,1986). Increasing gender segregation of preschoolchildren’s groups impacts on their friendships: girls’early friendships tend to be more stable than those ofboys (Howes & Phillipsen, 1992).

Gender segregation may derive from the differ-ential responding to same- and opposite-sex peersthat begins to be evident in the preschool years. Forexample, two- and three-year-old girls are morelikely to share with other girls than with boys, par-ticularly in response to peer requests; in observa-tions of toddlers with familiar peers, girls never gaveboys what they asked for (Hay et al., 1999). Conci-liatory gestures to resolve conflict are more likely tobe directed to same-sex peers (Sackin & Thelen,1984). Differences between girls and boys in the useof negotiation as opposed to coercive means toresolve disputes are seen more clearly when inter-acting with same-sex peers (Burford et al., 1996).Gender segregation may also derive from preferencesfor different types of activities, and from girls’


preferences for dyadic as opposed to group interac-tion (Benenson, 1993).

Dominance hierarchies. Research by human eth-nologists and sociobiologists has called attention tothe phenomenon of dominance in children’s groups(e.g., Strayer & Strayer, 1976). The notion of dom-inance implies that, when children are engaging inconflict with other group members, some are morelikely than others to yield to the peer. Even amongsttoddlers, the tendency to initiate conflict is a stablecharacteristic of individuals, but the tendency toyield to the peer depends very much on the identityof the peer (Hay & Ross, 1982). Transitive dominancestructures within young children’s groups are seenas early as the toddler years (Bakeman & Brownlee,1982). Dominant children tend to engage in conflictmost often with other dominant children (Strayer &Strayer, 1978). However, dominance relations affectplay as well as conflict; rough-and-tumble play,which is shown more often by boys than by girls, is acontext in which dominance is established and tes-ted (Pellegrini & Smith, 1998). It should be noted,however, that dominance hierarchies are apparentfor girls as well as boys (Sebanc et al., 2003; Strayer& Strayer, 1976). The influence of dominant childrenover their peers is attested to by the fact that dom-inant children find it easier to deceive their peers,often smiling while telling lies in a mannerreminiscent of some successful adult politicians(Keating & Heltman, 1994).

Peer acceptance and rejection. The foregoing phe-nomena serve to structure preschool children’sgroups, and all operate to determine the place ofindividual children in the group. Some children areaccepted by most of their peers, whereas others areignored or actively rejected by peers. Because peeracceptance is an important issue with respect to thechild’s risk for disorder and, conversely, becausechildren with disorders may be rejected by theirpeers, considerable attention has been paid to thedeterminants of peer acceptance in the preschoolyears. A variety of age-appropriate methods havebeen developed to measure sociometric preferencesand dislikes (e.g., Asher, Singleton, Tinsley, &Hymel, 1979; Musun-Miller, 19902 ).

It is important to note once again that peer rela-tionships take place within children’s broader socialnetworks, and so peer relations are affected by chil-dren’s relationships with parents (e.g., Kerns &Barth, 1995; Kochanska, 1992; Moss, Goesselin,Parent, Rousseau, & Dumont, 1997), siblings (e.g.,Herrera & Dunn, 1997), and teachers (Howes,Matheson, & Hamilton, 1994). Marital conflict be-tween parents is sometimes reflected in children’srelationships with peers (e.g., Gottman & Katz,1989). Mothers’ own support networks influencetheir children’s acceptance by peers (Bost, 1995;Melson, Ladd, & Hsu, 1993). Perhaps because of

access to wide networks, mothers who take theinitiative in arranging children’s contacts with peershave children who themselves initiate such contactsand are accepted by their peers (Ladd & Hart, 1992).However, despite mothers’ efforts, children who dis-play particular styles of interaction may fail to gainacceptance by their peers. Of particular importanceis the relative frequency of positive and negativeinteractions with peers.

Individual differences that affect peer acceptance

Prosocial behaviour. Stable individual differencesin prosocial behaviour emerge in the preschoolyears; long-term stability from early childhood toadolescence has been reported for spontaneoussharing (Eisenberg et al., 1999) and helpfulness(Cote, Tremblay, Nagin, Zoccolillo, & Vitaro, 2002).Furthermore, cooperation in early childhood pre-dicts low rates of externalising problems in earlyadolescence (Hay & Pawlby, 2003). Prosocial beha-viour is a key factor in gaining acceptance withpeers; preschool children report that they like theirprosocial classmates (Denham et al., 1990; Ladd,Price, & Hart, 1988). Sometimes it is the absence ofprosocial behaviour in kindergarten, not the pre-sence of aggression, that best predicts which chil-dren will be rejected a year later (Vitaro, Gagnon, &Tremblay, 1990).

Aggressiveness. The individual differences inaggressiveness that emerge in toddlerhood (e.g.,Cummings et al., 1989; Hay et al., 2000; Rubinet al., 1998) consolidate during the preschool years.Although for most children aggression occurs morefrequently in early than in later childhood (e.g.,Broidy et al., 2003), even then it is not a majorityphenomenon. In a normative sample of childrenbetween 3 and 5 years of age, in which a variety ofdifferent antisocial acts were recorded, approxim-ately 40% of the children exhibited at least oneantisocial act each day; however, only 10% showedantisocial behaviour at very high rates (Willoughby,Kupersmidt, & Bryant, 2001). These early individualdifferences tend to persist. Stability in aggressionand other externalising problems is evident fromthree years of age on (Loeber & Hay 1997; Lahey,Waldman, & McBurnett, 1999). Indeed, earlyaggression is part of the profile of behaviours anddeficits that characterise life-course persistent anti-social behaviour (Moffitt & Caspi, 2001).

When interviewed, young children condemnaggression that is hostile (as opposed to instru-mental), and not used in self-defence (Ferguson &Rule, 1988; Hay, Zahn-Waxler, Cummings, & Ian-notti, 1992). Given these views, it is not surprisingthat highly aggressive children are not well liked bytheir peers. This holds true not only for overt, phy-sical aggression but the sort of ‘relational aggression’that is shown more often by girls than by boys; this


sort of behaviour similarly leads to peer rejection(Crick, Casas, & Mosher, 1997). Furthermore, peerrejection may be underpinned by beliefs about thestability of aggressive behaviour in particular peers.Preschool children who believe in the stability ofantisocial behaviour over time are less likely toexhibit prosocial behaviour and are more acceptingof aggressive behaviour during peer interactions(Giles & Heyman, 2003). This implies that, evenbefore children begin primary school, they alreadyhold general belief patterns that they use to interpretpeer relationships and respond in a differential wayto aggressive and non-aggressive peers.

Shyness. Some children seem to find it difficult toengage with their peers to any great extent. Just asthere is stability in aggressive behaviours from pre-cursors in infancy, so general behavioural inhibitionin infancy predicts shyness with peers in the pre-school years (Rubin et al., 2002). Moreover, there iscontinuity from shyness across the childhood years,with many features similar to shyness amongstyoung adults (Asendorpf, 1992). Individual differ-ences in shyness have been linked to insecureattachment relationships in infancy, although thatmay depend on the infant’s temperament (Rubinet al., 2002). However, shyness has also been foundto be associated with mothers’ own social phobias(Cooper & Eke, 1999), suggesting a possible biolo-gical transmission of social anxiety from one gen-eration to the next. There has been considerableinterest in the biological bases of shyness. Severalreports suggest that shyness is associated with blueeye colour in males (e.g., Coplan, Coleman, & Rubin,1998) and with neuroendocrine functioning (e.g.,Dettling, Gunnar, & Donzella, 1999).

In evaluating the effects of shyness on peer rela-tions, it is important to take into account the child’ssex (Stevenson-Hinde & Glover, 1996), and also thedegree of heterogeneity in the category of shy chil-dren (Asendorpf, 1990). In a review of the literatureon solitary play, Katz and Buchholz (1999) focusedon the nature of solitary play in a preschool setting.Of those children who played alone more than wastypical, four groups emerged: shy, non-shy, isolatedand depressed children. Only the children in thelatter two groups present clear concerns during thepreschool years. It appears that children in thesegroups already show signs of psychopathology thatset them apart from other shy and non-shy chil-dren.

Underlying processes

In examining the individual differences in preschoolchildren’s relations with peers, and the conse-quences of those differences for peer acceptance andrejection, it is again important to attempt to identifythe underlying processes that contribute to suc-cessful peer interaction. The following processes

seem particularly important for peer relations in thepreschool years and beyond (see Figure 2).

Emotionality and emotion regulation. Children’sabilities to display and regulate emotions appropri-ately are linked to the quality of their friendshipswith their peers (McDowell, O’Neil, & Parke, 2000).Effective regulation strategies are positively corre-lated with ratings of social competence.

It is important to examine the joint contribution ofemotionality and emotion regulation, rather than theindependent contributions these constructs make tochildren’s competence with peers (Fabes et al.,1999). For example, both the experience of intenseanger, as well as the possession of strategies fordealing with anger, affect peer relations in earlychildhood. Inappropriate strategies for managinganger, such as aggression, or venting of anger, arelinked to higher rates of general emotionality, whichin turn is associated with lower social status(Eisenberg, Fabes, Nyman, Bernzweig, & Pinuelas,1994).

Skill at regulating emotions can help children dealwith their own anger. For example, observations of4-year-olds’ free play with peers were recorded every3 months during the spring semester of the schoolyear (Fabes et al., 1999). The findings indicated thatchildren’s experience of negative emotions was neg-atively correlated with social competence. However,the child’s ability to regulate emotion moderated thatrelationship; the higher the level of emotion regula-tion, the weaker the association between negativeemotional responding and social competence.

The experience of negative emotions, and prob-lems in regulating those feelings, have long-termconsequences for peer relations. In a longitudinal

Negative Emotion and Problems in Regulating

Emotion

Problems in

Social

Understanding

Problems in Executive

Function

Shyness

Aggressiveness

Lack of Prosocial Skills

Problems in Gaining

Acceptance by

Peers

Figure 2 In childhood, problems in emotion regulation,social understanding and executive function are pro-posed to underlie individual differences in shyness,aggression and prosocial behaviour, which in turninfluence the extent to which preschoolers are acceptedor rejected by their peers


study, (Eisenberg et al., 1997), emotionality and self-regulation at age 4 predicted social functioning up to4 years later. High levels of negative emotionality andlow levels of regulatory abilities were linked toproblem behaviours; these associations were stableover time.

Problems in regulating emotion that disrupt rela-tions with preschool peers may persist into laterchildhood, and interfere with competent, prosocialbehaviour. For example, in a sub-sample of childrentaking part in the NICHD Study of Early Child Careand Youth Development (Gallagher, 2002), negativeemotionality and emotion regulation significantlyinteracted at 7 to10 years of age to predict the child’ssocial competence at age 10. Those children who hadhigh levels of negative emotionality were more pro-social if they also showed a high degree of emotionregulation. These findings suggest that one of themajor developmental tasks of childhood is to learn tocope with one’s own negative emotions; the peersetting may be an arena in which such coping skillsare honed.

Social understanding. Much research has exam-ined the contribution of skill deficits and biases intheory of mind and social information-processing topeer relations. These processes begin to becomeimportant in the early childhood years, when chil-dren achieve an understanding of other people’sdesires and beliefs, and when they become betterable to generate solutions to interpersonal problems.For example, preschool children’s degree of socialunderstanding affects their competence at dyadicinteraction with peers; the peer partner’s degree ofsocial understanding is also important (McElwain &Volling, 2002).

Social understanding may play an important rolein determining which children become rejected.Conversely, peer rejection may interfere with thedevelopment of social understanding. For example,in the sociocultural view of development of theory ofmind, theory of mind skills are thought to developthrough social interactions with peers; the aggres-sive exchanges and lack of positivity that mark peer-rejected children’s social interactions may thusdeprive them of opportunities to develop theory ofmind skills through social interactions (Badenes,Estevan, & Bacete, 2000).

