peer rejection, aggressive or withdrawn behavior, and

Peer Rejection, Aggressive or Withdrawn Behavior, and Psychological

Maladjustment from Ages 5 to 12: An Examination of Four Predictive Models

Gary W. LaddArizona State University

Findings yielded a comprehensive portrait of the predictive relations among children’s aggressive or withdrawnbehaviors, peer rejection, and psychological maladjustment across the 5 – 12 age period. Examination of peer re-jection in different variable contexts and across repeated intervals throughout childhood revealed differences in thetiming, strength, and consistency of this risk factor as a distinct (additive) predictor of externalizing versus inter-nalizing problems. In conjunction with aggressive behavior, peer rejection proved to be a stronger additive predictorof externalizing problems during early rather than later childhood. Relative to withdrawn behavior, rejection’s ef-ficacy as a distinct predictor of internalizing problems was significant early in childhood and increased progres-sively thereafter. These additive path models fit the data better than did disorder-driven or transactional models.

The assumption that children’s early behavioralpropensities or their peer experiences engenderedpsychological maladjustment gained prominenceduring the 1950s and 1960s (cf. Freud & Dann, 1951;Morris, Soroker, & Burruss, 1954; Robins, 1966; Roff,1961, 1963), and fostered separate empirical tradi-tions that were devoted to explicating each of theseconstellations of variables as risk factors for early-and later-emerging dysfunction (for reviews, seeKupersmidt, Coie, & Dodge, 1990; Ladd, 2005;McDougall, Hymel, Vaillancourt, & Mercer, 2001;Parker & Asher, 1987). From the outset, researchersutilized longitudinal studiesFthe most valuable ofwhich were based on follow-up or follow-through(prospective) designs (see Parker & Asher, 1987)Ftodiscern which facets of children’s behavior, or theirpeer relationships, were most predictive of later-emerging dysfunctions (e.g., see Roff, Sells, &Golden, 1972; Roff & Wirt, 1984).

Children’s Behavior Versus Peer Relations: Main EffectPredictive Models

Within one empirical tradition, investigatorsworked from the hypothesis that children’s behav-

ioral propensities were the principal determinants ofpsychopathology, and a sizable proportion of thisresearch was designed to elucidate the predictivecontributions of aggressive and withdrawn behaviorto later dysfunction. An influential premise was thatchildren’s propensity to engage in confrontive formsof aggression became an enduring behavioral ori-entation that eventually brought about dysfunction(see Caspi, Elder, & Bem, 1987; Moss & Susman,1980; Olweus, 1979). Early and recent longitudinalfindings corroborated this premise by showing thataggressive behavior, even at early ages (e.g., seeTremblay, Pihl, Vitaro, & Dobkin, 1994), was fairlystable for boys and girls (see Cairns, Cairns, Neck-erman, Ferguson, & Gariepy, 1989; Olweus, 1979)and moderately predictive of maladjustment, par-ticularly externalizing problems (e.g., Cairns &Cairns, 1994; Caspi et al., 1987). Another influentialpremise was that children’s propensity to engage inwithdrawn behavior led to maladjustment. Al-though early findings suggested that withdrawnbehavior was not particularly stable or predictive ofdysfunction (see Michael, Morris, & Soroker, 1957;Morris et al., 1954; Wanlass & Prinz, 1982), laterlongitudinal evidence revealed that certain subtypesof this behavior (e.g., anxious – depressed, anxious –solitary styles) were moderately stable (Gazelle &Ladd, 2003; Harrist, Zaia, Bates, Dodge, & Pettit,1997) and modestly predictive of internalizingproblems (e.g., Caspi, Elder, & Bem, 1988; Gazelle &Ladd, 2003).

Investigators within another research traditionwere guided by the hypothesis that children’s

r 2006 by the Society for Research in Child Development, Inc.All rights reserved. 0009-3920/2006/7704-0002

Portions of this study were conducted as part of the PathwaysProject, a larger longitudinal investigation of children’s social/psychological/scholastic adjustment in school contexts that issupported by the National Institutes of Health (1 RO1MH-49223,2-RO1MH-49223, R01HD-045906 to Gary W. Ladd). Special ap-preciation is expressed to all the children and parents who madethis study possible, and to members of the Pathways Project forassistance with data collection.

Correspondence concerning this article should be addressed toGary Ladd, P.O. Box 852502, Arizona State University, Tempe, AZ85287-2502. Electronic mail may be sent to [email protected].

Child Development, July/August 2006, Volume 77, Number 4, Pages 822 – 846

exposure to adverse peer relations (e.g., peer grouprejection, low peer acceptance) was a fundamentaldeterminant of later maladjustment. Initial efforts topursue this research agenda were influenced byideological perspectives in which children’s devel-opment was seen as malleable and largely influencedby their social relationships, including those withage-mates (e.g., see Hartup, 1970; Sullivan, 1953).This premise was strengthened by experimental ev-idence indicating that age-mates influenced chil-dren’s (and some primate’s) socialization, and thataberrant peer socialization predicted later-emergingpsychopathology (e.g., Freud & Dann, 1951; Furman,Rahe, & Hartup, 1979; Harlow, Harlow, Dodsworth,& Arling, 1966; Suomi & Harlow, 1972). Past andrecent efforts to elucidate the contributions of chil-dren’s peer relationships to the development ofpsychopathology have consistently implicated poorpeer group relations, and peer group rejection inparticular, as a precursor of externalizing and inter-nalizing problems during the grade school years andthereafter (e.g., DeRosier, Kupersmidt, & Patterson,1994; Hymel, Rubin, Rowden, & LeMare, 1990;Kupersmidt & Coie, 1990; Ladd & Troop-Gordon,2003). Moreover, it was discovered that the predic-tive link between children’s peer acceptance/rejec-tion and subsequent psychological adjustment wasrobust even when it was evaluated in conjunctionwith other types of peer relationships (e.g., friend-ship, friendlessness, peer victimization; see Bagwell,Newcomb, & Bukowski, 1998; Ladd, Herald, & An-drews, 2006; Ladd & Troop-Gordon, 2003).

One consequence of this bifurcation in investiga-tive paradigms was that two rather distinct empiricalliteratures developed around these premises. Espe-cially during the first generation of longitudinalstudies, researchersFdepending on their theoreticalorientationsFtended to favor one of the aforemen-tioned premises over the other and, thus, investi-gated the predictors of psychological maladjustmentfrom a ‘‘main effects’’ perspective. As might be sur-mised, the evidence that accrued within these twobodies of evidence tended to implicate either ag-gression or peer group rejection as a principal riskfactor for later dysfunction.

Children’s Behavior and Peer Relations: Child andEnvironment Predictive Models

As the limitations of main effects perspectivesbecame apparent, investigators began to devise morecomplex prediction models in which multiple riskfactors were recognized as potential determinants ofdysfunction (see Coie et al., 1993). These models

were based on the premise that specific child be-haviors (e.g., aggression, withdrawal) and adversepeer experiences (e.g., peer rejection) co-determinechildren’s future psychological adjustment (seeLadd, 1989, 2003; McDougall et al., 2001; Parker,Rubin, Price, & DeRosier, 1995). This paradigm shiftwas first evident in a critical review published byParker and Asher (1987) in which the authors en-couraged researchers to evaluate competing explana-tionsFframed as ‘‘causal’’ versus ‘‘incidental’’modelsFof how children’s behavior and peer rela-tionships contributed to their future adjustment. Theincidental model was essentially a restatement of thehypothesis that children’s preexisting characteristics,including their behavioral dispositions, were theprincipal causes of later maladaptive outcomes. Itwas assumed that (a) premorbid forms of a later-emerging disorder, children’s behavioral propensi-ties, or both were the principal precursors of latermaladjustment and (b) these same child attributescaused children to develop poor peer relation-shipsFthat is, children’s peer relational difficultieswere ‘‘incidental’’ consequences (i.e., markers) oftheir behavioral dispositions, and played no causalrole in the development of later dysfunction. Incontrast, the main assumption of the ‘‘causal’’ modelwas that, along with children’s behavioral disposi-tions, participation in adverse peer relationships (i.e.,exposure to peer group rejection) was influential inshaping children’s later adjustment.

In part, this proposal was a forerunner of con-temporary ‘‘child and environment’’ frameworks.Although several variants of child and environmentmodels have been proposed to elucidate the distinctversus conjoint predictive contributions of children’sbehavior and peer relationships to later psychologi-cal maladjustment (see Ladd, 2003), most of the ev-idence amassed thus far conforms to an additivemodel. Additive models imply that, separate (par-tially independent) from the contributions of chil-dren’s behavior, the experiences they have in peerrelationships increase or decrease the probability ofmaladjustment.

Additive Child by Environment Prediction Models:Principal Premises and Extant Evidence

The hypothesis that aggression and peer grouprejection additively predict later maladjustment hasbeen based on assumptions similar to those depictedin the following exemplary model: Children whofrequently engage in confrontive aggression amongpeersFwhether because of inheritance, learning, orsome transaction between the two (see Caspi et al.,

Predicting Maladjustment 823

1987)Fdevelop a style of responding to interper-sonal contingencies that is eventually generalized toa larger range of environmental contingencies (e.g.,constraints that require delay of gratification; prohi-bitions or rules for behavior, etc.). Over time, thisprogression causes children to develop broader andmore serious forms of maladjustment, such as ex-ternalizing problems. Further, because children whopossess aggressive behavioral orientations often al-ienate peers, they are exposed to another (relational)risk factorFpeer group rejection. Once in a rejectingpeer milieu, children are exposed to debilitating re-lational processes (e.g., maltreatment by peers, ex-clusion from peer activities, isolation from peeraffirmation or support, etc.; see Asher, Rose, &Gabriel, 2001; Buhs & Ladd, 2001; Coie, 1990, 2004)that take an additional toll on their future adjustment(i.e., these stressors contribute to maladjustment inways that are distinct from aggressive behavior).Consistent with this model, there is evidence indi-cating that aggression and peer rejection duringchildhood make additive contributions to the pre-diction of future externalizing problems. For exam-ple, findings from recent longitudinal studies haveshown that aggression and peer rejection measuredin grade school made distinct contributions to theprediction of externalizing problems later in gradeschool and during preadolescence and adolescence(Coie, Lochman, Terry, & Hyman, 1992; Cowan &Cowan, 2004; Hymel et al., 1990).

The hypothesis that withdrawn behavioral stylesand peer group rejection additively predict latermaladjustment has been less well investigated. Oneof the principal assumptions guiding this line of in-vestigation can be portrayed as follows: Childrenwho manifest a withdrawn behavioral styleFwhether rooted in temperament, socialization, or theinteraction of the two (see Rubin & Asendorpf,1993)Fare prone to respond to social contingenciesby seeking solitude, avoiding contact, or interactinginfrequently with peers. Because withdrawn behav-ior may be interpreted by peers as disinterest or alack of social responsiveness, it may foster adverseinterpersonal reactions such as disliking and rejec-tion. In turn, exposure to peer rejection increases theprobability that withdrawn children will developinternalizing problems because the stressors associ-ated with this social position cause children to feelanxious or distressed and, thus, instigate and main-tain internalizing problems (Gazelle & Ladd, 2003;Rubin, LeMare, & Lollis, 1990).

Although sparse, corroborating evidence for thismodel has begun to accrue. Findings from a 1-yearlongitudinal study conducted by Renshaw and

Brown (1993) revealed that both withdrawn behaviorand low peer group acceptance made distinct pre-dictive contributions to children’s loneliness over a1-year interval. Similar linkages were reported byBoivin and Hymel (1997) for data gathered concur-rently rather than longitudinally. In a 5-year pro-spective longitudinal study, Gazelle and Ladd (2003)examined how withdrawn behavior and peer groupexclusion (one manifestation of peer group rejection)were linked with children’s trajectories toward de-pression. They found that children who were pronetoward anxious solitary behavior during the earlyschool years were more likely to manifest andmaintain depressive symptoms if they also had beensubjected to higher levels of peer exclusion.

