Reformulating infant and toddler social competence with peers

Infant Behavior & Development 30 (2007) 353–365

Reformulating infant and toddler social competence with peers

Shannon T. Williams ∗, Lenna L. Ontai, Ann M. MastergeorgeUniversity of California, Davis, United States

Received 21 January 2006; received in revised form 31 August 2006; accepted 5 October 2006

Abstract

Efforts to identify children at risk for social and academic difficulties at an early age have begun to apply conceptualizationsof social competence with peers from childhood to infancy and toddlerhood. These attempts have had limited success, in partbecause social behaviors studied in later childhood such as negative or aggressive acts may not capture relevant dimensions ofsocial competence with peers during the infant and toddler years. The present study conducted an exploratory factor analysis tobegin developing a conceptualization of social competence with peers that is appropriate for use with infants and toddlers, and thatcaptures individual differences capable of predicting later social functioning. Results indicate three dimensions of infant and toddlersocial competence with peers: peer sociability, active peer refusal, and passive peer avoidance. Negative and aggressive behaviorsloaded on each of the three factors rather than emerging as a unique dimension of social competence. Each factor demonstratedboth convergent and predictive validity, indicating that early identification of social skills remains a reasonable goal for practitionersand researchers alike. Finally, antecedents within children, families, and out-of-home child care environments helped to explainindividual variability in each of these three factors of social competence with peers.© 2006 Elsevier Inc. All rights reserved.

Keywords: Peers; Social competence; Aggression; Infancy

The development of social competence with peers begins as early as the first year of life (Brownell & Brown,1992; Eckerman, Whatley, & McGehee, 1979; Hay, 1985; Howes, 1988 (1, serial no. 217); Rauh, 1987; Vandell &Wilson, 1987). However, early peer interactions are generally quite rudimentary. They tend to be exploratory in nature,and are often limited to interactions around objects and intense watching or looking at the social partner (Brownell& Brown, 1992; Eckerman et al., 1979; Rauh, 1987; Rubin, Bukowski, & Parker, 1998). In contrast, by the earlyschool years children’s patterns of interacting with their peers become fairly stable, and serve as important predictorsof later social functioning, academic success, and mental health (Biovin, Vitaro, & Poulin, 2005; Campbell, Lamb,& Hwang, 2000; Coie, Terry, Lenox, Lochman, & Hyman, 1995; Ladd, 1999; Rubin et al., 1998). The contrast inpeer interactions between infancy and the early school years has resulted in qualitatively distinct research agendas andconceptualizations of social competence between these age groups.

Despite these conceptual differences, a new line of research is working to bridge the work on social competence dur-ing the school years with that from the infant and toddler years. These efforts aim to better understand the developmentof social difficulties and their origins at an early age, thus building a framework for preventive interventions (Campbellet al., 2000; Campbell, 2002; Howes, 1988; National Institute for Child Health and Human Development Early Child

∗ Corresponding author at: Department of Human and Community Development, University of California, Davis, One Shields Avenue, Davis,CA 95616, United States. Tel.: +1 530 754 4911; fax: +1 530 752 5660.

E-mail address: [email protected] (S.T. Williams).

0163-6383/$ – see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.infbeh.2006.10.008

mailto:[email protected]

dx.doi.org/10.1016/j.infbeh.2006.10.008

354 S.T. Williams et al. / Infant Behavior & Development 30 (2007) 353–365

Care Research Network [NICHD ECCRN], 2001). However, due, in part, to the differences in conceptualizations ofsocial competence between infancy and early childhood, these efforts have experienced limited success (Campbellet al., 2000; Howes, 1988; National Institute of Child Health and Human Development Early Child Care ResearchNetwork [NICHD ECCRN], 2001).

The present study aims to develop a conceptualization of social competence with peers that will facilitate longitudinalstudies of peer competence from infancy through preschool. This formulation intends to be appropriate for use withinfants and toddlers, and to capture individual differences in social competence that are capable of predicting latersocial functioning and that are sensitive to antecedent contributions from individual, family, and contextual factors.

1. Previous conceptualizations of social competence with peers

Frameworks for investigating social competence in early childhood differ from those for studying later childhood.While researchers of infant and toddler interactions with peers most often examine normative, age-related behaviors,investigations of older children’s social competence tend to focus on individual differences in patterns of behavior,using these differences to predict developmental outcomes. The separation of these two lines of inquiry has resulted indifferent types of focal behaviors between the two developmental periods. In particular, studies of infants and toddlerstend to focus on the development of the content and the complexity of peer interactions (Bronson, 1981; Brownell &Hazen, 1999; Eckerman et al., 1979; Hay, 1985; Howes, 1980, 1988; Ross & Lollis, 1987; Vandell & Wilson, 1987). Incontrast, the study of peer interactions during childhood tends to emphasize individual differences in social competence,such as popularity with peers (Coie, Dodge, & Kupersmidt, 1990; Coie et al., 1995; Rubin et al., 1998), and aggressive,or negative behavior (Biovin et al., 2005; Coie & Dodge, 1998; Rubin et al., 1998). This line of inquiry has establishedthe importance of aggressive behavior for children’s social status and their social, emotional, and academic functioning(Burks, Dodge, & Price, 1995; Coie et al., 1995; for a review see McDougall, Hymel, Vaillancourt, & Mercer, 2001).