Links between theory of mind and social compe-tence have indeed been found (e.g., Slaughter, Den-nis, & Pritchard, 2002). In particular, understandingof deception and white lies is associated with positivestatus within the peer group (Badenes et al., 2000;Peterson & Siegal, 2002). Theory of mind skills weregreater in those children who had greater verbalskills, a mutual friendship, and were considered tobe popular by the peer (Peterson & Siegal, 2002).However, children who are rejected by their peers arenot necessarily deficient in social understanding:Performance in theory of mind tasks did not differ in

a comparison of peer-rejected children with popularand average children (Badenes et al., 2000).

Problems in social understanding may take theform of biases as well as deficits. One aspect of the-ory of mind is the understanding of another person’sintentions; misunderstanding of another’s intent hasbeen linked to increasingly serious peer problems.Misattribution of peers’ intentions is associated withreactive aggression, involving angry retaliation inresponse to perceived provocation or frustration(Schwartz et al., 1998). Reactive aggression is neg-atively evaluated by peers, and correlated withvictimisation (Price & Dodge, 1989).

Whilst aggression has been linked to peer accep-tance, not all aggressive children are rejected by thepeer group; social cognitive biases appear to play arole in determining which children are rejected(Schonberg, Tussey, & Stickney, 2003). Americanboys aged 9 to11 years from low-income areas, andrepresentative of diverse ethnic backgrounds,attended a summer camp for a period of two weeks.Children who made hostile attributions about peers’intentions were more prone to reactive aggressivedisplays, which in turn predicted peer rejection.A subsequent study of somewhat younger childrenreplicated and extended these findings (Yoon,Hughes, Cavell, & Thompson, 2000). In that study,aggressive-rejected children found it difficult togenerate appropriate responses to social informa-tion. Their aggressive responses led to furtherrejection by the peer group. Thus, not only the mis-understanding of other children’s intentions, butalso an inability to generate alternative means ofsolving social problems, leads to difficulties withpeers.

Executive function. Deficits in social understandingmay interact with broader patterns of cognitivefunction to pose difficulties in relating to peers. In asample of British 4-year-olds, children who werecharacterised as ‘hard to manage’ showed deficits inboth theory of mind and executive function skills(Hughes, Dunn, & White, 1998). The children werefilmed whilst engaged in dyadic play with a friend(Hughes, White, Sharpen, & Dunn, 2000). The ‘hard-to-manage’ children were significantly more antiso-cial than other children, but deficits in theory ofmind were not associated with antisocial behaviour.Rather, a significant and negative associationbetween executive functioning and antisocial beha-viour was found for both ‘hard-to-manage’ andcomparison children. However, when the same chil-dren were filmed in dyadic competitive situations atages 5 and 7 years (Hughes, Cutting, & Dunn, 2001),the combined effects of theory of mind and executivefunction deficits at age 4 predicted antisocial beha-viour in peer relations at age 5 but not age 7 years.

The importance of one dimension of executivefunction, inhibitory control, was further attested toin a study of links between self-regulation at age


3 years and peer relations at age 4½ years in thenationally representative US sample drawn for theNICHD study of day care (Balaraman, 2003). Chil-dren with weak inhibitory control were involved inmore negative exchanges with peers than childrenwho showed good inhibitory control.

Thus, in addition to specific deficiencies in theform of a reduced capacity to understand mentalstates in others, cognitive impairments in the form ofreduced inhibitory control are also implicated in thedevelopment of problems in peer relations. It shouldbe noted, however, that inhibitory control and theoryof mind skills are themselves related; indeed, exe-cutive function abilities may contribute to the youngchild’s growing awareness of other people’s mentalstates (Carlson & Moses, 2001).

The foregoing evidence concerning the problemsfaced by insensitive, aggressive and ‘hard-to-man-age’ children draws attention to problematic rela-tions with peers likely to be experienced by youngchildren with serious disorder. In particular, highlydysregulated preschoolers – those already diagnosedwith Attention-Deficit/Hyperactivity Disorder andOppositional-Defiant Disorder – might be expected tohave particular problems dealing with their peers.Conversely, it may be in response to the challenge oflife in peer groups that these disorders becomeacknowledged and diagnosed.

The emergence of entrenched problemsin relating to peers

In general, friendships flower and group relationsbecome increasingly complex throughout the child-hood years (Rubin et al., 1998). Nevertheless, peerrelations go very badly wrong for some children.Some children do not develop satisfactory friend-ships, and feel very lonely. Others may be bullied,bully others, or both. Both of these phenomena areparticularly relevant when considering the contri-bution of peer relations to risk for disorder.

Loneliness and solitude

The subjective experience of loneliness. Animportant distinction must be made between playingalone and the subjective experience of loneliness(Qualter & Munn, 2002). Furthermore, the phe-nomenology of loneliness is itself multidimensional.Goossens and Beyers (2002) compared six self-rat-ing measures of childhood loneliness. The scalesmeasured four different but interrelated latent con-structs: peer-related loneliness, family-related lone-liness, negative attitudes towards being alone andpositive attitudes towards being alone. These differ-ent sorts of feelings about being alone emerge overthe course of childhood.

The correspondence between solitude and lone-liness seems most marked in the middle childhood

years. Toddlers and preschoolers spend a great dealof time engrossed in constructive solitary play. Inmiddle childhood, however, a tendency to play onone’s own is linked to the phenomenology of loneli-ness (Cassidy & Asher, 1992; Fordham & Stevenson-Hinde, 1999) and to social isolation and depressivesymptoms (Katz & Buchholz, 1999). The significanceof solitude then changes again between late child-hood and early adolescence: For adolescents, soli-tude has a positive effect on emotional state, withthose who spend an intermediate amount of theirtime alone being better adjusted than those whospent little or a great deal of time alone (Larson,1997).

Loneliness and competence with peers. In thechildhood years, loneliness is moderately correlatedwith lower levels of peer competence (Asher, Hymel,& Renshaw, 1984; Asher, Parkhurst, Hymel, & Wil-liams, 1990; Cassidy & Asher, 1992), although theseeffects appear to be mediated by social conditionsand emotional competence. Lonely children arerejected and victimised by their peers, becomingmore rejected over time (Boivin, Hymel, & Burkow-ski, 1995). However, it is important to distinguishdispositional ‘anxious solitude’ from solitude due topeer exclusion (Gazelle & Ladd, 2003). In long-itudinal analyses, the most severe depressive tra-jectories occurred when anxious solitary childrenalso experienced high levels of peer exclusion.Anxious solitude predicted significantly elevatedlevels of peer exclusion at school entry or soon after.Those anxious solitary children who experiencedearly peer exclusion were at greater risk for depres-sive symptoms over the course of middle childhood.Conversely, anxious solitary children who did notexperience significant peer exclusion displayeddecreased levels of anxious solitude over time.

Family factors. Although solitude and subjectiveloneliness may derive from peer exclusion, there arealso important links with parent–child relationships.Children who reported the most loneliness in earlychildhood were those who had been previouslyclassified during infancy as having insecure–ambi-valent attachment relationships (Berlin, Cassidy, &Belsky, 1995). Intergenerational continuities in theexperience of loneliness have also been identified.Positive associations have been found between theloneliness experienced by parents and their childrenat the time of school entry (Henwood & Solano,1994), and during the university years (Lobdell &Perlman, 1986); family conflict heightens feelings ofloneliness (Jones, 1992). Loneliness is also related tothe way in which parents socialise their children toresolve difficulties that arise during interactions withtheir peers; a controlling interaction style shown bymothers and fathers was positively related to chil-dren’s loneliness scores and depressive symptoms(McDowell, Parke, & Wang, 2003). In that study,


mothers’ (but not fathers’) warmth was related tolower levels of reported loneliness in children. Inaddition, parents’ controlling interaction style pre-dicted negative ratings by peers and teachers, whichmay in turn have influenced peers’ decisions to ex-clude a child, thus compounding the level of loneli-ness experienced by the child.

Bully–victim relationships

The incidence of bullying. Any attempt to determinehow many children experience bullying depends onthe use of definitions that apply across develop-mental epochs and, if possible, across cultural set-tings. What exactly does it mean to bully others? Alarge-scale international study of school bullyingattempted to address the problem of differing cul-tural definitions of the term bullying, whilst alsofocusing on age and gender differences (Smith,Cowie, Olafsson, & Liefooghe, 2002). Cartoons wereused to investigate the meanings given to variousterms, covering a range of social situations betweenpeers. Whilst 8-year-olds were able to discriminatebetween non-aggressive and aggressive social situa-tions, 14-year-olds were able to discriminateamongst fighting, physical bullying, verbal bullyingand social exclusion.

Overall, there were no gender differences in theunderstanding of social situations, confirming earl-ier research on the definition of bullying (Smith &Levan, 1995). The study demonstrated that there arenuances within each language with regard to theterms for bullying that children use in each culture,which must be carefully considered for future inter-national comparisons.

Increasing attention has been paid to the factthat some victims of aggression are themselvesaggressive, and so distinctions are made between‘passive victims’ and ‘bully/victims’. In assessmentsof primary-school-aged boys (Schwartz, Dodge, Pet-tit, & Bates, 1997), some boys were classified asaggressive victims (8%), others as passive victims(11%), and still others as non-victimised aggressors(17%). Thus over a third of the boys were activelyinvolved in bully–victim relationships.

Bully–victim relationships may be transient. Forexample, in a small-scale observational study ofclassroom bullying, 60 bullying episodes wererecorded in 28 hours of observations; on average,bullying occurred twice an hour, and was transient(Atlas & Pepler, 1998). Although only a minority ofchildren are classified as bullies, other peers doparticipate in interactions that involve bullying. In85% of the bullying episodes, other peers wereinvolved in some capacity, from active participationto passive onlooker.

Developmental trends. Bully–victim relationshipsbegin to emerge in the preschool years. For example,a social network analysis of 4- to 5-year-old children

showed that a ‘combined central victim/aggressormodel’ best described the pattern for both physicaland general aggression (Vermande, van den Oord,Goudena, & Rispens, 2000). Aggressive relation-ships were transient, with only a minority of childrenvictims and aggressors. Most children involved inaggressive relationships were dissimilar with respectto the level of aggression displayed, with one childthe aggressor and the other the victim; however,some reciprocal aggressive relationships did occur.

Does the rate of bullying change with age? An agedecline in the frequency of children being bullied atschool has been reported in many empirical studies(for a recent review see Smith, Madsen, & Moody,1999). However, many studies rely heavily on self-report data, with only a small number consideringother methods. Experiences of victimisation aresubjective. Indeed, Perry, Kusel, and Perry (1988)and later Graham and Juvonen (1998)4 confirmedthat there is a subset of children who report beingvictims of bullying but are not recognised as victimsby their peers. Another subset does not report anyvictimisation, but peers recognised these children asvictims of bullying. Similarly, Ladd and Ladd (2001)state that as many as 77% of school children reportthat they have experienced victimisation, yet only14% report that this affected them negatively.Clearly, the subjective nature of children’s percep-tion of bullying and victimisation and an over-reli-ance on self-report methods may have caused somedisparity in the literature.

Several possible explanations for a decline in bul-lying have been offered (Smith et al., 1999). Youngerchildren may report higher rates of victimisationbecause they have not yet acquired social andassertiveness skills to combat bullying incidents anddiscourage further incidents. Furthermore, youngerchildren are in a more vulnerable position; there aremany older children at school who are bigger and arein a position to bully them. The higher rates ofreported bullying amongst the younger children canalso be explained by differences in understanding ofthe definition of bullying (Smith et al., 1999, 2002).These age differences in children’s understanding ofbullying make it difficult to draw firm conclusionsabout developmental trends.