In sum, there is a modicum of evidence to suggestthat specific behavioral orientations (i.e., aggressive,withdrawn) and peer group rejection during child-hood constitute related, but not entirely redundant,risk factors for later psychological dysfunction. Atthis juncture, the overarching pattern of findingssuggests that, when aggression and rejection areevaluated conjointly as predictors of later externaliz-ing problems, aggression typically emerges as thestronger of the two risk factors or tends to be the onlypredictor to account for significant nonshared vari-ance in externalizing problems (for reviews, see Coie,2004; Ladd, 2003; McDougall et al., 2001). Whenthese same two risk factors are evaluated as predic-tors of internalizing problems, the additive contribu-tion of peer group rejection (relative to aggression)tends to be larger (McDougall et al., 2001). In con-trast, the hypothesis that withdrawn behavioralstyles and peer group rejection contribute additivelyto later maladjustment has been less well investi-gated. At present, there is preliminary evidence tosuggest that children disposed toward withdrawnbehavioral patterns are at risk for later internalizingproblems (e.g., Boivin, Hymel, & Bukowski, 1995;Caspi et al., 1988; Ladd & Troop-Gordon, 2003) andthat exposure to peer group rejection additively en-hances the prediction of these problems (see Ladd,2003; McDougall et al., 2001). Next to nothing isknown about whether children who are disposedtoward withdrawn behavior are at risk for external-izing problems and whether peer group rejectionexacerbates this risk (see Ladd, 2003).

Clearly, much remains to be learned about howpeer group rejection, in conjunction with children’saggressive or withdrawn behavioral styles, contrib-utes to the prediction of childhood externalizingand internalizing problems. Accordingly, this inves-tigation was undertaken to address two principalaims.

824 Ladd

Aim 1: Conduct a More Comprehensive Analysis of PeerRejection as an Additive Risk Factor

At present, knowledge remains limited with re-spect to when, to what extent, and how consistently peergroup rejection adds to (beyond aggressive or with-drawn behavioral styles) the prediction of external-izing or internalizing problems during the gradeschool years (i.e., from age 5 – 12). In large part, thisvoid stems from an overreliance on nonpanel, single-indicator, follow-up longitudinal designs. That is, instudies of additive risk factors, investigators havetended to assess single-indicator predictors at onetime point (e.g., childhood) and use this informationto forecast single-indicator criteria at one later timepoint (e.g., preadolescence or adolescence). Designssuch as these do not enable investigators to assesspatterns of continuity or change in predictor – criteriarelations across successive intervals in children’sdevelopment (e.g., age, grade levels), evaluate orcontrol for the stability of predictors or criteria whenexamining predictive associations (e.g., controllingfor current predictor – criteria associations when ex-amining antecedent predictor-criteria linkages),draw clear distinctions between constructs, or max-imize the psychometric properties of their measures(e.g., reliability, construct invariance over time, etc.).

Certain design problems must be avoided as well.There is a need to both distinguish among (ratherthan confound) the predictive contributions of be-havioral and relational risk factors, and ensure thatrisk factors are not imbued with information aboutthe to-be-predicted maladjustment criteria. To illus-trate, when investigators attempt to estimate ad-justment consequences for certain ‘‘subtypes’’ ofchildren, such as those who manifest multiple be-havioral and relational risks (e.g., aggressive – re-jected children) or those who exhibit a behavioralrisk and symptoms of maladjustment (e.g., anxious –withdrawn, depressed – withdrawn children), theyessentially confound (likely overestimate) the pre-dictive contributions of specific risk factors.

To address these limitations, a full-panel, follow-through longitudinal design was implemented inthis investigation. Multiple indicators of each riskfactor and maladjustment criterion were obtainedyearly across a 7-year period, and used to formulatelatent variables for each construct at each time ofassessment. This measurement model was utilizedso that all of the risk and maladjustment constructswere operationalized as separate latent variables(rather than confounding them, as occurs in somerisk-subtype designs) and the hypothesized predic-tor – criteria linkages could be examined within

latent stability models across six (lagged) 1-yearintervals. One set of specific aims was to determine(1) whether, and when (i.e., how early) during thegrade school years, exposure to peer group rejectionforecasts later externalizing or internalizing prob-lems, beyond that which can be predicted by chil-dren’s aggressive or withdrawn behavioralpropensities, and (2) whether the additive contribu-tion of peer group rejection to the prediction of ex-ternalizing or internalizing problems (relative tobehavioral styles) remains constant versus growsstronger or weaker across the grade school years(determine whether there is continuity vs. disconti-nuity in its additive, predictive contributions). Il-lustrations of the type of additive, latent stabilitymodels that was used to address these objectives aredepicted in Figure 1. The scientific merit of thefindings was expected to be substantial becauseneither of these aims has previously been examinedwith this age group, and the results would reflect onpast and contemporary assumptions about (a) when,during childhood, peer group rejection (and its at-tendant stressors) creates adjustment consequencesthat are distinct from those attributable to children’sbehavioral styles and (b) the strength and continuitywith which this form of peer adversity predisposeschildren to externalizing versus internalizing formsof dysfunction across early, middle, and later child-hood.

It has been proposed that attaining acceptance inpeer groups is a critical social task for children asthey enter grade school and progress through theprimary grades, and that failure to do so (i.e., peergroup rejection) creates sustained risk for psycho-logical difficulties. It is during the early grade schoolyears that children are challenged to establishthemselves (become accepted) as members of large,age-segregated peer groups (Ladd, 1989, 1996). Fur-thermore, in addition to being a key psychosocialtask, acceptance into peer society can be seen as apowerful social need that emerges during the 6 to 9-year age period (Buhrmester & Furman, 1986). Onthe basis of these rationales, it was hypothesized thatpeer group rejection would additively contribute tothe prediction of maladjustment as early as kinder-garten (children’s 1st year in grade school), and thatcontinuity would be apparent in this additive linkageacross the primary school years (childhood; ages 5 –9) and, to a lesser extent, across the later grade schoolyears (preadolescence; ages 10 – 12). Moreover, con-sonant with extant empirical evidence, it was antic-ipated that the strength and continuity of theseadditive contributions would be greater for inter-nalizing than for externalizing problems.


Aim 2: Evaluate Alternatives to Additive PredictionModels

Knowledge may also be incomplete because it hasnot been a priority to validate the prevailing pre-dictive paradigm (e.g., ‘‘additive risk models’’)against alternative models that stem from differentcausal premises. A predominant assumption withinthis area of inquiry has been that children’s behaviorand/or peer group rejection are determinants ofpsychological maladjustment. This theoretical he-gemony has dissuaded investigators from consider-ing, or empirically testing, models that implyalternative directions of prediction. Two such modelswere identified and targeted for investigation. First,it is conceivable that premorbid forms of a later-emerging disorder may predict children’s behavioralpropensities, peer group rejection, or both (seeParker & Asher, 1987). Second, it is possible that thecausal relations that exist among these variables maynot be temporally invariant, but rather shift as afunction of age-related changes in organismic factorsand environmental circumstances (see Sameroff,

1987; Sameroff & Chandler, 1975). Were this the case,permutations in the predictive relations amongchildren’s behavioral dispositions, classroom peerrelations, and psychological adjustment would beexpected to emerge over time, and these patternswould be better described by transactional ratherthan unidirectional models (see Parker et al., 1995).Accordingly, a second set of specific aims was todetermine whether there was empirical support forthese alternate models, which for ease of referencewere termed ‘‘disorder-driven’’ and ‘‘transactional’’prediction models, respectively.

The central premise of the disorder-driven modelis that early-emerging dysfunctionsFin this case,incipient externalizing or internalizing prob-lemsFcause children to act in aggressive or with-drawn ways and to alienate their classmates. Forexample, in classroom contexts, it is likely that chil-dren who manifest externalizing problems (a generalpattern of emotional/behavioral undercontrol; e.g.,angry, explosive, impulsive, distractible, overactivebehaviors, etc.) will devise forceful tactics or reactaggressively when confronted with peer problems or

EXT EXT EXT EXT EXT EXT EXT

AG AG AG AG AG AG AG

REJ REJ REJ REJ REJ REJ REJ

Kindergarten Grade 1 Grade 2 Grade 3 Grade 4 Grade 5 Grade 6

INT INT INT INT INT INT INT

WD WD WD WD WD WD WD



Figure 1. The upper and lower panels of this figure illustrate the types of additive prediction models that were evaluated in this inves-tigation. The model shown in the upper panel depicts aggressive behavior and peer rejection as additive predictors of externalizingproblems (AG1REJ model) across each of six targeted time lags. The model shown in the lower panel depicts withdrawn behavior andpeer rejection as additive predictors of internalizing problems (WD1REJ model) across each of six targeted time lags. AG 5 aggressivebehavior; WD 5 withdrawn behavior; REJ 5 peer rejection.

826 Ladd

interpersonal constraints. Moreover, children whounder-regulate their behavior or emotions may, dueto their impulsiveness and insensitivity towardothers, irritate or alienate peers and, thus, becomedisliked and rejected by peers (see Figure 2, upperpanel). Likewise, children’s early internalizingproblems may cause them to act in an anxious –withdrawn manner among peers (emit anxious cues,avoid peers), and their lack of social responsiveness(e.g., reservedness, diffidence, melancholy) may en-gender peer dislike or rejection (see Figure 2, lowerpanel). However, because externalizing problemsand peer aggression may stem from common un-derpinnings and therefore partially overlap as con-structs (i.e., externalizing disorder encompassessome forms of aggression, particularly confrontive,angry, explosive, reactive subtypes), empirical sup-port for disorder-driven models was expected to bestronger for children disposed toward early exter-nalizing rather than internalizing problems.

For purposes of this investigation, the concept of atransactional model was based on the principle that,temporally, the relations that transpire between risk

factors and dysfunction are dynamic rather thanstatic and unidirectionalFthat is, subject to ‘‘turna-bouts’’ or reversals in causal priority or directions ofeffect (Rutter, 1996). The two transactional predictivemodels evaluated in this investigation were based onthe premise that risk factors and maladjustment in-fluence each other cyclically over time (Ladd, 1989).The first of these models constitutes a synthesis (se-quential paring) of the ‘‘additive risk’’ and ‘‘disor-der-driven’’ models, respectively (see Figure 3,upper panel). To illustrate, children who initiallysucceed in addressing classroom social challengeswith aggression and are repeatedly subjected topeers’ rejecting behaviors may generalize their use offorce toward broader social contingencies and be-come desensitized to various forms of social censure(i.e., develop pervasive, more serious forms of ex-ternalizing problems). In turn, movement into thisform of dysfunction (i.e., acting/reacting impul-sively, pervasively violating rules and norms, mini-mizing or disregarding broader social consequences)may become the proximal cause for later aggression(an increasing reliance on the use of force) and peer









Figure 2. The upper and lower panels of this figure provide illustrations of the disorder-driven models that were evaluated in this in-vestigation. The model shown in the upper panel depicts externalizing problems as a predictor of aggression and rejection across each ofsix targeted time lags. The model shown in the lower panel depicts internalizing problems as a predictor of withdrawn behavior and peerrejection across each of six targeted time lags.


group rejection (acting in ways that generate nega-tive peer sentiments such as dislike, distrust, wari-ness, fearfulness, etc.; see Coie, 2004). Corroborationfor this ‘‘risk ! maladjustment ! risk’’ modelwould be obtained if observed predictive linkagesconformed to a pattern in which children’s behav-ioral orientations and peer group rejection duringearly grade school anteceded externalizing or inter-nalizing problems during middle childhood and,thereafter, the dysfunction forecasted children’s be-havioral propensities and rejection by peers.

The second of these transactional models depicts acycle in which the direction of prediction for thedisorder-driven model precedes that specified forthe additive risk model (i.e., a ‘‘maladjust-ment ! risk ! maladjustment’’ model; see Figure 3,lower panel). As an example, it may be the case thatchildren’s initial internalizing tendencies (e.g.,emitting anxious social signals, being emotionallyunresponsive to peers) cause them to act in a with-drawn manner and be rejected by peers, but it is thedevelopment of withdrawn behavior and the expe-rience of being rejected (exposure to peer maltreat-ment, and the accompanying emotional scars or

debilitating cognitions children acquire from it; seePomerantz & Rudolph, 2003) that subsequentlydrives the development of later internalizing prob-lems (i.e., predictive relations are congruent with adiathesis-stress interpretation; e.g., isolation frompeers and rejecting peer reactions alienate childrenand confirm their anxieties about social rejection,and thus perpetuate or increase their wariness,fearfulness, propensity toward depression, etc.).Support for this maladjustment ! risk ! malad-justment model would be obtained if evidence fit apattern in which early externalizing or internalizingproblems forecasted peer group rejection and/orchildren’s behavioral orientations in middle child-hood but, thereafter, one or both of these risk factorsbecame the principal antecedents of maladjustmentin later childhood.

Method

Participants

Participants were 399 children (206 boys; 193 girls)who were recruited as they entered kindergarten









Figure 3. The upper and lower panels of this figure provide illustrations of the transactional prediction models that were evaluated in thisinvestigation. The risk ! maladjustment ! risk model is depicted in the upper panel and the maladjustment ! risk ! maladjustmentmodel is shown in the lower panel.

828 Ladd

(average age 5 5.62 years) and followed prospec-tively until they completed sixth grade (averageage 5 11.71 years). Children and their parents wererecruited at kindergarten preregistration meetingsthat were held in school systems within multipleurban, suburban, and rural U.S. locations. Writteninformed parental consent was obtained beforechildren’s participation, and 95% of the recruitedfamilies agreed to participate. The sample containedchildren from different ethnic groups (77.4% Euro-pean American, 17.3% African American, 5.3% La-tina, mixed race, or other) and family socioeconomicbackgrounds (36.8% lower to middle income, range 5

$0 – 20,000; 30.6% middle income, range 5 $21,000 –40,000; and 32.6% upper-middle to high income;range 5 $41,000 and higher). At the time of the sixth-grade assessment, 96.49% of the children were stillactive participants in the longitudinal study(n 5 385). Data were also collected from children’steachers and classmates.

Over the course of this investigation, childrenbecame increasingly dispersed across classroomsand school districts. When children moved or chan-ged schools, permission was sought from adminis-trators, teachers, and parents to extend the projectinto their schools. After these consents were ob-tained, the parents of project children’s classmateswere contacted and asked to provide consent fortheir child’s participation. Only those classmateswho had written informed parental consent (andchild assent) participated in the study. The numberof teachers who participated in this study per as-sessment period ranged from 32 to 282, and thenumber of classmates who contributed data per as-sessment ranged from 964 to 3,030.

Measures

Multiple indicators of the targeted constructs wereobtained. To increase the probability that these in-dicators could be used to form latent variables thatwere invariant across measurement occasions, iden-tical measures were administered at each time ofassessment. Indicators were measured with instru-ments that are known to yield reliable and valid data.

Behavioral Orientations

Confrontive aggression. Indicators of this constructincluded two peer nomination measures (one forverbal and another for physical aggression) and oneteacher report measure of children’s aggressive be-havior toward peers. The reliability and validity ofall three measures have been substantiated (see Ladd

& Profilet, 1996). Peer nominations of verbal ag-gression (PN-VAG) and physical aggression (PN-PAG) were obtained by presenting children with aroster containing classmates’ names (those withpermission) and asking them to circle the names ofup to three classmates who could be described byeach of the following criteria, respectively (order ofadministration counterbalanced): (1) ‘‘Someone whoteases, calls names, or makes fun of other kids’’ (i.e.,verbal aggression toward peers) and (2) ‘‘Someonewho hits, kicks, or pushes other kids’’ (i.e., physicalaggression toward peers). For children in kinder-garten to grade 2, a felt board containing classmates’pictures was used instead of a roster and, duringindividual interviews, respondents were asked todesignate (by pointing) up to three nominees. Beforeadministering these measures, children were pre-sented with class rosters/picture boards, tested untilthey could identify all classmates, and trained tomake nominations contingent on a series of practicecriteria. The third measure of aggression (CBS-AG)was obtained by asking teachers to rate, on a 3-pointscale (1 5 doesn’t apply; 2 5 applies sometimes; 3 5 cer-tainly applies), each of the seven-items included in theAggressive with Peers subscale of the Child BehaviorScale (CBS; Ladd & Profilet, 1996). Items on thissubscale describe children’s use of verbal andphysical aggression against peers, and as acrosstesting occasions ranged from .90 to .93.

Withdrawn, asocial behavior. Two teacher-ratingmeasures served as indicators of this construct, in-cluding the Withdrawn Behavior subscale of theAchenbach Teacher Report Form (TRF-WD; Achen-bach, 1991) and the Asocial Behavior Subscale of theCBS (CBS-AB; Ladd & Profilet, 1996). Items on theTRF-WD and CBS-AB indexed children’s tendencyto engage in shy or solitary behaviors, avoid orwithdraw from peer activities, and abstain from so-cial overtures (e.g., talking to peers). One item fromthe TRF-WD subscale (i.e., ‘‘sad’’) was eliminateddue to its semantic similarity with indicators selectedto tap internalizing problems (anxiety/depression,see below). The remaining eight TRF-WD items (e.g.,withdrawn, shy, rather be alone) were rated on a 3-point scale (e.g., 0 5 not true; 1 5 sometimes true;2 5 often true), as were the six items included in theCBS-AB subscale (e.g., solitary child; keeps peers at adistance, likes to play alone; 1 5 doesn’t apply;2 5 applies sometimes; 3 5 certainly applies). Internalconsistency was acceptable across measurement oc-casions for both measures (range 5 .90 – .93 and .87 –.92, respectively).

Although it was of interest to obtain indicators ofthis construct from the same informants used to tap


aggressive behavior (i.e., peers and teachers), thiswas infeasible because peers appear to lack thenecessary schema to distinguish and report this formof behavior reliably until middle childhood (ap-proximately 8 years of age; see Ladd & Mars, 1986;Younger & Boyko, 1987; Younger, Gentile, & Burgess,1993). However, for prediction purposes, problemsassociated with shared-informant bias were reducedbecause different teachers rated children on thesemeasures at each time of assessment. Further, as acheck against the validity of these indicators, scoresfor the TRF-WD and CBS-AB were correlated withscores from a peer nomination measure of children’swithdrawn, asocial behavior (PN-ASOC) that wereobtained during grades 3 – 6. The procedures used toobtain this index were the same as those describedfor the PN-VAG and PN-PAG measures, except thatpeers responded to the criterion ‘‘Someone who isalone a lot.’’ Moderate convergence was found be-tween the peer- and the teacher-report measures.Correlations (all po.01) ranged from .37 to .53 be-tween the PN-ASOC and TRF-WD, and from .30 to.48 between the PN-ASOC and CBS-AB.

Peer Group Rejection

The three measures used as indicators of thisconstruct were obtained from two types of inform-ants: peers and teachers. Peer reports of children’sgroup acceptance/rejection were obtained with ros-ter-and-rating and peer nomination sociometricmeasures (see Asher & Dodge, 1986; Asher, Single-ton, Tinsley, & Hymel, 1979). The roster-and-ratingmeasure (RR-SMS) was administered in each class-room by asking children (those with permission) torate each of their classmates on the following crite-rion: ‘‘How much do you like to play with this per-son in school?’’ At all grades, ratings were made on a3-point scale, with 1 5 not much, 2 5 sometimes, and3 5 a lot (for details, see Asher et al., 1979). For chil-dren in kindergarten to grade 2, a felt board con-taining classmates’ pictures was used instead of aclass roster, and respondents rated classmates duringindividual interviews by sorting their pictures intoboxes marked (in text and graphics) with the same 3scale points. A second measure of peer group rejec-tion was acquired via a peer nomination procedure(PN-NEG) in which children nominated up to threeclassmates (i.e., ‘‘Kids who you don’t like to hang outwith [play with] at school’’) by circling names on aroster (pointing to individual photos). The ratingsand the nominations children received from class-mates were separately averaged and standardizedwithin classrooms. It has been established that the

RR-SMS and PN-NEG measures yield reliable andvalid data with both younger and older children (seeAsher et al., 1979; Cillessen & Bukowski, 2000; Laddet al., 2006). The third measure of peer group rejec-tion was obtained by asking teachers to rate partici-pants on the seven-item Excluded by Peers subscaleof the CBS (CBS-EP). Peer exclusion constitutes onemanifestation of peer group rejection (i.e., peers’ at-tempts to exclude children from participation inclassroom social activities; see Buhs & Ladd, 2001)that can be reliably observed by teachers in class-room settings and that has been shown to correlatesubstantially with peer report indices of peer grouprejection (see Buhs & Ladd, 2001; Ladd & Profilet,1996).

Psychological Maladjustment

Externalizing problems. Indicators of this broad-band construct consisted of three forms of childmisconductFclassroom disruptiveness, hyperac-tive – distractible behavior, and ‘‘delinquent’’ behav-iorFall of which were measured at each time ofassessment. Collectively, these measures were seenas indicators of children’s propensity to respond to abroad range of classroom contingencies in an emo-tionally and behaviorally impulsive or undercon-trolled manner. To acquire a measure of the firstindicator, classroom disruptiveness (CD), teachersrated children on five-items taken from the TeacherRating Scale of School Adjustment (TRSSA; seeLadd, Buhs, & Seid, 2000; e.g., ‘‘Has disciplineproblems’’; ‘‘Is easy for teacher to manage (reversescored)’’; ‘‘Needs constant supervision’’), each ofwhich was accompanied by a 3-point scale(1 5 doesn’t apply; 2 5 applies sometimes; 3 5 certainlyapplies). To tap the second indicator, teachers ratedeach of four items on the CBS Hyperactive – Dis-tractible subscale (CBS-HD; e.g., ‘‘Restless, doesn’tkeep still’’; ‘‘Poor concentration’’) on a 3-point scale(1 5 doesn’t apply; 2 5 applies sometimes; 3 5 certainlyapplies). The nine-item TRF Delinquent Behaviorsubscale (TRF-DB) was used to tap the third indica-tor. Teachers rated these items (‘‘Lying or cheating’’;‘‘Steals’’) on a 3-point scale (e.g., 1 5 not true;2 5 sometimes true; 3 5 very or often true). Scores foreach measure were created by averaging teachers’ratings across the items within each subscale. ascalculated for these measures across assessment oc-casions fell within the following ranges: .79 – .84(CD), .84 – .91(CBS-HD), and .68 – .78 (TRF-DB).

Internalizing problems. Indicators for this broad-band construct included measures of anxiety anddepression (i.e., emotional/behavioral manifesta-

830 Ladd

tions of anxious and sad affect). Teachers rated eachchild on 17 items from the TRF Anxious/Depressedsubscale (TRF-AD; one item was deleted due tocontent that overlapped with peer group rejection)using a 3-point scale (e.g., 0 5 not true; 1 5 sometimestrue; 2 5 very or often true). as for the TRF-AD rangedfrom .82 to .89 across all times of assessment.Teachers also rated children on the 4-item (e.g.,‘‘Appears miserable, distressed’’; Fearful or afraid’’)CBS Anxious Fearful subscale (CBS-AF) using a 3-point scale (1 5 doesn’t apply; 2 5 applies sometimes;3 5 certainly applies). Scores for the CBS-AF areknown to be reliable (see Ladd & Profilet, 1996; Ladd& Troop-Gordon, 2003) and yielded as that rangedfrom .69 to .78 with this sample.

Procedure

Peers and teachers completed each of the behav-ioral orientation, peer group rejection, and psycho-logical maladjustment measures each year, for aperiod of 7 consecutive years (from kindergarten tograde 6). Trained examiners administered thesemeasures in children’s and teachers’ classroomsduring the spring of each school year (i.e., approxi-mately March through May). Details about the exactprocedures used to administer each measure are re-ported in the Measures section. During each assess-ment interval, children were trained, and teachersreceived instructions about how to use the ratingscales that accompanied the items contained in eachof their assigned measures. Upon completion of theirassigned measures, children and their classmateswere thanked and given a small gift (e.g., pencil,stickers), and teachers received a cash honorarium.

Results

Data Analytic Strategy and Results of PreliminaryAnalyses

Preliminary analyses were conducted to examinevariable distributions, relations, and patterns ofmissing data, and to determine whether these andother properties of the data set conformed to theassumptions underlying parametric statistics anddata augmentation proceduresFspecifically, multi-ple imputation (e.g., values missing at random;variables normally distributed, etc.; Rubin, 1987;Schafer, 1999; Schafer & Olsen, 1998). Because it waspossible to satisfy these prerequisites, missing valueswere replaced with estimated scores, and additionalpreparatory analysis was undertaken. Multicolline-arity was evaluated by examining correlations

among predictors or criteria. Although substantialcovariance was evident among some predictors (e.g.,indicators of aggressive behavior and peer rejection),it was not of a magnitude that constituted seriousmulticollinearity (e.g., destabilized matrix inver-sion). Measurement models were initially evaluatedto determine whether the hypothesized alignmentsof construct indicators and corresponding latentvariables fit the data well, and whether there wasevidence of factorial invariance or constancy in therelations of indicators to constructs over time.Structural equations modeling was then utilized toexamine the fit between the data and the proposed(i.e., hypothesized and alternative) prediction mod-els. Two criteria were used to judge model fit (see Hu& Bentler, 1999): the root mean square error of ap-proximation (RMSEA � .08) and the standardizedroot mean square residual (SRMRo.08). CFI was notutilized because it tends to be a conservative fit in-dicator when measurement models are complex. Allestimated models were compared by gender usingmultisample SEM analyses, and the same analyticstrategy was used to compare model fit for the entiresample versus a reduced sample in which a smallnumber (n 5 4) of dual behavioral risk children (i.e.,those with scores41.00 SD on the averaged indicatorscores for both aggressive and withdrawn behavioracross four or more assessment occasions) had beeneliminated. Because there was no instance in whichmodel fit differed significantly by gender or by riskstatus, the full sample was utilized in all analyses.