If there is a true decline in victimisation as chil-dren grow older, this should also be confirmed bypeers and teachers. However, in a study of 48 schoolclasses, a decrease in self-reported victimisation wasfound between 9 and12 years of age, but this was notcorroborated by peers or teachers (Salmivalli, 2002)6 .The children were classified into one of four groups:self-identified victims, peer-identified victims, self-and peer-identified victims and non-victims. Therewas a significant decrease in victimisation as afunction of age for self-identified victims, whilst thenumber of children whose reports of being bulliedwere confirmed by their peers remained steadythrough the age range.


Longitudinal observations of a US sample showeda curvilinear trend, with bullying and aggressionincreasing with the transition to middle school anddeclining thereafter (Pellegrini & Long, 2002). Over-all, boys were victimised more than girls, with boystargeting boys more than they targeted girls, andgirls targeting girls more than they targeted boys.Gender differences in patterns of bullying and victi-misation have been noted in many different samples(e.g., Bjorkqvist, Lagerspetz, & Kaukiainen, 1992;Bukowski, Sippola, & Newcomb, 2000; Olweus &Endresen, 19983 ).

Family factors. Early family experiences predictwhich boys later emerge as bullies, victims and vic-timised aggressors (Schwartz et al., 1997). In a lon-gitudinal study, mothers of 5-year-old boys wereinterviewed and the home environment assessed.Five years later, those children who had experiencedmore punitive, hostile and abusive family treatmentduring their preschool years were both aggressiveand victimised by others. In contrast, children whowere aggressive but not victimised by their peers hadearlier received greater exposure to adult aggressionand conflict than children who were neither aggres-sive nor victimised; however, these aggressiveyoungsters who were not victims had not experi-enced victimisation by adults. Those victims whowere not aggressive did not differ from children whowere neither victims nor bullies on any dimension ofthe home environment.

Underlying processes. We have seen that problemsin emotion regulation and social understanding af-fect children’s acceptance by peers in early child-hood; it thus seems likely that deficits in thesedomains, which may indeed derive from early familyexperiences, also contribute to a child’s emergingroles as bully, victim, or victimised aggressor. Indeed,victims of bullying sometimes show problems inregulating emotion (Mahady Wilton, & Craig, 2000).However, the relationship between social cognitiveskills and bully-victim relationships is complex.

It might be thought that aggressive children havedeficits in social understanding, and that such prob-lems in social cognition are associated with bullyingothers. However, the assumption that bullies have apoor understanding of others can be challenged(Sutton, Smith, & Swettenham, 1999). Some bulliespossess good social cognition and theory of mindskills, which equip them well to manipulate others,often inflicting suffering upon others subtly to avoiddetection. Such social acumen may characterisemany bullies, but social cognitive skills are particu-larly likely to be used by ringleader bullies and inindirect forms of bullying, which are commonamongst girls (Sutton et al., 1999).

Self-control processes and, in particular, theability to regulate one’s attention and activity alsoinfluence bullying and victimisation. Dysregulated

children may be especially vulnerable to bullyingand may also bully others. It is of particular interestto examine the relationship between bullying andADHD. In a study of 1315 middle-school pupils, pathanalysis indicated that pupils who had taken med-ication for ADHD had low levels of self-control andwere at increased risk for bullying and victimisationby bullies (Unnever & Cornell, 2003). For thesepupils, it was not low self-control that increased theprobability of being bullied, but rather other symp-toms related to ADHD, including poor social skillsand frequent, inappropriate behaviour, which peersfound obtrusive. In contrast, lower levels of self-control were associated with a higher probability ofbullying other pupils.

Physical size has been linked to individual differ-ences in aggression (Raine, Reynolds, Venables,Mednick, & Farrington, 1998), and may also affectbullying; however, this appears to depend on aninteraction with self-control. In the sample justdescribed (Unnever & Cornell, 2003), pupils with lowself-control who perceived themselves to be strongerthan their peers were most likely to bully others, but,for students with high self-control, strength was notassociated with bullying. Low self-control plusgreater height and/or weight increased the likeli-hood of bullying others. However, pupils who wereboth overweight and had ADHD were the most likelyto be bullied. Thus, physical characteristics mayinteract with psychological processes to predict thequality of peer relationships, in the context ofchildhood disorder.

Do peer relationships play a causal rolein the genesis of disorder?

As we have just seen, children with disorders mayexperience problems in relating to peers, and prob-lematic peer relations are associated with greaterlevels of emotional and behavioural problems. How-ever, can we say confidently that peer relations playa causal role in the genesis of disorder?

Emotional disorders

Evidence concerning the contribution of peer rela-tions to emotional disorders remains equivocal.There is retrospective evidence that depressedwomen recall having poor peer relationships inchildhood (Hock & Lutz, 2001). Furthermore, thereare links between being victimised by peers andsubsequent internalising problems (Goodman,Stormshak, & Dishion, 2001), and between thesubjective feeling of loneliness and depressivesymptomatology (Gazelle & Ladd, 2003). Never-theless, there is also some counterevidence. Inlongitudinal observations of the Christchurch, NewZealand cohort, peer problems in childhood were notassociated with risk for anxiety or depressive


disorders in young adults (Woodward & Fergusson,1999). Furthermore, although prosocial children areusually seen favourably by their peers (e.g., Denhamet al., 1990; Vitaro et al., 1990), some prosocialchildren show severe levels of emotional problems(Hay & Pawlby, 2003). The question of links betweenpeer problems and risk for emotional disorderdeserves more intensive scrutiny.

Conduct disorder

The causal role of peer relations has primarily beeninvestigated with respect to the development ofconduct disorder. We explore this issue in detail, asthis literature illustrates some of the general com-plexities encountered when attempting to establishthe precise contribution of peers to psychologicaldevelopment. Investigators have been keen tounderstand whether problems with peers are simplymanifestations of underlying conduct disorder, oractual contributors to the development of conductdisorder in their own right. The two phenomena thathave received most attention are peer rejection inearly childhood and associations with deviant peersin later childhood and adolescence. A chain of eventshas been proposed whereby peer rejection in child-hood is followed by associations with deviant peersin later childhood and adolescence, which in turnincreases the risk of conduct problems and delin-quency (e.g., Cairns & Cairns, 1994; Quinton, Pick-les, Maughan, & Rutter, 1993). However, it could beargued that, because aggressive children are espe-cially likely to be rejected by their peers, anyassociation between peer rejection and later conductdisorder could be explained by intra-individualcontinuities in externalising problems throughoutchildhood. It thus becomes important to determinewhether there is evidence for a pathway from earlyaggression to peer rejection to affiliation with deviantpeers that aids prediction of risk for conduct disor-der, beyond simple autocorrelations between exter-nalising problems at early and later ages (Figure 3).To what extent does the existing evidence supportthe different links in this chain of events?

Does aggression lead to peer rejection? As we haveseen, even in early childhood, when the use of

aggression is not uncommon, aggressive children,and especially thosewho lackprosocial strategies, aredisliked by their peers (e.g., Denham et al., 1990;Schwartz, McFayden-Ketchum, Dodge, Pettit, &Bates, 1999). The relationship between externalisingproblems and peer relationships is seen for girls aswell as boys (e.g., Fergusson, Swain-Campbell, &Horwood, 2002), and for verbal as well as physicalaggression (Khatri & Kupersmidt, 2003). The phe-nomenon is also seen across cultures, having beendemonstrated in a number of longitudinal studies inWestern industrialised nations (e.g., Fergusson et al.,2002; Schwartz et al., 1999) and ina study of childrenin semi-rural India, in which both aggressors andvictims were rejected by their peers (Khatri & Ku-persmidt, 2003). Children who have been maltreatedare at special risk to show the levels of aggression thatresult in peer rejection (Bolger & Patterson, 2001).

Children do not have to be aggressive to berejected; social withdrawal and victimisation alsolead to peer rejection (Bierman, Smoot, & Aumiller,1993; Bolger & Patterson, 2001; Khatri & Ku-persmidt, 2003). However, within the group of chil-dren who are rejected by their peers, it appears to bethe aggressive children who show a broader array ofconduct problems, which in turn lead to continuedrejection (Bierman et al., 1993). As we have seen, thepresence of social cognitive problems increases thelikelihood that aggressive children are rejected(Schonberg et al., 2003).

Whether or not aggressive children are rejectedalso appears to depend on the base rates of aggres-sion within their peer group. Peers may evaluateaggressive behaviour in light of the social norms ofthe peer group; if the behaviour is deemed as fairlynormal for the group then peer rejection will notoccur (Stormshak, Bierman, Bruschi, Dodge, & Coie,1999). In support of this claim, in a sample of 1895American 6-year-olds, levels of classroom aggressionwere inversely correlated with the extent of negativeeffects of aggression on peer acceptance (Stormshaket al., 1999). This implies that aggressive childrenmay find friends in contexts where there is a higherthan usual amount of aggression in the peer group,which in turn suggests that children who are rejec-ted by their non-aggressive peers may aggregate intonew, increasingly aggressive peer groups. At theindividual level, this process may increase a child’slevel of aggression and other conduct problems.

Does peer rejection lead to further escalation ofaggression? There is accumulating evidence for alink between rejection by the peer group and conse-quent antisocial behaviour (e.g., Miller-Johnsonet al., 2002; Woodward & Fergusson, 1999). Forexample, in a sample of 566 American children whowere assessed first at the age of 5 and then annuallyuntil they were 10, peer rejection significantly pre-dicted externalising behaviours at 10 (Deater-Deck-ard, Dodge, Bates, & Pettit, 1998).

Early Aggression Conduct

Disorder and

Crime

Peer

Rejection

Friendship With Aggressive Peers

Figure 3 Do processes of peer rejection and theestablishment of friendships with deviant peers influ-ence a child’s risk for conduct disorder, beyond pre-diction on the basis of early aggression?


Being rejected by the peer group might directlyincrease the likelihood of becoming involved inantisocial activities, insofar as rejection leads tohostility and aggression by the child (Dodge et al.,2003). Rejection by the peer group might be especi-ally detrimental to children who are already showingsigns of maladaptive behaviour, acting like a stressorfor the already vulnerable child. However, in a sam-ple of 259 American children, early rejection by thepeer group was associated with antisocial behaviourfour years later, regardless of the extent of antisocialbehaviour evident at the time of peer rejection (Dodgeet al., 2003); the association was especially strongfor reactive aggression. Similarly, children who hadbeen rejected as 6-year-olds in first grade were likelyto manifest conduct problems four years later, evenwhen their levels of aggression and symptoms ofADHD in first grade were taken into account (Miller-Johnson et al., 2002). In contrast, in a secondsample of 585 children (Dodge et al., 2003), thedetrimental influence of peer rejection at age 5 wasonly apparent in those children who had alreadyshown aggressive tendencies. Girls and reactive-aggressive children were most vulnerable to thedetrimental consequences of peer rejection.

The effect of peer rejection on children’s later ten-dencies to be aggressive may partially be due to thefact that some children who are not accepted by theirpeers become actively victimised. For example, inlongitudinal observations, children with early beha-viour problems were likely to be rejected by the peergroup, which in turn led to victimisation by the peergroup at a later time, and thus peer rejection didmediate the association between behaviour prob-lems at age 5–6 years and victimization by the peergroup at ages 8–9 years (Schwartz et al., 1999).Some victims become aggressive, especially whenthey have also been victimised by the adults in theirlives (Schwartz et al., 1997).

The effect of victimisation on aggression maydepend on the victim’s sex (Egan, Monson, & Perry,1998). In a sample of American children between 8and 12 years of age, boys who were victims of peeraggression did not show an increase in aggressivebehaviour over time, despite having social cognitionssupporting the use of aggression. In fact, those boyswho were victims of peer aggression tended toevaluate aggressive responses less positively overtime. However, girls who were victims of peeraggression became more aggressive over time.

Deviant friendships and peer groups. As we haveseen, rejection by non-aggressive peers in earlychildhood may play a role in the establishment offriendships between aggressive children in laterchildhood. Peer rejection is predictive of associationswith deviant peers by early adolescence (Dishion,Patterson, Stoolmiller, & Skinner, 1991). It is there-fore important to examine the quality of peer rela-tions in groups of aggressive children.

In general, children who are identified as beingrejected or aggressive-rejected have fewer friendsthan their non-rejected non-aggressive counterparts(Deptula, 2003). However, by early adolescence,aggressive children may be as likely as non-aggres-sive children are to belong to a peer group (Hodgetts,2003). Groups of aggressive children may experiencea level of conflict that exceeds the norm (e.g., Espe-lage, Holt, & Henkel, 2003). For example, friendshipsbetween aggressive children have been observed tobe more fraught than those of non-aggressive chil-dren (Capaldi, Dishion, Stoolmiller, & Yoerger,2001). These authors suggested that the increasedlevels of conflict in these friendships lead to furtheraggressive episodes, which provides a furtherexplanation of how aggressive children’s friendshipsserve to facilitate aggression.

Nevertheless, peer groups comprised of aggressiveyouth establish group norms and provide socialsupport in the same manner as other peer groups.For example, in a sample of 684 children aged 11 to13 years, from diverse ethnic and economic back-grounds, aggressive/delinquent children’s friend-ships were characterised by as much instrumentaland emotional support as those of non-aggressive/non-delinquent children were (Hodgetts, 2003).Despite reporting higher levels of conflict withinfriendships, aggressive/delinquent children statedthat they were more satisfied with the friendshipsand received a greater level of support from theirfriendships than non-aggressive/non-delinquentchildren did. This suggests that friendships amongstaggressive youth constitute an important, emotion-ally charged arena for the further socialisation ofaggression and other forms of criminal offending.

In particular, the establishment of social normswithin aggressive peer groups may spur on furtheraggression. Group members are likely to influenceeach other’s views about the utility and acceptabilityof aggressive actions. For example, in a sample of206 Latino children aged 11 to 12 years, from mainlylow-income areas, aggressive friends provided asocial context within which aggressive thought andactions were endorsed, leading to increases inaggressive behaviour over time (Isaacs & Hodges,2003). Thus, as a function of social cognitive as wellas learning processes, aggressive children tend tomake each other more aggressive over time.

The negative effects aggressive youngsters have oneach other have been illustrated most dramaticallyby studies of interventions that unwittingly increasedrather than decreased levels of conduct problems intheir participants. A natural tendency on the part ofresearchers and policy-makers is to try to intervenewith aggressive youth efficiently by aggregatingtroubled individuals into groups, who then spendtime together in after-school programmes, summercamps and the like. This exposure to other aggressiveindividuals may be a more powerful influence thanthe planned intervention designed to deter aggres-


sion. For example, in the classic Cambridge–Somer-ville study, the long-term follow-up showed that theexperimental group, who received an interventiondesigned to reduce delinquency, fared more poorlythan the comparison subjects (McCord, 1992).Findings in recent studies (Dishion, Bullock, &Granic, 2002; Hektner, August, & Realmuto, 2000)similarly attest to the power of iatrogenic effects onconduct disorder. In contrast, family-based inter-ventions that reduce the time spent with deviantpeers may reduce risk (Dishion et al., 2002). Forexample, a prevention programme that did not isol-ate and aggregate highly aggressive children, but,rather, occurred relatively early in development andfocused on small groups of both aggressive and pro-social children, in tandem with a focus on parentingskills, deflected at-risk children from increasinglyviolent trajectories (Lacourse et al., 2002).

Does peer rejection lead to deviant peer groupsand thereby to conduct disorder? Although there iscertainly evidence for a link between peer rejectionand increasingly aggressive behaviour, and for theemergence of friendships amongst aggressive youth,which further encourages conduct problems, is thereclear evidence for the full causal pathway from earlyaggression to later disorder, mediated by peer rejec-tion and associations with deviant peers (Figure 3)?In addressing this question, it is important to takeinto account the child’s earlier level of aggressionand other behavioural problems. The evidence ismixed, and may depend on the population beingstudied. In a longitudinal follow-up of Britishyoungsters who spent time in care, peer rejectionwas associated with associations with deviant peers,and indeed assortative mating with deviant partners(Quinton et al., 1993). However, in the follow-ups ofthe Christchurch, New Zealand cohort, peer rejectionin middle childhood did not lead to associations withdeviant peers in adolescence, once the level of con-duct problems at age 9 were taken into account(Fergusson, Woodward, & Horwood, 1999). Fur-thermore, in that sample, the association betweenpeer rejection and problems in adolescence (criminaloffending and substance use) was largely explainedby family risk factors and other characteristics ofthe child, such as cognitive ability (Woodward &Fergusson, 1999). However, peer rejection continuedto predict educational underachievement and un-employment, even when other risk factors weretaken into account (Woodward & Fergusson, 2000).

In the Christchurch sample, there was indeed aneffect of deviant peers on criminal activity and sub-stance use, which was evident for girls as well asboys, and was especially marked if the associationwith deviant peers occurred around 14 to 15 years ofage, rather than in young adulthood (Fergussonet al., 2002). Thus, the pathway to conduct disorderand substance use observed in that sample seemedto be one of children with initial levels of conduct

problems beginning to associate with deviant peersin mid-adolescence, which increased risk. This is, ofcourse, exactly the sampling frame and procedurethat would be used in group-based interventionstudies designed to reduce aggression (Dishionet al., 2002).

Peer relations as a protective factorin development

Having enquired whether peers play a role in thegenesis of disorder, it is important to ask the com-plementary question of whether peer relationshipscan serve as a protective factor in psychologicaldevelopment. In very early childhood, peers fosterexploration of the physical and social world (e.g.,Gunnar, Senior, & Hartup, 1984; Ispa, 1981) andfacilitate cognitive development (e.g., Light & Gla-chan, 1985). Under conditions of extreme depriva-tion, in the absence of caring adults, peers provide asource of emotional security (Freud & Dann, 1949).Nonetheless, with respect to the development ofpsychopathology, it is clear that we know rathermore about the ill effects of peer rejection than thepositive effects of children’s friendships and mem-bership in well-functioning peer groups.

One exception to this general trend is a study ofthe protective effects of acceptance by the peer groupand the extent of children’s friendships, which con-trolled for the children’s level of social skills, whichmight influence peer acceptance whilst having adirect effect on later social adjustment (Criss, Pettit,Bates, Dodge, & Lapp, 2002). At the time of the firstassessment the sample consisted of 585 Americanchildren aged 3 to 6 years, from middle-class back-grounds. Social information processing skills wereassessed during home interviews when the childrenwere 3 years old. Measures of peer acceptance,friendships and mean levels of peer aggressivenesswere obtained via sociometric interview at differentpoints in the study, with teachers’ reports of child’sexternalising behaviours obtained at age 7 years.Peer acceptance and friendships served as protectivefactors against the development of externalisingproblems, with peer acceptance serving a somewhatenhanced protective function. Importantly, socialinformation processing skills were not found toaccount for the protective role of peer acceptanceand friendships, and the protective role of peeracceptance and friendship was not influenced bywhether the friendship was conflictual or not.Rather, simply having a friend at an early ageseemed to protect a child against the risk of furtheraggression and rejection.

In later childhood and adolescence, the protectivefeatures of peer relations must be studied withattention to the complex social networks that char-acterise the world of peers. Recent theoreticaladvances in the study of social networks provide new


means of understanding the structure of peer rela-tions and the place of an individual child vis-a-visvarious peer groups (e.g., Cairns, Cairns, Necker-man, Gest, & Gariepy, 1988; Moody, 2001). It isimportant to study affiliations amongst well-func-tioning youth as well as the networks of deviantpeers. For example, during the secondary schoolyears teenagers form friendships and peer groupsbased on academic achievement and aspirations, aswell as common interests, sport and charitableactivities. The relative levels of psychological prob-lems within one’s peer group may serve as riskfactors for an individual’s mental health and socio-educational adjustment. For example, associationswith deviant peers may deprive youngsters of age-normative experiences that foster educationalachievement and entry into the world of work(Woodward & Fergusson, 2000). The social‘homophily’ (Espelage et al., 2003) of childhood andadolescent peer groups deserves further investiga-tion.

Conclusions

To summarise, in the foregoing review of the litera-ture on peer relations, we have outlined a develop-mental pathway from the initial peer encounters ininfancy to complex peer networks in adolescence andyoung adulthood (with some underlying processessummarised in Figures 1–3). We have argued thatearly competence with peers depends on key skillsand self-regulatory abilities (Figure 1); individualdifferences in the level of aggression within peerrelationships, and in shyness, are already present inthe toddler years. Both of these tendencies, whichare affected by executive function, social under-standing and emotional regulation, play an import-ant role in the consolidation of social skills inchildren’s first peer groups, and influence the extentto which individual children are accepted or rejectedby peers (Figure 2). Peer acceptance or rejection inturn influences children’s risk for victimisation,associations with deviant peers and disorder,although the causal pathways are not completelyknown (Figure 3).

It is evident that an individual child’s ability toengage positively with his or her peers is affected bythat child’s individual skills and temperament, aswell as by factors in the family and greater socialenvironment at all points in development. When allof these individual and familial risk factors are takeninto account, the contribution of peers to the child’srisk for disorder may be attenuated (e.g., Woodward& Fergusson, 1999); put another way, family inter-ventions may militate against the negative effects ofdeviant peers (Dishion et al., 2002). Such findings,however, do not negate the importance of peers forpsychological development. Indeed, our search for adirect, causal influence of peer relations on disorder

may have missed the point so earnestly debatedby the 4-year-olds quoted at the beginning of thisarticle. To put it bluntly, peer relations may be boththe chicken and the egg, and it behoves us to exploitmethods for the analysis of longitudinal data toexplore the reciprocal relation between peers anddisorder more fully (see Kenny, 1979).

Whether or not problematic peer relations play adirect causal role in the development of disorder –and the iatrogenic effects of experimental interven-tions suggest that they sometimes do – the world ofpeers constitutes a challenging and sometimesunforgiving environment in which disorder mani-fests itself. Young toddlers with developmental dis-orders may have particular difficulties adjusting tothe demands of peer groups, and it may be in thepeer setting that the severity of their problems is firstrecognised. Children with internalising or external-ising problems may suffer academically as well associally, to the extent that difficulties with peersinterfere with adjustment to formal educational set-tings. Ultimately, problematic peer relationships anda growing association with deviant peers holdimportant implications for a choice of romanticpartner (Quinton et al., 1993); intimate relationshipswith deviant peers may encourage early sexualrelationships and parenthood, thereby conveyingadded risk for the next generation.

Nonetheless, despite the emerging evidence for thenegative effects of peer relations on children’sadjustment and risk for disorder, it is also salutaryto remember that, in the analyses of representativecommunity samples reported here, most childrenhave friends and ‘good enough’ relationships withtheir peers. Although the world of peers can indeedbe cruel at times (Deater-Deckard, 2001), it can alsobe enormously supportive, constructive and, most ofall, fun. Perhaps it is the fun we have with our peersthat conveys both the greatest risk if – the funinvolves antisocial flirtation with physical and socialdangers – and the greatest positive contributionpeers make to psychological development.

We suggest that it is time for psychologists andpsychiatrists to turn their attention once again to theserious study of fun. It is worth noting that, in theindex of the most recent Handbook of Child Psy-chology (Damon, 1998), in contrast to editions fromearlier decades, there is a full chapter on peer rela-tions, with many citations pertaining to problematicpeer relationships (Rubin et al., 1998). However,throughout all four volumes, only a handful of ref-erences are made to the main thing that children invirtually all cultures do with their peers, namely,play. Play with peers was once a major topic indevelopmental psychology, and deserves to bestudied anew. It is our hope that future studies ofpeer relationships will focus more extensively on thepositive and protective features of children’s rela-tionships with their friends and acquaintances,whilst not forgetting that, for some children,


difficulty getting along with peers is a harbinger ofvery serious psychological problems.