Estimation of Missing Values Via Multiple Imputation

Attrition was minimal, with 385 (96.49%) of theoriginal 399 participants remaining in the samplefrom kindergarten to grade 6. Constraints on datacollection (e.g., absences, moves, dropouts, etc.)caused 3.49% of the data points within the entirelongitudinal data set to be missing. The percentageof missing data by measure across all times of ad-ministration averaged 3.52% (range 5 0.00 – 9.00%).

Missing data were estimated via NORM (Schafer,1999), a multiple imputation program in whichmissing multivariate data are simulated m41 times.Three imputations were generated (n 5 3) because,for data sets containing 10% or fewer missing datapoints, it has been established that n 5 3 imputationsyield sufficient estimation efficiency (i.e., 97%; Ru-bin, 1987; Schafer, 1999). Mplus 3.01 (Muthen &Muthen, 1998 – 2004) was used to estimate each ofthe hypothesized models from the multiple imputeddata sets, and to average parameter estimates andobtain combined standard errors.


Formation of Latent Variables and Factorial Invariance

The multiple prediction models that were stipu-lated for this investigation necessitated the creationof five latent variables: confrontive aggression,withdrawn behavior, peer group rejection, internal-izing problems, and externalizing problems. Eachlatent variable was constructed from its manifestindicators (i.e., see measures listed under corre-sponding constructs in the Method section). For ex-ample, the latent confrontive aggression measurewas formed from three indicators: the PN-VAG, PN-PAG, and CBS-AG measures. The coherence betweenindicators and their respective latent variables wasevaluated as part of the SEM analyses that wereconducted for each prediction model; without ex-ception, indicators loaded only on the latent varia-bles that they were intended to represent (seemeasurement model results in subsequent sections).

To determine whether the factor structure for eachlatent variable was invariant over time, five multi-sample CFAs were calculated. In each of theseanalyses, grade level was the grouping variable andthe measurement model contained a single latentvariable that was constructed from its correspondingindicators. Using a marker variable strategy, theunstandardized loading for one indicator (per factor)was fixed at 1.0, and all remaining factor loadingswere freely estimated, as were error covariances be-tween identical indicators across measurement oc-casions (Rensvold & Cheung, 2001). The results ofthese analyses (ranges for fit statistics: w2/df 5 1.44 –3.85; RMSEA 5 .03 – .08; SRMR 5 .03 – .07) supportedthe assumption that each latent variable’s indicatorsformed a single factor that was invariant acrossmeasurement occasions.

Stabilities of the Latent Variables

SEM analyses were also used to estimate the sta-bility of each latent variable across the seven as-sessment occasions. Each of these analyses wasconducted on a model in which the same latentvariable was estimated on each of seven measure-ment occasions from its corresponding indicators.Factor loadings were freely estimated, as were errorcovariances for identical indicators across measure-ment occasions. The results are shown by constructin Figure 4. Of the five latent variables, aggressionand peer rejection were the most stable from kin-dergarten to grade 6; the stability coefficients forthese variables were substantial at lag 1 and re-mained high across all subsequent lags. The with-drawn behavior and externalizing latent variables

exhibited a pattern of increasing stability over time;the stability coefficients for these variables were lowto moderate in magnitude during the early grades(across lags 1 – 2) but became progressively strongerthereafter (e.g., across lags 3 – 6). Only modest sta-bility was found for internalizing problems and, overtime (lags 1 – 6), the stability coefficients for this la-tent variable increased only slightly.

Estimation of Additive Continuity Models

The hypothesis that each of two variables makes adistinct (additive) predictive contribution to a crite-rion rests on the assumption that both variables are,independently, predictors of that criterion. Thus, as aprecondition to estimating the proposed additivemodels, six ‘‘main effects’’ models were tested viaSEM to determine whether each of the latent behav-ioral and peer rejection variables, when evaluatedindependently, were significant predictors of the latentexternalizing or internalizing variables across one ormore of the six lagged, 1-year intervals. These anal-yses were conducted on a model that contained onelatent predictor variable measured on seven occa-sions (either aggressive behavior, withdrawn be-havior, or peer rejection at each of the seven gradelevels) and one latent criterion variable measured onseven occasions (either externalizing or internalizingbehavior at each of the 7 grade levels). Each of thelatent variables included in these analyses was esti-mated from its corresponding indicators. Factorloadings for latent variables were freely estimated, aswere error covariances for matching indicatorsacross measurement occasions and same-informantindicators within measurement occasions (see Cole& Maxwell, 2003). For example, in the analysis whereaggressive behavior predicted externalizing prob-lems, structural paths were estimated across each ofsix lagged, 1-year intervalsFthat is, from the latentaggressive behavior predictor measured at one gradelevel to latent externalizing problems criterion meas-ured at the next grade level (e.g., AG-K ! EXT-G1;AG-G1 ! EXT-G2; AG-G2 ! EXT-G3; AG-G3 !EXT-G4; AG-G4 ! EXT-G5; & AG-G5 ! EXT-G6FK, kindergarten; G1, grade 1).

For externalizing problems, the model in whichaggression was the sole predictor fit the data ade-quately, w2(721) 5 2,362.04, po.01; RMSEA 5 .08,SRMR 5 .07, as did the model in which peer rejectionwas a lone predictor, w2(721) 5 1,922.32, po.01;RMSEA 5 .06, SRMR 5 .07. Positive and statisticallysignificant path coefficients were found from ag-gression to externalizing problems, and theseweights increased in magnitude across the six lagged

832 Ladd

intervals (e.g., .67, .73, .78, .79, .80, .89, respectively).The path weights from peer rejection to externalizingproblems were also positive and significant, butthese coefficients declined in magnitude across thelatter of the six lags (i.e., .73, .70, .74, .68, .54, .55,respectively). In contrast, the model for withdrawnbehavior did not fit the data, w2(487) 5 2,207.40,po.01; RMSEA 5 .10, SRMR 5 .18, and none of thepath coefficients were significant.

For internalizing problems, model fit was ade-quate for the withdrawn model, w2(279) 5 795.27,po.01; RMSEA 5 .07, SRMR 5 .07, and for the peerrejection model, w2(487) 5 1,176.78, po.01;RMSEA 5 .06, SRMR 5 .07. All but the first of the sixpath coefficients from withdrawn behavior to inter-nalizing problems were significant, and theseweights increased in magnitude across the six lags(.09, .15, .19, .31, .26, .40, respectively). Similarly, allpath coefficients from peer rejection to internalizingproblems were significant and tended to increase inmagnitude across the six lags (.14, .22, .43, .38, .32,.42, respectively). Although the aggression modelapproximated a fit to the data, w2(487) 5 1,302.18,

po.01; RMSEA 5 .07, SRMR 5 .10, none of the pathweights from aggression to internalizing problemswere significant.

These findings indicated that it was feasible toevaluate two of the hypothesized additive models.The first of these models (i.e., the AG1REJ additivemodel) provided a test of the hypothesis that, inaddition to children’s propensity to engage in ag-gressive behavior, peer group rejection contributesadditively to the prediction of externalizing prob-lems. The second model (i.e., the WD1REJ additivemodel) addressed the hypothesis that beyond chil-dren’s propensity to engage in asocial – withdrawnbehavior, peer group rejection contributes additivelyto internalizing problems. In each case, it was ofinterest to determine (1) whether peer rejectioncontributed additively to the prediction of malad-justment, (2) when during childhood such contribu-tions were evident, and (3) how the observedcontributions differed in strength and continuityacross grade levels (i.e., the six lagged, 1-year inter-vals). Because results from the main effects analysesdid not implicate withdrawn behavior as a signi-







.90 .82 .95 .91 .93 .93

.38 .60 .48 .56 .56 .31

.87 .93 .82 .73 .95 .91

.26

.67 .87 .63 .78 .83 .91

.27 .40 .37 .27 .33

Figure 4. Stability coefficients are shown for each of the six latent variables across each of the six lags that were examined in this inves-tigation. AG 5 aggressive behavior; WD 5 withdrawn behavior; REJ 5 peer rejection; EXT 5externalizing problems; INT 5 internalizingproblems.


ficant predictor of externalizing problems, noraggressive behavior as a predictor of internalizingproblems, the predictive contributions of peer rejec-tion were not estimated in conjunction with theselatent variables.

Both the measurement and structural componentsof the AG1REJ and WD1REJ models are depictedgraphically in Figures 5 and 6. In the SEM analysesperformed on these models, the latent predictor andcriterion variables were estimated from their corre-sponding indicators. Factor loadings for all latentvariables were freely estimated, as were error co-variances for matching indicators across measure-ment occasions and for same-informant indicatorswithin measurement occasions.

AG1REJ additive model. Calculation of the corre-lations among the multiple indicators of the latentvariables included in this model (averaged acrossimputed data sets) yielded a very large number ofcoefficients (n 5 1,953). For the sake of brevity, theranges of coefficients obtained between predictors(AG & REJ) across all lags, and between the predic-tors and the criterion (AG & EXT; REJ & EXT) acrossall lags, are reported by measure in the left column ofTable 1. SEM findings (see Figure 5) showed that thismodel fit the data adequately, w2(1,636) 5 4,138.90,po.01; RMSEA 5 .06, SRMR 5 .06. Indicators loadedsignificantly (all ls po.01) on their designated latentvariables. Across each of the six lags (1-year inter-vals), significant positive path weights were foundnot only from aggression to externalizing problemsbut also from peer rejection to externalizing prob-lems (po.01). Thus, from kindergarten to grade 6,both aggression and peer rejection at one grade levelwere found to be distinct and significant predictorsof externalizing problems at the next grade level.There was evidence of continuity in the predictive(additive) contributions of peer group rejection toexternalizing problems in the sense that the pathcoefficients for peer rejection were significant (afteradjusting for aggression) across every grade levelfrom kindergarten to grade 6. Discontinuity wasobserved, however, in the sense that the magnitudesof these path coefficients varied across grade levels;of the weights obtained, those for the early primarygrades (particularly those spanning grades 1 – 2 and2 – 3) were among the strongest.

WD1REJ additive model. In this model (see Figure6), withdrawn behavior and peer rejection served aspredictors, and it was of interest to examine thecontinuity and strength of peer rejection as an ad-ditive predictor of internalizing problems across theinvestigated grade intervals. The ranges of the cor-relations obtained between predictors (WD & REJ)

across all lags, and between the predictors and thecriterion (WD & INT; REJ & INT) across all lags, arereported by measure in the right column of Table 1.SEM results indicated an adequate fit between thismodel and the data, w2(924) 5 2,213.76, po.01;RMSEA 5 .06, SRMR 5 .07. Measurement model re-sults showed that indicators loaded on their desig-nated latent variables (all ls po.01). Results for thestructural model revealed substantial continuity inthe predictive (additive) contributions of peer grouprejection to internalizing problems in the sense thatthe path coefficients obtained for this latent variablewere significant (after adjusting for asocial – with-drawn behavior) across all grade intervals exceptK ! G1. In contrast, the path weights for asocial –withdrawn behavior were significant only for thelast three grade intervals (i.e., G3! G4, G4! G5,and G5! G6), suggesting that this variable’s addi-tive contribution to the prediction of internalizingproblems was not substantial until later in child-hood. The importance of peer rejection as a distinctpredictor of internalizing problems was also illus-trated by the fact that, across each of the lagged in-tervals, the path weights obtained for this variablewere larger in magnitude than those obtained forasocial – withdrawn behavior.