Correspondence to

Dale Hay, School of Psychology, University of Cardiff,Tower Building, Park Place, PO Box 901, CardiffCF10 3YG, Wales; Email: [email protected]

ReferencesAdamson, L.B., & Bakeman, R. (1985). Affect andattention: Infants observed with mothers and peers.Child Development, 56, 582–593.

Asendorpf, J.B. (1990). Beyond social withdrawal:Shyness, unsociability, and peer avoidance. HumanDevelopment, 33, 250–259.

Asendorpf, J.B. (1992). A Brunswikean approach totrait continuity: Application to shyness. Journal ofPersonality, 60, 53–77.

Asher, S.R., Hymel, S., & Renshaw, P.D. (1984). Lone-liness in children. Child Development, 55, 1456–1464.

Asher, S.R., Parkhurst, J.T., Hymel, S., & Williams,G.A. (1990). Peer rejection and loneliness in child-hood. In S.R. Asher & J.D. Coie (Eds.), Peer rejectionin childhood (pp. 253–273). New York: CambridgeUniversity Press.

Asher, S.R., Singleton, L.C., Tinsley, B.R., & Hymel, S.(1979). A reliable sociometric measure for preschoolchildren. Developmental Psychology, 15, 443–444.

Atlas, R.S., & Pepler, D.J. (1998). Observations ofbullying in the classroom. Journal of EducationalResearch, 92, 86–99.

Babcock, F., Hartle, L., & Lamme, L.L. (1995). Prosocialbehaviors of five-year-old children in sixteen learn-ing/activity centers. Journal of Research in ChildhoodEducation, 9, 113–127.

Badenes, L.V., Estevan, R.A.C., & Bacete, F.J.G. (2000).Theory of mind and peer rejection at school. SocialDevelopment, 9, 271–283.

Bakeman, R., & Adamson, L.B. (1984). Coordinatingattention to people and objects in mother–infant andpeer–infant interaction. Child Development, 55,1278–1289.

Bakeman, R., & Brownlee, J.R. (1982). Social rulesgoverning object conflicts in toddlers and preschoo-lers. In K.H. Rubin & H.S. Ross (Eds.), Peer relationsand social skills in childhood (pp. 99–110). New York:Springer-Verlag.

Balaraman, G.R. (2003). Children’s self regulation andpeer interaction at 36 and 54 months: Concurrent andlongitudinal relations. Paper presented at the BiennialMeeting of the Society for Research in Child Develop-ment, Tampa, Florida.

Belanger, N.D., & Desrochers, S. (2001). Can 6-month-old infants process causality in different types ofcausal events? British Journal of DevelopmentalPsychology, 19, 11–21.

Benenson, J.F. (1993). Greater preference amongfemales than males for dyadic interaction in earlychildhood. Child Development, 64, 544–555.

Berlin, L.J., Cassidy, J., & Belsky, J. (1995). Lonelinessin young children and infant–mother attachment:A longitudinal study. Merrill Palmer Quarterly, 41,91–103.

Bierman, K.L., Smoot, D.L., & Aumiller, K. (1993).Characteristics of aggressive-rejected, aggressive(nonrejected), and rejected (nonaggressive). boys.Child Development, 64, 139–151.

Bjorkqvist, K., Lagerspetz, K.M.J., & Kaukiainen, A.(1992). Do girls manipulate and boys fight? Aggres-sive Behavior, 18, 117–127.

Black, B. (1989). Interactive pretense: Social and sym-bolic skills in preschool play groups. Merrill PalmerQuarterly, 35, 379–397.

Black, B., & Hazen N. (1990). Social status and patternsof communication in acquainted and unacquaintedpreschool children. Developmental Psychology, 26,379–387.

Boivin, M., Hymel, S., & Burkowski, W.M. (1995). Theroles of social withdrawal, peer rejection and victimi-sation by peers in predicting loneliness and de-pressed mood in childhood. Development andPsychopathology, 7, 765–785.

Bolger, K.E., & Patterson, C.J. (2001). Developmentalpathways from child maltreatment to peer rejection.Child Development, 72, 549–568.

Bost, K.K. (1995). Mother and child reports of preschoolchildren’s social support networks: Network corre-lates of peer acceptance. Social Development, 4, 149–164.

Broidy, L., Nagin, D.S., Tremblay, R.E., Bates, J.,Brame, B., Dodge, K.A., Fergusson, D., Horwood,J.L., Loeber, R., Laird, R., Lynam, D.R., Moffitt, T.E.,& Pettit, G.S. (2003). Developmental trajectories ofchildhood disruptive behaviors and adolescent delin-quency: A six-site, cross-national study. Developmen-tal Psychology, 39, 222–245.

Brooks, J., & Lewis, M. (1976). Infants’ responses tostrangers: Midget, adult, and child. Child Develop-ment, 47, 323–332.

Bukowski, W.M., Sippola, L.A., & Newcomb, A.F.(2000). Variations in patterns of attraction to sameand other sex peers during early adolescence. Devel-opmental Psychology, 36, 147–154.

Burford, H.C., Foley, L.A., Rollins, P.G., & Rosario, K.S.(1996). Gender differences in preschoolers’ sharingbehavior. Journal of Social Behavior and Personality,11, 17–25.

Buss, K.A., & Goldsmith, H.H. (1998). Fear and angerregulation in infancy: Effects on the temporal dy-namics of affective expression. Child Development,69, 359–374.

Butterworth, G. (2001). Joint visual attention ininfancy. In G. Bremner & A. Fogel (Eds.), Blackwellhandbook of infant development. Handbooks of devel-opmental psychology (pp. 213–240). Malden, MA:Blackwell Publishers.

Cairns, R., & Cairns, B.D. (1994). Lifelines and risks:Pathways of youth in our time. Cambridge: Cam-bridge University Press.

Cairns, R.B., Cairns, B.D., Neckerman, H.J., Gest,S.D., & Gariepy, J.-L. (1988). Social networks andaggressive behavior: peer support or peer rejection?Developmental Psychology, 24, 815–823.


Camaioni, L., Baumgartner, E., & Perucchini, P. (1991).Content and structure in toddlers’ social competencewith peers from 12 to 36 months of age. Early ChildDevelopment and Care, 67, 17–27.

Capaldi, D.M., Dishion, T.J., Stoolmiller, M., & Yoerger,K. (2001). Aggression toward female partners by at-risk young men: The contribution of male adolescentfriendships. Developmental Psychology, 37, 61–73.

Caplan, M.Z. (1986). The role of aggression in preschoo-lers’ conflicts with their peers. Unpublished PhDdissertation, State University of New York at StonyBrook.

Caplan, M.Z., & Hay, D.F. (1989). Preschooler’s re-sponses to peers’ distress and beliefs about bystanderintervention. Journal of Child Psychology and Psychi-atry, 30, 231–242.

Caplan, M., Vespo, J.E., Pedersen, J., & Hay, D.F.(1991). Conflict and its resolution in small groups ofone- and two-year-olds. Child Development, 62,1513–1524.

Carlson, S.M., & Moses, L.J. (2001). Individual differ-ences in inhibitory control and theory of mind. ChildDevelopment, 72, 1032–1053.

Cassidy, J., & Asher, S.R. (1992). Loneliness and peerrelations in young children. Child Development, 63,350–365.

Charman, T., Baron-Cohen, S., Swettenham, J., Baird,G., Cox, A., & Drew, A. (2001). Testing joint attention,imitation, and play as infancy precursors to languageand theory of mind. Cognitive Development, 15, 481–498.

Charman, T., Swettenham, J., Baron-Cohen, S., Cox,A., Baird, G., & Drew, A. (1997). Infants with autism:An investigation of empathy, pretend play, jointattention, and imitation. Developmental Psychology,33, 781–789.

Chen, D.W., Fein, G., & Tam, H.P. (2001). Peer conflictsof preschool children: Issues, resolution, incidence,and age-related patterns. Early Education and Devel-opment, 12, 523–544.

Cohen, L.B., Rundell, L.J., Spellman, B.A., & Cashon,C.H. (1999). Infants’ perception of causal chains.Psychological Science, 10, 412–418.

Cooper, P.J., & Eke, M. (1999). Childhood shyness andmaternal social phobia: A community study. BritishJournal of Psychiatry, 174, 439–443.

Coplan, R.J., Coleman, B., & Rubin, K.H. (1998).Shyness and little boy blue: Iris pigmentation,gender, and social wariness and preschoolers. Devel-opmental Psychobiology, 32, 37–44.

Cote, S., Tremblay, R.E., Nagin, D., Zoccolillo, M., &Vitaro, F. (2002). The development of impulsivity,fearfulness, and helpfulness during childhood: Pat-terns of consistency and change in the trajectories ofboys and girls. Journal of Child Psychology andPsychiatry, 43, 609–618.

Crick, N.R., Casas, J.F., & Mosher, M. (1997). Rela-tional and overt aggression in preschool. ChildDevelopment, 33, 579–588.

Criss, M.M., Pettit, G.S., Bates, J.E., Dodge, K.A., &Lapp, A.L. (2002). Family adversity, positive peerrelationships, and children’s externalising behavior:A longitudinal perspective on risk and resilience.Child Development, 73, 1220–1237.

Cummings, E.M., Iannotti, R.J., & Zahn-Waxler, C.(1989). Aggression between peers in early childhood:Individual continuity and change. Child Development,60, 887–895.

Damon, W. (Ed.). (1998). Handbook of child psycho-logy. Chichester: Wiley.

Davis, S., Howell, P., & Cooke, F. (2002). Sociodynamicrelationships between children who stutter and theirnon-stuttering classmates. Journal of Child Psycho-logy and Psychiatry, 43, 939–947.

Deater-Deckard, K. (2001). Annotation: Recent re-search examining the role of peer relations in thedevelopment of psychopathology. Journal of ChildPsychology and Psychiatry, 42, 565–579.

Deater-Deckard, K., Dodge, K.A., Bates, J.E., & Pettit,G.S. (1998). Multiple risk factors in the developmentof externalizing behavior problems: Group and in-dividual differences. Development and Psychopatho-logy, 10, 469–493.

Denham, S.A., McKinley, M., Couchoud, E.A., & Holt,R. (1990). Emotional and behavioral predictors ofpreschool peer ratings. Child Development, 61, 1145–1152.

Deptula, D. (2003). Social behaviour and social stand-ing: Comparing the friendships of aggressive and/orrejected children. Poster presented at the BiennialMeeting of the Society for Research in Child Develop-ment, Tampa, Florida.

Desrochers, S., Ricard, M., Decarie T.G., & Allard, L.(1994). Developmental synchrony between socialreferencing and Piagetian sensorimotor causality.Infant Behavior and Development, 17, 303–309.

Dettling, A.C., Gunnar, M.R., & Donzella, B. (1999).Cortisol levels of young children in full-day childcarecentres: Relations with age and temperament. Psy-choneuroendocrinology, 24, 519–536.

Diener, M.L., Mangelsdorf, S.C., McHale, J.L., &Frosch, C.A. (2002). Infants’ behavioral strategiesfor emotion regulation with fathers and mothers:Associations with emotional expression and attach-ment quality. Infancy, 3, 153–174.

Dionne, G., Tremblay, R.E., Boivin, M., Laplante, D., &Perusse, D. (2003). Physical aggression and expres-sive vocabulary in 19-month-old twins. Developmen-tal Psychology, 39, 261–273.