Estimation of Disorder-Driven Models

Another series of SEM analyses were conducted toevaluate models founded on the premise that early-emerging maladjustment predicts children’s subse-quent propensities to engage in specific behaviors(aggressive or withdrawn) and be rejected by theirclassmates. Here again, main effects models wereinitially estimated to determine whether early ex-ternalizing problems or internalizing problems,when evaluated independently, were significantpredictors of each of the latent behavioral risk andpeer rejection variables across one or more of the sixlagged, 1-year intervals. Each of these analyses wasconducted on a model that contained one latentpredictor variable measured on seven occasions (ei-ther externalizing or internalizing behavior at each ofthe seven grade levels) and one latent criterion var-iable measured on seven occasions (either aggressivebehavior, withdrawn behavior, or peer rejection ateach of the seven grade levels). Each latent variablewas estimated from its corresponding indicators,and factor loadings for latent variables were freelyestimated, as were error covariances for matchingindicators across measurement occasions and same-informant indicators within measurement occasions.To illustrate, in the analysis where externalizing

834 Ladd

Lag

1

Lag

2

Lag

3

Lag

4

Lag

5

Lag

6

Kin

derg

arte

n G

rade

1

Gra

de 2

G

rade

3

Gra

de 4

G

rade

5

Gra

de 6

AG

K

RE

J 1

AG

2

AG

3

RE

J 4

RE

J 5

RE

J 6

AG

1

AG

6

AG

5

AG

4

RE

JK

RE

J 2

RE

J 3

EX

TK

EX

T1

EX

T4

EX

T3

EX

T2

EX

T5

EX

T6

.47

.27

.68

.64

.56

.23

.18

.20

2

2

3 2

2 1

3 1

3 2

1 3

2 1

3 1

3 2

1 2

3

3

3 1

2

1 3

1 2

2 3

1

1

.47

3 2

1 3

2 1

3 2

1 3

2 1

3 2

1

.72

.76

.36

2 2

2 2

2 3

3 3

3 3

1 1

1 1

1

.85

.68

.90

.72

.71

.79

.76

.65

.78

.87

.82

.75

.82

.79

.77

.87

.85

.76

.92

.85

.71

.83

–.76

.6

9

.76

.71

.72

–.77

.6

4 .6

6 –.

59

.63

.78

–.75

.6

7 .8

4

.68

.71

–.70

.6

8 .8

5 –.

82

.52

–.76

.8

7 .6

4

.82

.76

.74

.85

.76

.77

.78

.73

.88

.75

.84

.91

.74

.76

.88

.72

.73

Fig

ure

5.P

ath

dia

gra

mfo

rth

eA

G1

RE

Jad

dit

ive

mo

del

.A

G5

late

nt

var

iab

lefo

rag

gre

ssiv

eb

ehav

ior;

RE

J5la

ten

tv

aria

ble

for

pee

rre

ject

ion

;E

XT

5la

ten

tv

aria

ble

for

exte

rnal

izin

gp

rob

lem

s.In

dic

ato

rsfo

rla

ten

tv

aria

ble

s:A

G(1

:P

N-P

AG

;2:

PN

-VA

G;

3:C

BS

-AG

);R

EJ

(1:

PN

-NE

G;

2:R

R-S

MS

;3:

CB

S-E

P);

and

EX

T(1

:C

D;

2:T

RF

-DB

;3:

CB

S-H

D).

All

pat

hw

eig

hts

sho

wn

inth

ism

od

elw

ere

stat

isti

call

ysi

gn

ific

ant

(po

.01)

.


Lag

1

Lag

2

Lag

3

L

ag 4

L

ag 5

L

ag 6

Kin

derg

arte

nG

rade

1

Gra

de 2

G

rade

3

Gra

de 4

G

rade

5

Gra

de 6

WD

K

RE

J 1

WD

2

WD

3

RE

J 4

RE

J 5

RE

J 6

WD

1

WD

6

WD

5

WD

4

RE

JK

RE

J 2

RE

J 3

INT

KIN

T

1IN

T

4IN

T

3IN

T

2IN

T

5IN

T

6

.05

.11

.17

.13

.09

.28

.23

.32

2

2 2

2 1

1 2

1 2

1 1

2 1

2

3 1

2

1 1

2

2 3

1

1

.18

3 2

1 3

2 1

3 2

1 3

2 1

3 2

1

.16

.17

.37

2 2

2 2

2 1

1 1

1 1

.80

.84

.67

.75

.64

.84

.85

.80

.76

.87

.67

.73

.60

.91

.75

–.78

.7

5

.84

.86

–.80

.7

5 .7

3 –.

73

.67

.68

–.73

.7

0 .7

8 .8

3 –.

90

.58

.83

–.86

.5

6 –.

76

.76

.70

.83

.91

.88

.77

.98

.91

.83

.69

.91

.89

.84

.82

Fig

ure

6.P

ath

dia

gra

mfo

rth

eW

D1

RE

Jad

dit

ive

mo

del

.W

D5

late

nt

var

iab

lefo

rw

ith

dra

wn

beh

avio

r;R

EJ5

late

nt

var

iab

lefo

rp

eer

reje

ctio

n;

INT

5la

ten

tv

aria

ble

for

inte

rnal

izin

gp

rob

lem

s.In

dic

ato

rsfo

rla

ten

tv

aria

ble

s:W

D(1

:C

BS

-AB

;2:

TR

F-W

D);

RE

J(1

:P

N-N

EG

;2:

RR

-SM

S;

3:C

BS

-EP

);an

dIN

T(1

:T

RF

-AD

;2:

CB

S-A

F).

All

pat

hw

eig

hts

abo

ve

.16

wer

est

atis

tica

lly

sig

nif

ican

t(po

.01)

.

836 Ladd

problems were evaluated as predictors of aggressivebehavior, structural paths were estimated acrosseach of six lagged, 1-year intervalsFthat is, from thelatent externalizing predictor measured at one gradelevel to latent aggressive variable measured at thenext grade level (e.g., EXT-K ! AG-G1; EXT-G1 ! AG-G2; EXT-G2 ! AG-G3; EXT-G3 ! AG-G4; EXT-G4 ! AG-G5; & EXT-G5 ! AG-G6).

As the following fit statistics reveal, there was lessthan adequate coherence between the data and each ofthese six models: externalizing problems! aggres-sive behavior, w2(721) 5 2,821.55, po.01; RMSEA 5 .09,SRMR 5 .14; externalizing problems! withdrawnbehavior, w2(487) 5 1,976.28, po.01; RMSEA 5 .09,SRMR 5 .09; externalizing problems! peer rejec-tion, w2(721) 5 2,073.41, po.01; RMSEA 5 .07,SRMR 5 .14; internalizing problems! aggressivebehavior, w2(487) 5 1,679.58, po.01; RMSEA 5 .08,SRMR 5 .17; internalizing problems! withdrawn

behavior, w2(279) 5 1,106.77, po.01; RMSEA 5 .09,SRMR 5 .13; internalizing problems! peer rejection,w2(487) 5 1,973.89, po.01; RMSEA 5 .09, SRMR 5 .21.Because none of these models fit the data well, noattempt was made to evaluate larger disorder-drivenmodels such as those in which antecedent malad-justment variables predicted both children’s behavior(e.g., aggressive, withdrawn behavior) and peer re-jection across multiple grade intervals.

Estimation of Transactional Models

Two final SEM analyses were undertaken to de-termine whether the predictive relations observedbetween the investigated risk factors and dysfunc-tions followed a cyclical pattern that conformed toone of the two proposed transactional models. Bothof these models were estimated such that the direc-tion of effect between risk factors and maladjustment

Table 1

Range of Correlations Obtained Between Latent Variable Indicators Across Six Assessment Lags

AG1REJ model WD1REJ model

Aggressive behavior and peer rejection Withdrawn behavior and peer rejection

PN-PAG & RR-SMS � .21 to � .56 TRF-WD & RR-SMS � .12 to � .39

PN-PAG & PN-NEG .21 to .60 TRF-WD & PN-NEG .05 to .32

PN-PAG & CBS-EP .11 to .35 TRF-WD & CBS-EP .10 to .53

PN-VAG & RR-SMS � .19 to � .51 CBS-AB & RR-SMS � .08 to � .44

PN-VAG & PN-NEG .21 to .58 CBS-AB & PN-NEG .07 to .37

PN-VAG & CBS-EP .11 to .40 CBS-AB & CBS-EP .07 to .60

CBS-AG & RR-SMS � .23 to � .45

CBS-AG & PN-NEG .21 to .56

CBS-AG & CBS-EP .11 to .57

Aggressive behavior and externalizing problems Withdrawn behavior and internalizing problems

PN-PAG & CD .36 to .61

PN-PAG & CBS-HD .27 to .58 TRF-WD & TRF-AD .04 to .58

PN-PAG & TRF-DB .38 to .60 TRF-WD & CBS-AF .04 to .56

PN-VAG & CD .31 to .63 CBS-AB & TRF-AD .01 to .42

PN-VAG & CBS-HD .22 to .52 CBS-AB & CBS-AF .01 to .48

PN-VAG & TRF-DB .33 to .57

CBS-AG & CD .35 to .61

CBS-AG & CBS-HD .21 to .60

CBS-AG & TRF-DB .35 to .62

Peer rejection and externalizing problems Peer rejection and internalizing problems

RR-SMS & CD � .18 to � .54 RR-SMS & TRF-AD .01 to .26

RR-SMS & CBS-HD � .14 to � .56 RR-SMS & CBS-AF .01 to .29

RR-SMS & TRF-DB � .14 to � .44

PN-NEG & CD .19 to .55 PN-NEG & TRF-AD � .03 to � .24

PN-NEG & CBS-HD .20 to .53 PN-NEG & CBS-AF � .05 to � .28

PN-NEG & TRF-DB .18 to .56

CBS-EP & CD .13 to .53 CBS-EX & TRF-AD .03 to .52

CBS-EP & CBS-HD .16 to .50 CBS-EX & CBS-AF .05 to .51

CBS-EP & TRF-AD .16 to .46

AG 5 aggressive behavior; WD 5 withdrawn behavior; REJ 5 peer rejection.


were reversed beginning at grade 3 (at lag 4, asshown in the upper and lower panels of Figure 3).The first of these modelsFthe risk ! maladjust-ment ! risk model (see Figure 3, upper pan-el)Fcontained three latent variables, each of whichwas measured on seven occasions (at each of theseven grade levels). This model was estimated suchthat (a) in kindergarten to grade 2, the latent ag-gressive behavior and peer rejection variables thatwere measured at each of these grade levels pre-dicted the latent externalizing problems variable thatwas measured at the next grade level (each of thesestructural paths traversed a 1-year lag), and (b) ingrades 3 – 6, the direction of prediction reversed sothat the latent externalizing problems variables thatwere measured at each of these grade levels pre-dicted the latent aggressive behavior and peer re-jection variables that were measured at eachsubsequent grade level (each path traversed a 1-yearlag). The second of the two modelsFthe malad-justment ! risk ! maladjustment model (see Fig-ure 3, lower panel) was estimated as follows: (a) inkindergarten to grade 2, the latent internalizingproblems variable that was measured at each ofthese grade levels predicted the latent withdrawnbehavior and peer rejection variables that weremeasured at the next grade level (each of thesestructural paths traversed a 1-year lag), and (b) ingrades 3 – 6, the direction of prediction reversed sothat the latent withdrawn behavior and peer rejec-tion variables that were measured at each of thesegrade levels predicted the latent internalizing prob-lems variable that was measured at each subsequentgrade level (each path traversed a 1-year lag). In bothanalyses, the latent predictor and criterion variableswere estimated from their corresponding indicators.Factor loadings for all latent variables were freelyestimated, as were error covariances for matchingindicators across measurement occasions and forsame-informant indicators within measurement oc-casions. Results showed that neither the risk !maladjustment ! risk model, w2(1,664) 5 4,626.61,po.01; RMSEA 5 .07, SRMR 5 .11, nor the malad-justment ! risk ! maladjustment model, w2(1,672)5 5,099.54, po.01; RMSEA 5 .07, SRMR 5 .13, fit thedata adequately. Altering the structural paths suchthat the reversals occurred at earlier or later gradelevels did not improve the fit of either model.

Discussion

The results of this investigation advance extantknowledge in several novel, specific ways. Unlike

the evidence obtained in past investigations, the dataobtained from this study yielded a more compre-hensive, developmental portrait of the relations thatevolve among children’s behavioral orientations,peer rejection, and psychological maladjustmentacross the entire grade school period (ages 5 – 12). Inaddition, this study’s prospective longitudinal paneldesign permitted an evaluation of differing premisesabout the causal forces that may underlie the rela-tions that develop among the targeted risk andmaladjustment variables. By estimating traditionalmodels along with conceptually viable but previ-ously untested alternative models, it was possible todraw stronger inferences about the relative impor-tance (predictive significance) of peer behavioral andrelational risks in the development of childhoodpsychological maladjustment and vice versa.

At the broadest level of analysis, the combinedmodel evaluation results were consistent with a childand environment perspective, and implicated bothchild attributes (i.e., aggressive, withdrawn behav-ioral orientations) and features of the child’s peermilieu (i.e., peer rejection) as risk factors for psy-chological maladjustment (Ladd, 1989, 2003, 2005;McDougall et al., 2001; Parker & Asher, 1987; Parkeret al., 1995). More importantly, the results werecongruent with an additive child and environmentmodel (see Ladd, 2003). An additive model impliesthat maladjustment has multiple triggering mecha-nisms and accelerants (exacerbating factors) that,when concurrently operative, act as separate forces(complements) on the inception and development ofdysfunction. The fact that the data conformed to anadditive model is significant because, theoretically,these findings strengthen the view that peer rejectioncontributes to children’s maladjustment, and does soin a way that is at least partially independent of theirbehavioral orientations. That is, the findings areconsistent with the inference that, in addition toagentic forces (the child’s performance of aggressiveor withdrawn behaviors among peers), relationalinfluences (peers’ rejection of the child) shape earlypsychological maladjustment and its developmentalcourse. On the basis of this, it can be argued thatmaladjustment becomes more likely when either ofthese risk factors is present, and its probability ofoccurrence or severity is exacerbated when both areoperative. This is in contrast to several competingtheoretical perspectives (for a review, see Ladd,2003), including those in which it has been arguedthat (a) child (e.g., behavioral) rather than relationalforces, or vice versa, constitute the determinants ofmaladjustment (i.e., main effects models), (b) peerrejection is a consequence of deviant behavior rather

838 Ladd

than a cause of maladjustment (e.g., incidental, be-havior continuity models), (c) children’s behavioralstyles and peers’ rejecting reactions derive fromemergent dysfunction rather than vice versa (i.e.,disorder-driven models), and (d) child behavior,peer rejection, and maladjustment influence eachother and evolve cyclically over time (i.e., transac-tional models).

It must be recognized, however, that even thoughadditive models better represented the covariancestructures within this longitudinal data set than didthe investigated alternative models, such findings donot render implausible other hypothesized but un-tested effect patterns. Investigators have, for exam-ple, proposed models other than those tested here toexplain how behavioral and relational risks mightcombine to influence psychological maladjustment.Examples would include models in which one riskfactor’s effects on children’s maladjustment is hy-pothesized to be (a) contingent upon another varia-ble (i.e., moderator models) or (b) transmittedthrough one or more variables (i.e., mediator models;see Ladd, 2003, for further explication of thesemodels). Models such as these were not evaluated inthis study, however, because they could not be ope-rationalized or analyzed in a way that made themcomparable, from a sample size (power) or designperspective, to the investigated main effects, addi-tive, disorder-driven, and transactional models. Forexample, evaluation of a mediated model, using thisstudy’s panel design, would have required estimat-ing latent variable linkages across multiple ratherthan single time lags.

Similarly, interpretive certainty is curtailed bysampling constraints. It remains to be seen whetherthe reported data patterns, which were obtainedwith a normative community sample (one that wasfairly diverse with respect to gender, SES, ethnicity,and geographic locales), will replicate in other sam-ples, including ones that incorporate other culturalgroups, peer environments, and child populations.For example, stronger empirical support for disor-der-driven models might be found in clinical sam-ples, such as those composed entirely of childrenwho manifest diagnosable early-onset psychopa-thology. Conceivable, too, are differences in the rel-ative importance of additive risk factors acrosssamples. It is possible, for example, that the strengthof a particular child attribute as a predictor of mal-adjustment may vary as a function of the ‘‘norma-tiveness’’ of that characteristic within the child’s peerenvironment. The additive contribution of aggres-sion, for example, relative to peer rejection, might bediminished within peer environments where ag-

gression is widespread (e.g., exhibited by mostgroup members) or so frequent as to desensitizechildren to its occurrence and effects (see Wright,Giammarino, & Parad, 1986). For these reasons, itwill be necessary to undertake additional studies toachieve a fuller, more comprehensive understandingof how peer behavioral and relational risks antecedepsychological maladjustment.

In the meantime, however, the present findingsenhance the stature of additive modelsFthat is, theylend support to the premise that, in addition tospecific behavioral orientations, peer rejection in-creases the probability that children will becomemaladjusted. This inference, however, calls attentionto the question of why peer rejection forecasts or, ifconstrued as a determinant, ‘‘contributes’’ to dys-function in ways that are partially distinct from thoseattributable to children’s behaviors. Pertinent to thisissue is the contention that, even though children’sbehaviors contribute to their rejection (see Coie &Kupersmidt, 1983; Dodge, 1983), the subsequent es-tablishment of a rejecting peer milieu exposes chil-dren to other (i.e., nonshared) processes that fosterthe development of psychological dysfunction. Inother words, behaviors such as aggression may causepeer rejection, but once children come to occupy thissocial position (i.e., acquire rejected status), theirexperiences in the peer group may change in waysthat affect their adjustment. As Coie (2004) has ar-gued, children’s behavioral characteristics may havea pathogenic influence on their development, butpeer rejection has its own debilitating consequences.

Therefore, why is peer rejection debilitating tochildren’s psychological health in a way that is dis-tinct from their behavioral propensities? Insight intothis question can be found, in part, within emergentmetatheories where it has been proposed that peerrejection is not just a social ‘‘address’’ (e.g., a child’sposition or status in a peer group) but also a socialexperienceFone that exposes children to adverse orpainful interpersonal processes (see Bukowski &Hoza, 1989; Coie, 1990, 2004; Sandstrom & Zakriski,2004). For example, it has been hypothesized that theacquisition of rejected status (an awareness amongpeers that an individual is disliked) signals tomembers of the peer group that disliked individualsare suitable targets for maltreatment (see Buhs &Ladd, 2001; Coie, 1990). Coie (1990, 2004), for ex-ample, contends that once children are rejected,peers treat them more negatively, and that this abuseand the isolation and stigma it often produces aredistinct causes of maladjustment. In this way, peerrejection can be seen as having effects on childrenthat are at least partially distinct from those attrib-


utable to aggression or other behaviors that mayantecede rejection. Support for this contention hasbegun to accrue; evidence indicates that peers per-petrate many forms of abuse against rejected chil-dren (e.g., ignoring, exclusion, threats, mocking,verbal and physical abuse; see Asher et al., 2001) andthat peer maltreatment predicts psychological andschool maladjustment even after controlling for in-dividual differences in children’s aggressive andwithdrawn behaviors (see Buhs & Ladd, 2001; Ga-zelle & Ladd, 2003; Ladd, Birch, & Buhs, 1999; Ladd& Troop-Gordon, 2003).

In contrast, little evidentiary support was foundfor alternative or competing prediction models basedon the view that premorbid forms of psychologicaldysfunction foster the development of specific be-havioral propensities or peer group rejection (i.e.,disorder-driven models; see Parker & Asher, 1987),or that the investigated forms of risk and dysfunc-tion influence each other cyclically over time (i.e.,transactional models; see Parker et al., 1995). Col-lectively, the lack of unanimity between these modelsand the data, coupled with the stronger fit found foradditive models, implies that developments withinthe peer milieu (children’s behavior and rejectionamong age-mates) are more influential in shapingthe course of dysfunction than vice versa. Moreover,this implied direction of effect remained evidentover a 7-year period, suggesting that both peer be-haviors and relations continue to play an importantrole in the emergence and maintenance of psycho-logical dysfunction throughout childhood.

In part, the basis for the lesser efficacy of the dis-order-driven and transactional models was reflectedin the pattern of stabilities and main effect linkagesthat were observed among the investigated latentvariables. Consider, for example, the relative stabilityof the three latent variables that were included in theAG1REJ and WD1REJ additive models. Within eachmodel, the two latent peer predictors (i.e., aggressivebehavior and peer rejection, withdrawn behaviorand peer rejection, respectively) exhibited greaterstability across the six lagged, 1-year intervals thandid the latent maladjustment criterion (i.e., exter-nalizing or internalizing problems, respectively).This relative difference in stabilities was particularlyapparent during the early grade school years, sug-gesting that individual differences in maladjustmentwere subject to greater temporal variability (changeover time) than were differences in children’s behav-ioral orientations or peer rejection. This apparentinstability may reflect the fact that internalizing andexternalizing problems, when measured at early agesand in normative samples such as this one, constitute

incipient rather than mature forms of dysfunction. Ifso, it is likely that the symptoms, or outward ex-pressions of these dysfunctions, were few in number,sporadically expressed, and only moderately dis-crepant from age norms. For this reason, these dys-functionsFas they were initially manifestedFmaynot have constituted a force strong or consistent en-ough to induce moderate to severe behavioral devi-ance (e.g., aggressive or withdrawn behavior) or peercensure (e.g., peer group rejection). Additionally, re-sults from the main effects models revealed that,across each of the six lags, the predictive links be-tween both of the peer variables and the maladjust-ment criteria remained substantial (statisticallysignificant) or became stronger over time. Collec-tively, these findings are consistent with the inferencethat children’s behavioral orientations and peer re-jection anteceded emergent forms of dysfunction or,conversely, that manifestations of maladjustmentchanged over time so as to become increasinglypredictable from children’s behavior and social re-jection within the peer milieu.

Second, the fact that peer rejection increased theefficacy with which both externalizing and internal-izing problems could be predicted (beyond that at-tributable to differing behavioral orientations)suggests that this aspect of children’s peer experi-ence is a risk factor for both externalizing and in-ternalizing problemsFthat is, multiple forms ofpsychopathology. The findings also suggest, how-ever, that it would be imprudent to conclude that thestrength and pattern of these predictive contribu-tions were identical or invariant over time. Rather,examination of peer rejection in different variablecontexts (e.g., in conjunction with aggression as apredictor of externalizing problems, and in combi-nation with withdrawn behavior as a predictor ofinternalizing problems), and across repeated timeintervals throughout childhood (yearly, from kin-dergarten to grade 6), revealed differences in thetiming, strength, and consistency of peer rejection asa distinct predictor of externalizing versus internal-izing problems.

Results from the AG1REJ additive model cor-roborated extant evidence by indicating that peerrejection, as well as children’s aggressive behavior,presages the development of externalizing problems.More importantly, these findings expanded extantknowledge by showing that the distinct predictivelink between peer rejection and externalizing prob-lems (1) was significant and sustained across a sub-stantial period of child development, from childhoodthrough preadolescence, and (2) tended to bestronger during the early rather than the later grade

840 Ladd

school years. Thus, along with aggressive behavior,peer rejection was found to be an enduring risk fac-tor throughout childhood, but one that proved to bea stronger prognosticator of impending externalizingproblems during the early rather than later schoolyears. These age-related differences in predictionwere consistent with the hypothesis that peer rejec-tion during the primary grades (ages 5 – 9) causeschildren to resolve insipient social needs and tasks inways that orient them toward antisocial rather thannormative social and behavioral development. Onthe basis of Sullivanian premises, Buhrmester andFurman (1986) characterized this particular age pe-riod (i.e., the ‘‘juvenile era’’) as one in which childrenexperience a growing need for peer acceptance (in-clusion) and a heightened fear of peer rejection (ex-clusion). As these needs become prominent, peerrejection experiences (e.g., ignoring, exclusion, mal-treatment) may become an impetus for negativeemotions (e.g., anger at exclusion), forceful behaviors(e.g., attacks on peers, revenge seeking), rule viola-tions (e.g., rejecting norms established or followed bynonrejected peers), and other maladaptive reactionsthat lead to the development of externalizing prob-lems (see Coie, 2004; Kupersmidt & DeRosier, 2004).Similarly, it has been proposed that peer rejectionduring the early grades institutes thought and be-havior patterns (e.g., attributional biases, negativepeer beliefs, coercive cycles) that persist over timeand increase children’s propensity to engage inmisconduct and delinquency (see Coie, 2004; Ladd &Troop-Gordon, 2003; Sandstrom & Zakriski, 2004).

Within this same variable context, the efficacy ofaggressive behavior as a distinct predictor of exter-nalizing problems not only remained significant butalso increased in magnitude from childhood throughpreadolescence. Although not previously docu-mented longitudinally across this entire age period,perhaps it should not be surprising that aggressivebehavior became a stronger predictor of externaliz-ing problems over time. The current findings cor-roborated the view that children’s use of forceagainst age-mates is an enduring behavioral pro-pensity for both boys and girls (Cairns et al., 1989;Ladd, 2003; Olweus, 1979), and one that is likely tobe perpetuated by its social consequences (Caspi etal., 1987) and other mediating factors (e.g., aggres-sive children’s tendency to infer that forceful be-haviors are effective; see Boldizar, Perry, & Perry,1989; Crick & Ladd, 1990; Egan, Monson, & Perry,1998; Slaby & Guerra, 1988). Eventually, the use ofconfrontive aggression in the peer milieu may gen-eralize to a larger range of environmental contin-gencies and cause children to develop broader and

more serious forms of externalizing problems.Moreover, as children approach preadolescence, theyare less likely to be restricted by the contextualconstraints of self-contained classrooms and morelikely to engage in social niche seeking. It has beensubstantiated that as children mature, those withaggressive tendencies tend to seek the company ofother aggressive peers who, in turn, may socialize orencourage (e.g., condone) more frequent and seriousforms of externalizing behavior (Brown, Dolcini, &Leventhal, 1997; Cairns & Cairns, 1994; Cairns,Cairns, Neckerman, Gest, & Gariepy, 1988). Moreo-ver, in this context, aggressive children may find thatit is necessary to engage in more frequent and ex-treme conduct problems to establish and maintaintheir social positions among aggressive peers (Coie,2004).