Dishion, T.J., Patterson, G.R., Stoolmiller, M., &Skinner, M.L. (1991). Family, school, and behavioralantecedents to early adolescent involvement withantisocial peers. Developmental Psychology, 27,172–180.

Dishion, T.J., Bullock, B.M., & Granic, I. (2002).Pragmatism in modeling peer influence: Dynamics,outcomes, and change processes. Development andPsychopathology, 14, 969–981.

Dodge, K.A., Lansford, J.E., Burks, V.S., Bates, J.E.,Pettit, G.S., Fontaine, R., & Price, J.M. (2003). Peerrejection and social information-processing factors inthe development of aggressive behavior problems inchildren. Child Development, 74, 374–393.

Doyle, A.B., Ceschin, F., Tessier, O., & Doehring, P.(1991). The relation of age and social class factors inchildren’s social pretend play to cognitive and sym-bolic ability. International Journal of BehaviouralDevelopment, 14, 395–410.


Dunn, J., Cutting, A.L., & Fisher, N. (2002). Old friends,new friends: Predictors of children’s perspective ontheir friends at school. Child Development, 73, 621–635.

Easterbrooks, M.A., & Lamb, M.E. (1979). The relation-ship between quality of infant–mother attachmentand infant competence in initial encounters withpeers. Child Development, 73, 621–635.

Eckerman, C.O., & Whitehead, H. (1999). How toddlerpeers generate coordinated action: A cross-culturalexploration. Early Education and Development, 10,226–241.

Eckerman, C.O., Davis, C.C., & Didow, S.M. (1989).Toddlers’ emerging ways of achieving social coordina-tion with a peer. Child Development, 60, 440–453.

Egan, S.K., Monson, T.C., & Perry, D.G. (1998). Social-cognitive influences on change in aggression overtime. Developmental Psychology, 34, 996–1006.

Eisenberg, N., & Fabes, R.A. (1998). Prosocial develop-ment. In W. Damon (Ed.), Handbook of child psycho-logy (vol. 3, N. Eisenberg, vol. ed., pp. 701–778).Chichester: Wiley.

Eisenberg, N., Fabes, R.A., Nyman, M., Bernzweig, J., &Pinuelas, (1994). The relations of emotionality andregulation to children’s anger-related reactions. ChildDevelopment, 65, 109–128.

Eisenberg, N., Fabes, R.A., Shepard, S.A., Murphy,B.C., Guthrie, I.K., Jones, S., Friedman, J., Poulin,R., & Maszk, P. (1997). Contemporaneous and long-itudinal prediction of children’s social functioningfrom regulation and emotionality. Child Development,68, 642–664.

Eisenberg, N., Guthrie, I.K., Murphy, B., Shepard, S.,Cumberland, A., & Carlo, G. (1999). Consistency anddevelopment of prosocial dispositions: A longitudinalstudy. Child Development, 70, 1360–1372.

Espelage, D.L., Holt, M.K., & Henkel, R.R. (2003).Examination of peer-group-contextual effects onaggression during early adolescence. Child Develop-ment, 74, 205–220.

Fabes, R.A., Eisenberg, N., Jones, S., Smith, M.,Guthrie, I.K., Poulin, R., Shepard, S., & Friedman,J. (1999). Regulation, emotionality, and preschoolers’socially competent peer interactions. Child Develop-ment, 70, 432–442.

Fagot, B.I., & Hagan, R. (1985). Aggression in toddlers:Responses to the assertive acts of boys and girls. SexRoles, 12, 341–351.

Fagot B., & Leve, L.D. (1998). Teacher ratings ofexternalising behaviour at school entry for boys andgirls: Similar early predictors and different correlates.Journal of Child Psychology and Psychiatry, 39, 555–546.

Farver, J.A.M., Kim, Y.K., & Lee, Y. (1995). Culturaldifferences in Korean-American and Anglo-Americanpreschoolers’ social interaction and play behaviors.Child Development, 66, 1088–1099.

Farver, J.A.M., Kim, Y.K., & Lee-Shim, Y. (2000). Withincultural differences: Examining individual differencesin Korean-American and European-American pre-schoolers’ social pretend play. Journal of Cross-Cul-tural Psychology, 31, 583–602.

Ferguson, T.J., & Rule, B. (1988). Children’s attribu-tions of retaliatory aggression. Child Development, 59,961–968.

Fergusson, D.M., Woodward, L.J., & Horwood, L.J.(1999). Childhood peer relationship problems andyoung people’s involvement with deviant peers inadolescence. Journal of Abnormal Child Psychology,27, 357–370.

Fergusson, D.M., Swain-Campbell, N.R., & Horwood,L.J. (2002). Deviant peer affiliations, crime andsubstance use: A fixed effects regression analysis.Journal of Abnormal Child Psychology, 30, 419–430.

Field, T.M. (1979). Infant behaviors directed towardpeers and adults in the presence and absence ofmother. Infant Behavior and Development, 2, 47–54.

Fordham, K., & Stevenson-Hinde, J. (1999). Shyness,friendship quality, and adjustment during middlechildhood. Journal of Child Psychology and Psychi-atry, 40, 757–768.

Freud, A., & Dann, S. (1949). An experiment in groupupbringing. In A. Freud & H. Hartmann (Eds.), Thepsychoanalytic study of the child (pp. 127–168).Oxford, England: International Universities Press.

Furman, L.N., & Walden, T.A. (1990). Effect of scriptknowledge on preschool children’s communicativeinteractions. Developmental Psychology, 26, 227–233.

Gallagher, K.C. (2002). Does child temperament mod-erate the influence of parenting on adjustment?Developmental Review, 22, 623–643.

Garner, P.W., Robertson, S., & Smith, G. (1997).Preschool children’s emotional expression with peers:The roles of gender and emotion socialization. SexRoles, 36, 675–691.

Garvey, C., & Kramer, T.L. (1989). The language of socialpretend play. Developmental Review, 9, 364–382.

Gazelle, H., & Ladd, G.W. (2003). Anxious solitude andpeer exclusion: A diathesis-stress model of internalis-ing trajectories in childhood. Child Development, 74,257–278.

Giles, J.W., & Heyman, G.D. (2003). Preschoolers’beliefs about the stability of anxious behaviour:Implications for navigating social challenges. SocialDevelopment, 12, 182–197.

Goodman, M.R., Stormshak, E.A., & Dishion, T.J.(2001). The significance of peer victimization at twopoints in development. Journal of Applied Develop-mental Psychology, 22, 507–526.

Goossens, L., & Beyers, W. (2002). Comparing meas-ures of childhood loneliness Internal consistency andconfirmatory factor analysis. Journal of Clinical Childand Adolescent Psychology, 31, 252–262.

Gottman, J.M. (1983). How children become friends.Monographs of the Society for Research in ChildDevelopment, 48 (Serial No. 86).

Gottman, J.M., & Katz, L.F. (1989). Effects of maritaldiscord on young children’s peer interaction andhealth. Developmental Psychology, 25, 373–381.

Graham, S., & Juvonen, J. (1998). Self-blame and peervictimization in middle school: An attributionalanalysis. Developmental Psychology, 34, 587–599.

Gunnar, M.R., Senior, K., & Hartup, W.W. (1984). Peerpresence and the exploratory behavior of eighteen-and thirty-month-old children. Child Development,55, 1103–1109.

Gunnar, M.R., & Donzella, B. (2002). Social regulationof the cortisol levels in early human development.Psychoneuroendocrinology, 27, 199–220.


Guralnick, M.J., Connor, R.T., Hammond, M.A., Gott-man, J.M., & Kinnish, K. (1996). The peer relations ofpreschool children with communication disorders.Child Development, 67, 471–489.

Guralnick, M.J., Paul-Brown, D., Groom, J.M., Booth,C. L., Hammond, M.A., Tupper, D.B., & Gelenter, A.(1998). Conflict resolution patterns of preschoolchildren with and without developmental delays inheterogeneous playgroups. Early Education andDevelopment, 9, 49–77.

Harding, C.G., & Golinkoff, R.M. (1979). The origins ofintentional vocalisations in prelinguistic infants.Child Development, 50, 33–40.

Harkness, S. (2002). Culture and social development:Explanations and evidence. In P.K. Smith & C.H.Hart (Eds.), Blackwell handbook of childhood socialdevelopment (pp. 60–77). Oxford: Blackwell.

Hay, D.F. (1984). Social conflict in early childhood.Annals of Child Development, 1, 1–44.

Hay, D.F. (1985). Learning to form relationships ininfancy: Parallel attainments with parents and peers.Developmental Review, 5, 122–161.

Hay, D.F. (2000). Toddlers talk about property tofamiliar peers. Poster presented at the InternationalConference on Infant Studies, Brighton.

Hay, D.F. (in press). The beginnings of aggression andconflict in infancy. In R.E. Tremblay (Ed.), Thedevelopmental origins of aggression.

Hay, D.F., & Nash, A. (2002). Social development indifferent family arrangements. In P.K. Smith & C.H.Hart (Eds.), Blackwell handbook of childhood socialdevelopment (pp. 238–2610). Oxford: Blackwell.

Hay, D.F., & Pawlby, S. (2003). Prosocial developmentin relation to children’s and mothers’ psychologicalproblems. Child Development, 74, 1295–1308.

Hay, D.F., & Ross, H.S. (1982). The social nature ofearly conflict. Child Development, 53, 105–111.

Hay, D.F., Castle, J., & Davies, L. (2000). Toddlers’ useof force against familiar peers: A precursor of seriousaggression? Child Development, 71, 457–467.

Hay, D.F., Nash, A., & Pedersen, J. (1981). Responses ofsix-month-olds to the distress of their peers. ChildDevelopment, 52, 1071–1075.

Hay, D.F., Nash, A., & Pedersen, J. (1983). Interactionbetween six-month-old peers. Child Development, 54,557–562.

Hay, D.F., Vespo, J.E., & Zahn-Waxler, C. (1998).Young children’s conflicts with their siblings andmothers: Links with maternal depression and bipolarillness. British Journal of Developmental Psychology,16, 519–538.

Hay, D.F., Caplan, M.Z., Castle, J., & Stimson, C.A.(1991). Does sharing become increasingly rational inthe second year of life? Developmental Psychology,27, 987–993.

Hay, D.F., Castle, J., Davies, L., Demetriou, H., &Stimson, C. (1999). Prosocial action in very earlychildhood. Journal of Child Psychology and Psychi-atry, 40, 905–916.

Hay, D.F., Zahn-Waxler, C., Cummings, E.M., &Iannotti, R. (1992). Young children’s views aboutconflict with peers: A comparison of the daughtersand sons of depressed and well women. Journal ofChild Psychology and Psychiatry, 33, 669–683.

Hektner, J.M., August, G.J., & Realmuto, G.M. (2000).Patterns and temporal changes in peer affiliationamong aggressive and nonaggressive children parti-cipating in a summer school program. Journal ofClinical Child Psychology, 29, 603–614.

Henwood, P.G., & Solano, C.H. (1994). Loneliness inyoung children and their parents. Journal of GeneticPsychology, 155, 35–45.

Herrera C., & Dunn, J. (1997). Early experiences withfamily conflict: Implications for arguments with aclose friend. Developmental Psychology, 33, 869–881.

Hock, E., & Lutz, W.J. (2001). Peer rejection in child-hood: Effects on maternal depression and behaviorproblems in toddlers. Journal of Genetic Psychology,162, 167–177.

Hodgetts, J. (2003). Support quality and conflict withinaggressive/delinquent and non-aggressive/non-delin-quent friendship networks. Paper presented at theBiennial Meeting of the Society of Research in ChildDevelopment, April 2003, Tampa, Florida.

Hold-Cavell, B.C., Attili, G., & Schleidt, M. (1986). Across-cultural comparison of children’s behaviourduring their first year in a preschool. InternationalJournal of Behavioural Development, 9, 471–483.

Howes, C. (1980). Peer play scale as an index ofcomplexity of peer interaction. Developmental Psy-chology, 16, 371–372.

Howes, C., & Farver, J. (1987). Social pretend play in2-year-olds: Effects of age of partner. Early ChildhoodResearch Quarterly, 2, 305–314.

Howes, C., & Matheson, C.C. (1992). Sequences in thedevelopment of competent play with peers: Social andpretend play. Developmental Psychology, 28, 961–974.

Howes, C., & Phillipsen, L. (1992). Gender and friend-ship: Relationships within peer groups of youngchildren. Social Development, 1, 230–242.

Howes, C., & Phillipsen, L. (1998). Continuity inchildren’s relationships with peers. Social Develop-ment, 7, 340–349.

Howes, C., Matheson, C.C., & Hamilton, C.E. (1994).Maternal, teacher, and child-care history correlates ofchildren’s relationships with peers. Child Develop-ment, 65, 264–273.

Hughes, C., Dunn, J., & White, A. (1998). Trick or treat?Uneven understanding of mind and emotion andexecutive dysfunction in ‘hard-to-manage’ preschool-ers. Journal of Child Psychology and Psychiatry, 39,981–994.

Hughes, C., White, A., Sharpen, J., & Dunn, J. (2000).Antisocial, angry and unsympathetic: ‘Hard-to-man-age’ preschoolers’ peer problems and possible cognit-ive influences. Journal of Child Psychology andPsychiatry, 41, 169–179.

Hughes, C., Cutting, A.L., & Dunn, J. (2001). Actingnasty in the face of failure? Longitudinal observationsof ‘hard-to-manage’ children playing a rigged com-petitive game with a friend. Journal of Abnormal ChildPsychology, 29, 403–416.

Isaacs, J., & Hodges, E.V.E. (2003). Friendship contextsthat potentiate or inhibit the translation of aggressivethought into action. Paper presented at the BiennialMeeting of the Society for Research in Child Develop-ment, Tampa, Florida.


Ishikawa, F. (2003). Dyadic and triadic interactionbetween newly acquainted two-year-olds. Unpub-lished PhD thesis, Cambridge University.

Iskandar, N., Laursen, B., Finkelstein, B., & Fredrick-son, L. (1995). Conflict resolution among preschoolchildren: The appeal of negotiation in hypotheticaldisputes. Early Education and Development, 6, 359–376.

Ispa, J. (1981). Peer support among Soviet day caretoddlers. International Journal of Behavioural Devel-opment, 4, 255–269.

Jacobson, J.L. (1981). The role of inanimate objects inearly peer interaction. Child Development, 52, 618–626.

Jones, D.C. (1992). Parental divorce, family conflict,and friendship networks. Journal of Social andPersonal Relationships, 9, 219–235.

Katz, J.C., & Buchholz, E.S. (1999). ‘I did it myself’: Thenecessity of solo play for preschoolers. Early ChildDevelopment and Care, 155, 39–50.

Keating, C.F., & Heltman, K.R. (1994). Dominance anddeception in children and adults: Are leaders the bestmisleaders? Personality and Social Psychology Bulle-tin, 20, 312–321.

Keenan, K., & Shaw, D. (1997). Developmental andsocial influences on young girls’ early problembehavior. Psychological Bulletin, 121, 95–113.

Keenan, K., & Wakschlag, L.S. (2002). Can a validdiagnosis of disruptive behavior disorder be made inpreschool children? American Journal of Psychiatry,159, 351–358.

Kemple, K., Speranza, H., & Hazen, N. (1992). Cohesivediscourse and peer acceptance: longitudinal relationsin the pre-school years. Merrill Palmer Quarterly, 38,364–381.

Kenny, D.A. (1979). Correlation and causality. NewYork: Wiley.

Kerns, K.A., & Barth, J.M. (1995). Attachment andplay-convergence across components of parent–childrelationships and their relations to peer competence.Journal of Social and Personal Relationships, 12, 243–260.

Khatri, P., & Kupersmidt, J. (2003). Aggression, peervictimisation, and social relationships among Indianyouth. International Journal of Behavioural Develop-ment, 27, 87–95.

Kochanska, G. (1992). Children’s interpersonal influ-ence with mothers and peers. Developmental Psychol-ogy, 28, 491–499.

Kuentay, A.C. (2001). Occasions for providing resolu-tions (or not) in Turkish preschool conversationalnarratives. Narrative Inquiry, 11, 411–450.

Lacourse, E., Cote, S., Nagin, D.S., Vitaro, F.,Brendgen, M., & Tremblay, R.E. (2002). A lon-gitudinal-experimental approach to testing theoriesof antisocial behavior development. Development andPsychopathology, 14, 909–924.

Ladd, B.K., & Ladd, G.W. (2001). Variations in peervictimization: Relations to children’s maladjustment.In J. Juvonen & S. Graham (Eds.), Peer harassment inschool: The plight of the vulnerable and victimised(pp. 25–48). New York: Guilford.

Ladd, G.W., & Hart, C.H. (1992). Creating informalplay opportunities: Are parents’ and preschoolers’

initiations related to children’s competence withpeers? Developmental Psychology, 28, 1179–1187.

Ladd, G.W., Price, J.M., & Hart, C.H. (1988). Predictingpreschoolers’ peer status from their playgroundbehaviors. Child Development, 59, 986–992.

Lafreniere, P.J. (1996). Cooperation as a conditionalstrategy among peers: Influence of social ecology andkin relations. International Journal of BehaviouralDevelopment, 19, 39–52.

Lafreniere, P.J., & Charlesworth, W. (1987). Effects offriendship and dominance status on preschoolers’resource utilization in a cooperative competitive situ-ation. International Journal of Behavioral Develop-ment, 10, 345–358.

Lahey, B.B., Waldman, I.D., & McBurnett, K. (1999).The development of antisocial behaviour: An integra-tive causal model. Journal of Child Psychology andPsychiatry, 40, 669–682.

Larson, R.W. (1997). The emergence of solitude as aconstructive domain of experience in early adoles-cence. Child Development, 68, 80–93.

Laursen, B., & Hartup, W.W. (1989). The dynamics ofpreschool children’s conflicts. Merrill-Palmer Quar-terly, 35, 281–297.

Lewis, M., & Ramsay, D.S. (1995). Developmentalchange in infants’ responsiveness to stress. ChildDevelopment, 66, 657–670.

Light, P., & Glachan, M. (1985). Facilitation of indi-vidual problem solving through peer interaction.Educational Psychology, 5, 217–225.

Lobdell, J., & Perlman, D. (1986). The intergenerationaltransmission of loneliness: A study of college femalesand their parents. Journal of Marriage and Family, 3,589–595.

Loeber, R., & Hay, D.F. (1997). Key issues in thedevelopment of aggression and violence from child-hood to early adulthood. Annual Review of Psycho-logy, 48, 371–410.

Løkken, G. (2000). Tracing the social style of toddlerpeers. Scandinavian Journal of Educational Research,44, 163–167.

Maccoby, E.E. (1998). The two sexes: Growing up apart,coming together. Cambridge, MA: Harvard UniversityPress.

Mahady Wilton, M.M., & Craig, W.M. (2000). Emotionalregulation and display in classroom victims of bully-ing: Characteristic expressions of affect, coping stylesand relevant contextual factors. Social Development,9, 226–245.

Maudry, M., & Nekula, M. (1939). Social relationsbetween children of the same age during the firsttwo years of life. Journal of Genetic Psychology, 54,193–215.

McCord, J. (1992). The Cambridge-Somerville Study: Apioneering longitudinal-experimental study of delin-quency prevention. In J. McCord & R.E. Tremblay(Eds.), Preventing antisocial behavior: Interventionsfrom birth through adolescence (pp. 196–206). NewYork: Guilford Press.

McDowell, D.J., O’Neil, R., & Parke, R.D. (2000).Display rule application in a disappointing situationand children’s emotional reactivity: Relations withsocial competence. Merrill-Palmer Quarterly, 46, 306–324.


McDowell, D.J., Parke, R.D., & Wang, S.J. (2003).Differences between mothers’ and fathers’ advice-giving style and content: Relations with social com-petence and psychological functioning in middlechildhood. Merrill-Palmer Quarterly, 49, 55–77.

McElwain, N.L., & Volling, B. (2002). Relating indi-vidual control, social understanding, and gender tochild–friend interaction: A relationships perspective.Social Development, 11, 362–385.

Melson, G.F., Ladd, G.W., & Hsu, H.-C. (1993).Maternal support networks, maternal cognitions,and young children’s social and cognitive develop-ment. Child Development, 64, 1401–1417.

Meyers, S.C. (1990). Verbal behaviors of preschoolstutterers and conversational partners: Observingreciprocal relationships. Journal of Speech and Hear-ing Disorders, 55, 706–712.

Miller, A.L., McDonough, S.C., Rosenblum, K.L., &Sameroff, A.J. (2002). Emotion regulation in context:Situational effects on infant and caregiver behavior.Infancy, 3, 403–433.

Miller-Johnson, S., Coie, J.D., Maumary, G.A., & Bier-man, K. Conduct Problems Prevention ResearchGroup. (2002). Peer rejection and aggression andearly starter models of conduct disorder. Journal ofAbnormal Child Psychology, 30, 217–230.

Moffitt, T.E., & Caspi, A. (2001). Childhood predictorsdifferentiate life-course persistent and adolescent-limited antisocial pathways among males andfemales. Development and Psychopathology, 13,355–375.

Moody, J. (2001). Peer influence groups: Identifyingdense clusters in large networks. Social Networks, 23,261–283.

Moss, E., Goesselin, C., Parent, S., Rousseau, D., &Dumont, M. (1997). Attachment and joint problem-solving experiences during the preschool period.Social Development, 6, 1–17.

Mundy, P., Card, J., & Fox, N. (2000). EEG correlates ofthe development of infant joint attention skills.Developmental Psychobiology, 36, 325–338.

Mundy, P., Kasari, C., & Sigman, M. (1992). Nonverbalcommunication, affective sharing, and intersubjec-tivity. Infant Behavior and Development, 15, 377–381.

Musatti, T., & Mayer, S. (1987). Object substitution: Itsnature and function in early pretend play. HumanDevelopment, 30, 225–235.

Musun-Miller, L. (1990). Sociometrics with preschoolchildren: Agreement between different strategies.Journal of Applied Developmental Psychology, 11,195–207.

Oakes, L.M. (1994). Development of infants’ use ofcontinuity cues in their perception of causality.Developmental Psychology, 30, 869–879.

Olweus, D., & Endresen, I.M. (1998). The importance ofsex-of-stimulus object: Age trends and sex differencesin empathetic responses. Social Development, 7, 370–388.

Pastor, D.L. (1981). The quality of mother–infantattachment and its relationship to toddlers’ initialsociability with peers. Developmental Psychology, 17,323–335.

Pellegrini, A.D., & Long, J.D. (2002). A longitudinalstudy of bullying, dominance, and victimizationduring the transition from primary school through

secondary school. British Journal of DevelopmentalPsychology, 20, 259–280.

Pellegrini, A.D., & Smith, P.K. (1998). Physical activityplay: The nature and function of a neglected aspect ofplay. Child Development, 69, 577–598.

Perry, D., Kusel, S., & Perry, L. (1988). Victims of peeraggression. Developmental Psychology, 24, 807–814.

Peterson, C.C., & Siegal, M. (2002). Mindreading andmoral awareness in popular and rejected preschool-ers. British Journal of Developmental Psychology, 20,205–224.

Pickert, S.M. (1985). Functions of repetition in theconversations of preschool children. Journal of Gen-eral Psychology, 112, 35–41.