Results from the WD1REJ additive model im-plied that the efficacy with which peer rejection andchildren’s withdrawn behavior predicted internaliz-ing problems increased as children matured, or werestronger during the later rather than the early gradeschool years. Compared with results for the AG1REJmodel, these findings suggest that internalizingproblems are less stable than externalizing problemsthroughout grade school, and that peer rejection andwithdrawn behavior are less powerful predictors(risk factors) for this type of dysfunction, especiallyduring the early elementary school years. Of the tworisk factors, peer rejection was the stronger; unlikewithdrawn behavior, it emerged as a significantpredictor of internalizing problems early in gradeschool (as indicated by main effects results), and itsdistinct predictive contributions increased progres-sively over time (implied by additive model find-ings). This was in contrast to rejection’s distinctpredictive contributions (when evaluated with ag-gressive behavior) to externalizing problems, which,as previously noted, declined as children matured.These contrasting data patterns were consistent withMcDougall et al.’s (2001) contention that peer rejec-tion experiences are particularly influential in thedevelopment of internalizing problems. Moreover,these data extend that inference by suggesting thatrejection’s role becomes progressively more impor-tant as an antecedent of internalizing problems aschildren mature. This may be the case because re-peated exposure to rejection experiences (e.g., peermaltreatment) may intensify children’s feelings ofdistress and strengthen processes that instigate in-ternalizing problems (e.g., uncertainty, mistrust, de-bilitating self-attributions, etc.; see Chorpita &Barlow, 1998; Pomerantz & Rudolph, 2003; Rubin etal., 1990; Troop-Gordon & Ladd, 2005). However, a


comparison of the absolute magnitude of rejection’sdistinct predictive contributions (additive pathweights) with externalizing versus internalizing prob-lems over time does not appear to justifyFat least forthe age period investigated hereFthe conclusion thatpeer rejection is a more powerful risk factor for in-ternalizing than for externalizing problems.

Additionally, the findings expand what is knownabout withdrawn behavior as a risk factor for psy-chopathology. The data were consistent with theconclusion that the risk posed by withdrawn be-havior for internalizing problems was low duringthe early grade school years, but higher during themiddle grade school years when, perhaps not sur-prisingly, peers begin to judge withdrawn behavioras atypical or deviant (Younger & Boyko, 1987;Younger et al., 1993). It may be the case that, aschildren grow older, those prone to engage in with-drawn behavior become increasingly marginalizedfrom the mainstream of social interaction and, ac-cordingly, begin to experience stronger feelings ofalienation, anxiety, loneliness, and other internaliz-ing symptoms. Furthermore, by middle childhoodand thereafter, withdrawn behavior may become aless mature way of coping with the social demandsof peer environments, and norms may shift such thatchildren who engage in such behaviors are at greaterrisk for peer censure. Were this the case, both with-drawn behavior and peer rejection would becomemore powerful risk factors, or predictors of subse-quent internalizing problems. Indeed, the findingsreported here were congruent with this inference.

Comparatively, however, it would appear thatwithdrawn behavior does not carry the same level ofrisk for internalizing problems during the gradeschool years as aggression does for externalizingproblems. In part, this appears to be attributable tothe fact that withdrawn behavior exhibited lesserstability over time as compared with aggressive be-havior, and the fact that internalizing problemsmanifested lesser stability than did externalizingproblems. This result was consistent with past evi-dence indicating that, at early ages, withdrawn be-havior is not a particularly powerful risk factor formaladjustment (Ladd & Burgess, 1999; Rubin, 1982,1993). Still, it may be judicious to reexamine thisconclusion in light of additional evidence. Whetheror not these findings accurately gauge the risk as-sociated with early withdrawn behavior may de-pend, in part, on investigators’ definitions of thisconstruct and their ability to distinguish it concep-tually and empirically from its psychological conse-quences (particularly, problems of an internalizingnature). A key aim of this investigation was to define

and measure withdrawn behavior in a way thatwould distinguish its attributes (i.e., avoidance of,and infrequent interaction with peers) from thoseindicative of its hypothesized or to-be-predictedconsequences (internalizing problems). Whether thisaim was fully realized using existing measures ofwithdrawn behavior (e.g., TRF-WD, CBS-AB sub-scales) warrants further theoretical and empiricalscrutiny and, perhaps, cannot be determined untilgreater precision is achieved in defining and mea-suring this construct and distinguishing it from itsreputed consequences. Interpretation is also com-plicated by the conceptual question of whetherwithdrawn behaviorFonce distinguished from itspsychological consequencesFis multifaceted ortakes more than one form (see Rubin & Coplan,2004). If the measures used to construct the latentvariable used in this study tapped more than oneform of withdrawn behavior, then the reportedfindings may not accurately reflect the risk associ-ated with constituent subtypes.

In sum, the results of this investigation reflect on along-standing controversy about the efficacy of twochildhood risk factorsFchildren’s behavioralorientations (i.e., aggressive, withdrawn) and peergroup rejectionFas predictors of incipient and later-emerging forms of psychological maladjustment.Unlike prior studies, the results provide a more nu-anced, developmental perspective on when, to whatextent, and how consistently children’s behavioralstyles and peer rejection forecast the development ofexternalizing and internalizing problems during the5 – 12 age period. Peer group rejection, when evalu-ated in conjunction with children’s behavior, madedistinct contributions to the prediction of both ex-ternalizing and internalizing problems, and did soconsistently throughout the 5 – 12 age period. As an-ticipated, aggressive behavior was a stronger pre-dictor of externalizing problems than was peerrejection over the entire course of this age period, butno support was found for the conclusion that onlyaggression accounts for significant nonshared vari-ance in externalizing problems. The predictive con-tributions of withdrawn behavior to internalizingproblems was eclipsed by peer rejection earlier inchildhood, but became progressively stronger (al-though not stronger than rejection) as children ap-proached preadolescence. No support was found forthe premise that withdrawn behavior increaseschildren’s risk for externalizing problems, nor thesupposition that aggressive behavior increases chil-dren’s risk for internalizing problems.

Thus, these data add to a growing corpus of evi-dence that contests the validity of ‘‘incidental mod-

842 Ladd

els’’ (Parker & Asher, 1987; Rudolph & Asher, 2000) orthe proposition that peer rejection is simply a marker(derivative) of the risks posed by children’s preexist-ing behavioral tendencies, such as aggressive andwithdrawn behavior. Rather, the overall pattern offindings bolstered the assumption that both children’sbehavior and their exposure to peer group rejectionare determinants of psychological maladjustment.

References

Achenbach, T. M. (1991). Manual for the teacher’s report formand 1991 profile. Burlington: Department of Psychiatry,University of Vermont.

Asher, S. R., & Dodge, K. A. (1986). Identifying childrenwho are rejected by their peers. Developmental Psycholo-gy, 22, 442 – 449.

Asher, S. R., Rose, A. J., & Gabriel, S. W. (2001). Peer re-jection in everyday life. In M. R. Leary (Ed.), Interpersonalrejection (pp. 105 – 142). Oxford, UK: Oxford UniversityPress.

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

Bagwell, C. L., Newcomb, A. F., & Bukowski, W. M. (1998).Early adolescent friendship and peer rejection as pre-dictors of adult adjustment. Child Development, 69,140 – 153.

Boivin, M., & Hymel, S. (1997). Peer experiences and socialself-perceptions: A sequential model. DevelopmentalPsychology, 33, 135 – 145.

Boivin, M., Hymel, S., & Bukowski, W. M. (1995). Theroles of social withdrawal, peer rejection, and victimi-zation by peers in predicting loneliness and depressedmood in childhood. Development and Psychopathology, 7,765 – 785.

Boldizar, J. P., Perry, D. G., & Perry, L. C. (1989). Outcomevalues and aggression. Child Development, 60, 571 – 579.

Brown, B. B., Dolcini, M. N., & Leventhal, A. (1997).Transformations in peer relationships at adolescence:Implications for health-related behavior. In J. Schulenb-erg, J. L. Maggs, & K. Hurrelmann (Eds.), Health risks anddevelopmental transitions during adolescence (pp. 161 – 189).New York: Cambridge University Press.

Buhrmester, D., & Furman, W. (1986). The changing func-tions of friends in childhood: A neo-Sullivanian per-spective. In V. J. Derlega & B. A. Winstead (Eds.),Friendship and social interaction (pp. 41 – 61). New York:Springer-Verlag.

Buhs, E. S., & Ladd, G. W. (2001). Peer rejection as ante-cedent of young children’s school adjustment: An ex-amination of mediating processes. DevelopmentalPsychology, 37, 550 – 560.

Bukowski, W. M., & Hoza, B. (1989). Popularity andfriendship: Issues in theory, measurement, and outcome.In T. J. Berndt & G. W. Ladd (Eds.), Peer relationships inchild development (pp. 15 – 45). New York: Wiley.

Cairns, R. B., & Cairns, B. D. (1994). Lifelines and risks:Pathways of youth in our time. New York: CambridgeUniversity Press.

Cairns, R. B., Cairns, B. D., Neckerman, H. J., Ferguson, L.L., & Gariepy, J. L. (1989). Growth and aggression I:Childhood to early adolescence. Developmental Psychol-ogy, 25, 320 – 330.

Cairns, R., Cairns, B., Neckerman, H. J., Gest, S., & Gariepy,J. (1988). Peer networks and aggressive behavior: Peersupport or peer rejection? Developmental Psychology, 24,815 – 823.

Caspi, A., Elder, G. H., & Bem, D. J. (1987). Moving againstthe world: Life-course patterns of explosive children.Developmental Psychology, 23, 308 – 313.

Caspi, A., Elder, G. H., & Bem, D. J. (1988). Moving awayfrom the world: Life-course patterns of shy children.Developmental Psychology, 24, 824 – 831.

Chorpita, B. F., & Barlow, D. H. (1998). The development ofanxiety: The role of control in the early environment.Psychological Bulletin, 124, 3 – 21.

Cillessen, A. H. N., & Bukowski, W. M. (Eds.) (2000). Recentadvances in the measurement of acceptance and rejection inthe peer system (New Directions for Child Development, No.88). San Francisco: Jossey-Bass.

Coie, J. D. (1990). Toward a theory of peer rejection. InS. R. Asher & J. D. Coie (Eds.), Peer rejection in child-hood (pp. 365 – 401). New York: Cambridge UniversityPress.

Coie, J. D. (2004). The impact of negative social experi-ences on the development of antisocial behavior. InJ. B. Kupersmidt & K. A. Dodge (Eds.), Children’speer relations: From development to intervention (pp.243 – 267). Washington, DC: American PsychologicalAssociation.

Coie, J. D., & Kupersmidt, J. B. (1983). A behavioral anal-ysis of emerging social status in boys’ groups. ChildDevelopment, 54, 1400 – 1416.

Coie, J. D., Lochman, J. E., Terry, R., & Hyman, C. (1992).Predicting early adolescent disorder from childhoodaggression and peer rejection. Journal of Consulting andClinical Psychology, 60, 783 – 792.

Coie, J. D., Watt, N. F., West, S. G., Hawkins, D., Asarnow,J. R., Markman, H. J., et al. (1993). The science of pre-vention: A conceptual framework and some directionsfor a national research program. American Psychologist,48, 1013 – 1022.

Cole, D. A., & Maxwell, S. E. (2003). Testing mediationalmodels with longitudinal data: Questions and tips in theuse of structural equation modeling. Journal of AbnormalPsychology, 112, 558 – 577.

Cowan, P. A., & Cowan, C. P. (2004). From family rela-tionships to peer rejection to antisocial behavior inmiddle childhood. In J. B. Kupersmidt & K. A. Dodge(Eds.), Children’s peer relations: From development to inter-vention (pp. 159 – 177). Washington, DC: American Psy-chological Association.

Crick, N. R., & Ladd, G. W. (1990). Children’s perceptionsof the consequences of aggressive strategies: Do the ends


justify being mean? Developmental Psychology, 26,612 – 620.

DeRosier, M. E., Kupersmidt, J. B., & Patterson, C. J. (1994).Children’s academic and behavioral adjustment as afunction of the chronicity and proximity of peer rejec-tion. Child Development, 65, 1799 – 1813.

Dodge, K. A. (1983). Behavioral antecedents of peer socialstatus. Child Development, 54, 1386 – 1399.

Egan, S. E., Monson, T. C., & Perry, D. G. (1998). Socialcognitive influences on change in aggression over time.Developmental Psychology, 34, 996 – 1006.

Freud, A., & Dann, S. (1951). An experiment in groupupbringing. In R. Eisler, A. Freud, H. Hartmann, &E. Kris (Eds.), Psychoanalytic study of the child (Vol. 6,pp. 127 – 168). New York: International UniversityPress.

Furman, W., Rahe, D., & Hartup, W. (1979). The rehabili-tation of socially withdrawn preschool children throughmixed-age and same-age socialization. Child Develop-ment, 50, 915 – 922.

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

Harlow, H. F., Harlow, M. K., Dodsworth, R. O., & Arling,G. L. (1966). Maternal behavior of rhesus monkeys de-prived of mothering and peer associations in infancy.Proceedings of the American Philosophical Society, 110,58 – 66.

Harrist, A. W., Zaia, A. F., Bates, J. E., Dodge, K. A., &Pettit, G. S. (1997). Subtypes of social withdrawal inearly childhood: Sociometric status and social-cognitivedifferences across four years. Child Development, 68,278 – 294.

Hartup, W. (1970). Peer interaction and social organization.In P. H. Mussen (Ed.), Carmichael’s manual of child psy-chology (Vol. 2, 3rd ed., pp. 360 – 456). New York: Wiley.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indicesin covariance structure analysis: Conventional criteriaversus new alternatives. Structural Equation Modeling, 6,1 – 55.

Hymel, S., Rubin, K. H., Rowden, L., & LeMare, L. (1990).Children’s peer relationships: Longitudinal predictionof internalizing and externalizing problems from middleto late childhood. Child Development, 61, 2004 – 2021.

Kupersmidt, J. B., & Coie, J. D. (1990). Preadolescent peerstatus, aggression, and school adjustment as predictorsof externalizing problems in adolescence. Child Devel-opment, 61, 1350 – 1362.

Kupersmidt, J. B., Coie, J. D., & Dodge, K. A. (1990). Therole of poor peer relationships in the development ofdisorder. In S. R. Asher & J. D. Coie (Eds.), Peer rejectionin childhood (pp. 274 – 305). New York: Cambridge Uni-versity Press.

Kupersmidt, J. B., & DeRosier, M. E. (2004). How peerproblems lead to negative outcomes: An integrativemediational model. In J. B. Kupersmidt & K. A. Dodge(Eds.), Children’s peer relations: From development to inter-

vention (pp. 119 – 138). Washington, D.C.: AmericanPsychological Association.

Ladd, G. W. (1989). Children’s social competence and so-cial supports: Precursors of early school adjustment? InB. Schneider, G. Attili, J. Nadel, & R. Weissberg (Eds.),Social competence in developmental perspective (pp.277 – 292). Amsterdam: Kluwer Academic Publishers.

Ladd, G. W. (1996). Shifting ecologies during the 5 – 7 yearperiod: Predicting children’s adjustment to gradeschool. In A. Sameroff & M. Haith (Eds.) The five to sevenyear shift (pp. 363 – 386). Chicago: University of ChicagoPress.

Ladd, G. W. (2003). Probing the adaptive significance ofchildren’s behavior and relationships in the schoolcontext: A child by environment perspective. In R. Kail(Ed.), Advances in child behavior and development (31, pp.43 – 104). New York: Wiley.

Ladd, G. W. (2005). Children’s peer relationships and socialcompetence: A century of progress. New Haven, CT: YaleUniversity Press.

Ladd, G. W., Birch, S. H., & Buhs, E. (1999). Children’ssocial and scholastic lives in kindergarten: RelatedSpheres of Influence? Child Development, 70, 1373 –1400.

Ladd, G. W., Buhs, E., & Seid, M. (2000). Children’s initialsentiments about kindergarten: Is school liking an an-tecedent of early classroom participation and achieve-ment? Merrill-Palmer Quarterly, 46, 255 – 279.

Ladd, G. W., & Burgess, K. B. (1999). Charting the rela-tionship trajectories of aggressive, withdrawn, and ag-gressive/withdrawn children during early grade school.Child Development, 70, 910 – 929.

Ladd, G. W., Herald, S., & Andrews, K. (2006). Youngchildren’s peer relations and social competence. InB. Spodek & O. Saracho (Eds.), Handbook of research onthe education of young children (Vol. 2, pp. 23 – 54). NewYork: Macmillan.

Ladd, G. W., & Mars, K. T. (1986). Reliability and validity ofpreschoolers’ perceptions of peer behavior. Journal ofClinical Child Psychology, 15, 16 – 25.

Ladd, G. W., & Profilet, S. M. (1996). The Child BehaviorScale: A teacher-report measure of young children’saggressive, withdrawn, and prosocial behaviors. Devel-opmental Psychology, 32, 1008 – 1024.

Ladd, G. W., & Troop-Gordon, W. (2003). The role ofchronic peer adversity in the development of children’spsychological adjustment problems. Child Development,74, 1325 – 1348.

McDougall, P., Hymel, S., Vaillancourt, T., & Mercer, L.(2001). The consequences of childhood peer rejection. InM. R. Leary (Ed.), Interpersonal rejection (pp. 213 – 247).Oxford, UK: Oxford University Press.

Michael, C. M., Morris, D. P., & Soroker, E. (1957). Follow-up studies of shy, withdrawn children II. Relativeincidence of schizophrenia. American Journal of Ortho-psychiatry, 27, 331 – 337.

Morris, D. P., Soroker, E., & Burruss, G. (1954). Follow-upstudies of shy, withdrawn children I: Evaluation of later

844 Ladd

adjustment. American Journal of Orthopsychiatry, 24,743 – 754.

Moss, H. A., & Susman, E. J. (1980). Longitudinal studyof personality development. In O. G. Brim Jr. & J.Kagan (Eds.), Constancy and change in human development(pp. 530 – 595). Cambridge, MA: Harvard UniversityPress.

Olweus, D. (1979). Stability of aggressive reaction pat-terns in males: A review. Psychological Bulletin, 86, 852 –875.

Parker, J. G., & Asher, S. R. (1987). Peer relations and laterpersonal adjustment: Are low-accepted children ‘‘atrisk’’? Psychological Bulletin, 102, 357 – 389.

Parker, J. G., Rubin, K. H., Price, J. M., & DeRosier, M.(1995). Peer relationships, child development, and ad-justment: A developmental psychopathology perspec-tive. In D. Cicchetti & D. Cohen (Eds.), Developmentalpsychopathology: Vol. 2. Risk, disorder and adaptation (pp.96 – 161). New York: Wiley.

Pomerantz, E. M., & Rudolph, K. D. (2003). What ensuesfrom emotional distress? Implications for competenceestimation. Child Development, 74, 329 – 345.

Renshaw, P., & Brown, P. J. (1993). Loneliness in middlechildhood: Concurrent and longitudinal predictors.Child Development, 64, 1271 – 1284.

Rensvold, R. B., & Cheung, G. W. (2001). Testing for metricinvariance using structural equations models: Solvingthe standardization problem. In C. A. Schriesheim &L. L. Neider (Eds.), Research in management: Vol 1.Equivalence in measurement (pp. 21 – 51). Greenwich, CT:Information Age.

Robins, L. N. (1966). Deviant children grow up. Baltimore:Williams & Wilkins Press.

Roff, J. D., & Wirt, D. (1984). Childhood aggression andsocial adjustment as antecedents of delinquency. Journalof Abnormal Child Psychology, 12, 111 – 126.

Roff, M. (1961). Childhood social interaction and youngadult bad conduct. Journal of Abnormal Social Psychology,63, 333 – 337.

Roff, M. (1963). Childhood social interactions and youngadult psychosis. Journal of Clinical Psychology, 19,152 – 157.

Roff, M., Sells, S. B., & Golden, M. M. (1972). Social ad-justment and personality development in children. Min-neapolis: University of Minnesota Press.

Rubin, D. B. (1987). Multiple imputation for nonresponse insurveys. New York: Wiley.

Rubin, K. H. (1982). Nonsocial play in preschoolers: Nec-essary evil? Child Development, 53, 651 – 657.

Rubin, K. H. (1993). The Waterloo Longitudinal Project:Correlates and consequences of social withdrawal fromchildhood to adolescence. In K. H. Rubin & J. B. As-endorpf (Eds.), Social withdrawal, inhibition, and shyness inchildhood (pp. 291 – 314). Hillsdale, NJ: Lawrence Er-lbaum.

Rubin, K. H., & Asendorpf, J. (1993). Social withdrawal, in-hibition, and shyness in childhood. Hillsdale, NJ: LawrenceErlbaum.

Rubin, K. H., & Coplan, R. (2004). Paying attention to andnot neglecting social withdrawal and social isolation.Merrill-Palmer Quarterly, 50, 506 – 534.

Rubin, K. H., LeMare, L., & Lollis, S. (1990). Social with-drawal in childhood: Developmental pathways to peerrejection. In S. R. Asher & J. D. Coie (Eds.), Peer rejectionin childhood (pp. 217 – 249). New York: Cambridge Uni-versity Press.

Rudolph, K. D., & Asher, S. R. (2000). Adaptation andmaladaptation in the peer system: Developmentalprocesses and outcomes. In A. J. Sameroff, M. Lewis, &S. Z. Miller (Eds.), Handbook of developmental psychopa-thology (pp. 157 – 175). New York: Kluwer Academic/Plenum.

Rutter, M. (1996). Transitions and turning points in devel-opmental psychopathology: As applied to the age spanbetween childhood and mid-adulthood. InternationalJournal of Behavioral Development, 19, 603 – 626.

Sameroff, A. J. (1987). The social context of development.In N. Eisenberg (Ed.), Contemporary topics in develop-mental psychology (pp. 273 – 291). New York: Wiley.

Sameroff, A. J., & Chandler, M. J. (1975). Reproductiverisk and the continuum of caretaking causality. In F. D.Horowitz (Ed.). Review of child development research(Vol. 4, pp. 187 – 244). Chicago: University of ChicagoPress.

Sandstrom, M. J., & Zakriski, A. L. (2004). Understandingthe experience of peer rejection. In J. B. Kupersmidt &K. A. Dodge (Eds.), Children’s peer relations: From devel-opment to intervention (pp. 101 – 118). Washington, DC:American Psychological Association.

Schafer, J. L. (1999). NORM: Multiple imputation of in-complete multivariate data under a normal model, ver-sion 2. Software for Windows 95/98/NT, available fromhttp://www.stat.psu.edu/� jls/misoftwa.html.

Schafer, J. L., & Olsen, M. K. (1998). Multiple imputa-tion for multivariate missing-data problems: A dataanalyst’s perspective. Multivariate Behavioral Research, 33,545 – 571.

Slaby, R. G., & Guerra, N. G. (1988). Cognitive mediators ofaggression in adolescent offenders: I. Assessment. De-velopmental Psychology, 24, 580 – 588.

Sullivan, H. S. (1953). The interpersonal theory of psychiatry.New York: W.W. Norton & Company.

Suomi, S. J., & Harlow, H. F. (1972). Social rehabilitation ofisolate reared monkeys. Developmental Psychology, 6,487 – 496.

Tremblay, R. E., Pihl, R. O., Vitaro, F., & Dobkin, P. L. (1994).Predicting early onset of male antisocial behavior frompreschool behavior. Archives of General Psychiatry, 51,732 – 739.

Troop-Gordon, W., & Ladd, G. W. (2005). Trajectories ofpeer victimization and perceptions of the self andschoolmates: Precursors to internalizing and external-izing problems. Child Development. 76, 1072 – 1091.

Wanlass, R. L., & Prinz, R. J. (1982). Methodological issuesin conceptualizing and treating childhood social isola-tion. Psychological Bulletin, 92, 39 – 55.


Wright, J. C., Giammarino, M., & Parad, H. W. (1986). So-cial status in small groups: Individual-group similarityand the social ‘‘misfit’’. Journal of Personality and SocialPsychology, 50, 523 – 536.

Younger, A. J., & Boyko, K. A. (1987). Aggression andwithdrawal as social schemas underlying children’speer perceptions. Child Development, 58, 1094 – 1100.

Younger, A. J., Gentile, C., & Burgess, K. B. (1993). Chil-dren’s perceptions of social withdrawal: changes acrossage. In K. Rubin & J. Asendorpf (Eds.), Social withdrawal,inhibition, and shyness in childhood (pp. 215 – 235). Hills-dale, NJ: Lawrence Erlbaum.

846 Ladd

peer rejection, aggressive or withdrawn behavior, and

Documents