Premack D., & Premack A.J. (1997). Infants attributevalue+- to the goal-directed actions of self-propelledobjects. Journal of Cognitive Neuroscience, 9, 848–856

Price, J.M., & Dodge, K.A. (1989). Reactive andproactive aggression in childhood: Relations to peerstatus and social context dimensions. Journal ofAbnormal Child Psychology, 17, 455–471.

Principe, G.F., & Ceci, S.J. (2002). ‘I saw it with my ownears’: The effects of peer conversations on preschool-ers’ reports of nonexperienced events. Journal ofExperimental Child Psychology, 83, 1–25.

Qualter, P., & Munn, P. (2002). The separateness ofsocial and emotional loneliness in childhood. Journalof Child Psychology and Psychiatry, 43, 233–244.

Quinton, D., Pickles, A., Maughan, B., & Rutter, M.(1993). Partners, peers, and pathways: Assortativepairing and continuities in conduct disorder. Devel-opment and Psychopathology, 5, 763–783.

Rabbitt, P. (Ed.). (1997). Methodology of frontal andexecutive function. Hove: Psychology Press.

Raine, A., Reynolds, C., Venables, P.H., Mednick, S.A.,& Farrington, D.P. (1998). Fearlessness, stimulation-seeking, and large body size at age 3 years as earlypredispositions to childhood aggression at age 11years. Archives of General Psychiatry, 55, 745–751.

Rochat, P., Morgan, R., & Carpenter, V. (1997). Younginfants’ sensitivity to movement information specify-ing social causality. Cognitive Development, 12, 537–561.

Rogers, S.J., Hepburn, S.L., Stackhouse, T., & Wehner,E. (2003). Imitation performance in toddlers withautism and those with other developmental disor-ders. Journal of Child Psychology and Psychiatry, 44,763–781.

Ross, H.S. (1982). Establishment of social gamesamong toddlers. Developmental Psychology, 18,509–518.

Ross, H.S., Tesla, C., Kenyon, B., & Lollis, S.P. (1990).Maternal intervention in toddler peer conflict: thesocialization of principles of justice. DevelopmentalPsychology, 26, 994–1003.

Ross, H., Martin, J., Perlman, M., Smith, M., Black-more, E., & Hunter, J. (1996). Autonomy andauthority in the resolution of sibling disputes. NewDirections for Child Development, 13, 71–90.

Rothstein-Fisch, C., & Howes, C. (1988). Toddler peerinteraction in mixed-age groups. Journal of AppliedDevelopmental Psychology, 9, 211–218.

Rubenstein, J.L., Howes, C., & Pedersen, F.A. (1982).Second order effects of peers on mother–toddler


interaction. Infant Behavior and Development, 5,185–194.

Rubin, K.H., Bukowski, W., & Parker, J.G. (1998). Peerinteractions, relationships, and groups. In W. Damon(Ed.), Handbook of child psychology (vol. 3, N. Eisen-berg, vol. ed., pp. 619–700). Chichester: Wiley.

Rubin, K.H., Burgess, K.B., & Coplan, R.J. (2002).Social withdrawal and shyness. In P.K. Smith & C.H.Hart (Eds.), Blackwell handbook of childhood socialdevelopment (pp. 329–352). Oxford: Blackwell.

Rubin, K.H., Hastings, P., Stewart, S., & McNichol, K.(1998). Intrapersonal and maternal correlates ofaggression, conflict, and externalising problems intoddlers. Child Development, 69, 1614–1629.

Rubin, K.H., Burgess, K.B., & Hastings, P.D. (2002).Stability and social-behavioral consequences of tod-dlers’ inhibited temperament and parenting behav-iors. Child Development, 73, 483–495.

Rubin, K.H., Burgess, K.B., Dwyer, K.M., & Hastings,P.D. (2003). Predicting preschoolers’ externalisingbehaviors from toddler temperament, conflict andmaternal negativity. Developmental Psychology, 39,164–176.

Sackin, S., & Thelen, E. (1984). An ethological study ofpeaceful associative outcomes to conflict in preschoolchildren. Child Development, 55, 1098–1102.

Sagi, A., & Hoffman, M.L. (1976). Empathic distress inthenewborn.Developmental Psychology,12, 175–176.

Salmivalli, C. (2002). Is there an age decline in victim-ization by peers at school? Educational Research, 44,269–277.

Schlottman, A., & Surian, L. (1999). Do 9-month-oldsperceive causation-at-a-distance? Perception, 28,1105–1113.

Schonberg, M.A., Tussey, J.L., & Stickney, J.A. (2003).Moderators of the relationships between proactive andreactive aggression and peer rejection. Paper presen-ted at the Biennial Meeting of the Society for Researchin Child Development, Tampa, Florida.

Schwartz, D., Dodge, K.A., Pettit, G.S., & Bates J.E.(1997). The early socialization of aggressive victims ofbullying. Child Development, 68, 665–675.

Schwartz, D., McFayden-Ketchum, S., Dodge, K.A.,Pettit, G.S., & Bates, J.E. (1999). Early behaviorproblems as a predictor of later peer group victimiza-tion: Moderators and mediators in the pathways ofsocial risk. Journal of Abnormal Child Psychology, 27,191–201.

Schwartz, D., Dodge, K.A., Coie, J.D., Hubbard, J.A.,Cillessen, A.H., Lemerise, E.A., & Bateman, (1998).Social-cognitive and behavioral correlates of aggres-sion and victimization in boys’ play groups. Journal ofAbnormal Child Psychology, 26, 431–440.

Sebanc, A.M., Pierce, Cheatham, C.L., & Gunnar, M.(2003). Gendered social worlds in preschool: Dom-inance, peer acceptance and assertive social skills inboys’ and girls’ peer groups. Social Development, 12,91–106.

Shantz, C.U. (1987). Conflicts between children. ChildDevelopment, 58, 282–305.

Shaw, D.S., Keenan, K., & Vondra, J.I. (1994). Devel-opmental precursors of externalizing behavior: Ages 1to 3. Developmental Psychology, 30, 355–364.

Shweder, R.A., Goodnow, J., Hatano, G., LeVine, R.A.,Markus, H., & Miller, P. (1998). The cultural

psychology of development: One mind, many men-talities. In W. Damon (Ed.), Handbook of childpsychology (vol. 1, R. Lerner, vol. ed., pp. 865–937).Chichester: Wiley.

Simner, M.L. (1971). Newborn’s response to the cry ofanother infant. Developmental Psychology, 5, 136–150.

Slaughter, V., Dennis, M.J., & Pritchard, M. (2002).Theory of mind and peer acceptance in preschoolchildren. British Journal of Developmental Psycho-logy, 20, 545–564.

Slomkowski, C.L., & Killen, M. (1992). Young children’sconceptions of transgressions with friends and non-friends. International Journal of Behavioural Develop-ment, 15, 247–258.

Smith, P.K., & Levan, S. (1995). Perceptions andexperiences of bullying in younger pupils. BritishJournal of Educational Psychology, 65, 489–500.

Smith, P.K., Cowie, H., Olafsson, R.F., & Liefooghe, P.D.(2002). Definitions of bullying: A comparison of termsused, and age and gender differences, in a fourteencountry international comparison. Child Develop-ment, 73, 1119–1133.

Smith, P.K., Madsen, K., & Moody, J. (1999). Whatcauses the age decline in reports of being bullied atschool? Towards a developmental analysis of risks ofbeing bullied. Educational Research, 41, 267–285.

Sparkes, K.K. (1991). Cooperative and competitivebehavior in dyadic game-playing: A comparison ofAnglo-American and Chinese children. Early ChildDevelopment and Care, 68, 37–47.

Stefani, L.H., & Camaioni, L. (1983). Effects of famil-iarity on peer interaction in the first year of life. EarlyChild Development and Care, 11, 45–54.

Stevenson-Hinde, J., & Glover, A. (1996). Shy girls andboys: A new look. Journal of Child Psychology andPsychiatry, 37, 181–187.

Stifter, C.A., & Spinrad, T. (2002). The effect ofexcessive crying on the development of emotionregulation. Infancy, 3, 133–152.

Stormshak, E.A., Bierman, K.L., Bruschi, C., Dodge,K.A., & Coie, J. D. (1999). The relation betweenbehavior problems and peer preference in differentclassroom contexts. Child Development, 70, 169–182.

Strayer, F.F., & Strayer, J. (1976). An ethologicalanalysis of social agonism and dominance relationsamong preschool children. Child Development, 47,980–989.

Strayer, J., & Strayer, F.F. (1978). Social aggressionand power relations among preschool children.Aggressive Behavior, 4, 173–182.

Striano, T., & Rochat, P. (1999). Developmental linksbetween dyadic and triadic social competence ininfancy. British Journal of Developmental Psychology,17, 563–584.

Sutton, J., Smith, P.K., & Swettenham, J. (1999).Bullying and ‘theory of mind’: A critique of the ‘socialskills deficit’ view of antisocial behaviour. SocialDevelopment, 8, 117–127.

Tremblay, R.E. (2001). Antisocial behaviour in child-hood. Paper presented to European Society of Crim-inology, Lausanne.

Unnever, J.D., & Cornell, D.G. (2003). Bullying, self-control and ADHD. Journal of Interpersonal Violence,18, 129–147.


Vandell, D.L. (1980). Sociability with peer and motherduring the first year. Developmental Psychology, 16,355–361.

Vandell, D.L., & Wilson, K.S. (1987). Infants’ interac-tions with mother, sibling, and peer: Contrasts andrelations between interaction systems. Child Devel-opment, 58, 176–186.

Vandell, D.L., Wilson, K.S., & Buchanan, N.R. (1980).Peer interaction in the first year of life: An examina-tion of its structure, content, and sensitivity to toys.Child Development, 51, 481–488.

Vandell, D.L., Wilson, K.S., & Whalen, W.T. (1981).Birth-order and social-experience differences in in-fant–peer interaction. Developmental Psychology, 17,438–445.

Vermande, M., van den Oord, E.J.C.G., Goudena, P.P.,& Rispens, J. (2000). Structural characteristics ofaggressor–victim relationships in Dutch school clas-ses of 4 to 5 year olds. Aggressive Behavior, 26, 11–31.

Vitaro, F., Gagnon, C., & Tremblay, R. (1990). Predict-ing stable peer rejection from kindergarten to Gradeone. Journal of Clinical Child Psychology, 19, 257–264.

Willoughby, M., Kupersmidt, J., & Bryant, D. (2001).Overt and covert dimensions of antisocial behavior in

early childhood. Journal of Abnormal Child Psycho-logy, 29, 177–187.

Woodward, L.J., & Fergusson, D. (1999). Childhoodpeer relationship problems and psychosocial adjust-ment in late adolescence. Journal of Abnormal ChildPsychology, 27, 87–104.

Woodward, L.J., & Fergusson, D.M. (2000). Childhoodpeer relationship problems and later risks of educa-tional under-achievement and unemployment. Jour-nal of Child Psychology and Psychiatry, 41, 191–201.

Yoon, J.S., Hughes, J.N., Cavell, T.A., & Thompson, B.(2000). Social cognitive differences between aggres-sive-rejected and aggressive-nonrejected children.Journal of School Psychology, 38, 551–570.

Young, G., & Lewis, M. (1979). Effects of familiarity andmaternal attention on infant peer relations. MerrillPalmer Quarterly, 25, 105–119.

Zahn-Waxler, C., Radke-Yarrow, M., Wagner, E., &Chapman, M. (1992). Development of concern forothers. Developmental Psychology, 28, 126–136.

Zaslow, M. (1980). Relationships among peers in kib-butz toddler groups. Child Psychiatry and HumanDevelopment, 10, 178–189.

Manuscript accepted 12 July 2003


Documents

Peerrelationsin childhood - Cardiff Universitypsych.cf.ac.uk/home2/hay/hay2004.pdf · Peerrelationsin childhood Dale F. Hay, ... than entering into ultimately unsatisfying debates