In the attempt to identify these children early in their development, and to study their development over time,researchers have begun investigating negative and aggressive peer interactions in toddler and preschool children(Campbell et al., 2000; Howes, 1988; NICHD ECCRN, 2001; Vaughn, Vollenweider, Bost, Azria-Evans, & Snider,2003; for a review see Odom & Ogawa, 1992). Findings depict instability of individual differences in negative andaggressive peer behaviors between the infant and preschool years (Campbell et al., 2000; Howes, 1988). In addition,early measures of negative, aggressive peer interactions fail to predict children’s subsequent social competence withpeers (Howes, 1988). It is not clear exactly when negative interactions do reach stability, though results from longitu-dinal investigations suggest that stability begins to occur around age two (NICHD ECCRN, 2001) or three (Campbellet al., 2000).

This instability in aggressive and negative behaviors may result from a shift in the meaning of these behaviorsfrom the infant and toddler period to the preschool and school-aged years. While many infants and toddlers exhibitelevated levels of aggressive or difficult interactive behavior (Biovin et al., 2005; NICHD ECCRN, 2004; Shantz,1987; Tremblay, 2000; Tremblay & Nagin, 2005), it may be inappropriate to define aggressive behavior as poorsocial competence during these early years. Rather, aggression during the earliest years represents a normative age-related pattern of development. Longitudinal studies suggest that children’s patterns of aggressive behavior tend toincrease during the toddler and preschool years and to decline thereafter (NICHD ECCRN, 2004; Tremblay, 2000;Tremblay & Nagin, 2005). From an evolutionary perspective, this period of heightened aggression can be viewed asadaptive, allowing children to exercise their impulses within safe environments. With the help of parents and otheradult caregivers, young children learn how to express their aggression, physical play, and negative emotions in anappropriate manner (Cairns & Cairns, 2000; NICHD ECCRN, 2004; Suomi, 2000; Shantz, 1987; Tremblay & Nagin,2005). Thus, as infants and toddlers learn to control these behaviors, they can begin to develop more adaptive socialbehaviors. Research suggests that infants and toddlers who are aggressive also tend to be more socially competent thannon-aggressive children (Biovin et al., 2005; NICHD ECCRN, 2001; Vaughn et al., 2003).

1.1. Working toward a re-conceptualization of peer competence for infants and toddlers

Given the inconsistency between social behaviors in infancy and early childhood, there is a need for a measure thatcaptures individual differences in social competence with peers that is appropriate for use in the infant and toddleryears. As previously discussed, recent attempts to employ assessments of negative or aggressive behaviors during this

S.T. Williams et al. / Infant Behavior & Development 30 (2007) 353–365 355

period have revealed that these behaviors may represent either generalized sociability or part of a process for learningappropriate social behaviors. As a result, experts in the field have called for researchers to begin examining otherdimensions of early social competence with peers that might hold more potential to predict later social functioning(Brownell & Hazen, 1999; Howes, 1988; NICHD ECCRN, 2001; Vaughn et al., 2003).

Previous work by Howes (1988) has taken the initial step toward reaching this goal. In an attempt to investigateindividual differences in social competence with peers, Howes found that a teacher-rated measure of peer sociabilitydemonstrated moderate stability over time, and also predicted various aspects of children’s later social competencewith peers, including sociometric status. While this measure included infants and toddlers, it was developed and factoranalyzed for children through 6 years of age; thus, it may not be directly applicable to the specific interactive behaviorsof infants and toddlers (Howes, 1988). In particular, the measure results in three factors: sociability, hesitant, anddifficult, which is comprised of negative and aggressive items. Consistent with other research (Campbell et al., 2000),individual differences in this difficult dimension during infancy and toddlerhood did not predict later social competence.Thus, there remains a need for research that systematically examines peer interactions during the infant and toddleryears in order to develop measures that accurately reflect relevant social behaviors during this period of development.

1.2. Antecedents of early individual differences in social competence with peers

In order to develop effective prevention efforts of child aggression, it is important that early measures of socialbehavior accurately identify antecedents of later aggressive tendencies (Campbell, 2002). Findings from investigationsof the complexity of peer play, sociometric ratings, and negative interactions suggest that young children’s experiences inboth their homes and their out-of-home child care environments contribute the development of social skills. Specifically,family demographics such as income and maternal education (Lamb, 1998; NICHD ECCRN, 2001), and structuralfacets of child care environments such as caregiver training, group size, and caregiver–child ratio (Fitzgerald, Mann,Cabrera, & Wong, 2003; Howes, Phillips, & Whitebook, 1992; Lamb, 1998; NICHD ECCRN, 2001, 2002; Rosenthal,1991) have been found to be important antecedents. Even more important than these structural aspects, children’sinteractions with adults have been found to contribute to the development of social and emotional competencies(Lamb, 1998; Howes et al., 1992; NICHD ECCRN, 2002; Peisner-Feinberg et al., 2001). In particular, research hasfound that caregivers may directly influence the course of peer interactions in young children (Kontos, 1999), eitherthrough involvement in their resolutions of conflicts (Bayer, Whaley, & May, 1995; Chen, 2003; Goncu & Cannella,1996), or through obstruction of their interactions by interacting with them individually (Harper & McClusky, 2003).The present study aims to add to this dialogue by investigating the effect of caregiver involvement in infant and toddler’speer interactions.

In addition to interactions with adults, individual characteristics such as temperament may also assume an importantrole in the development of social competence with peers. Temperamental qualities tend to be fairly stable throughoutchildhood and are closely linked to children’s tendencies to approach or avoid social situations, therefore playing animportant role in early social development (for reviews see Kagan, 1998; Rubin et al., 1998; Sanson, Hemphill, &Smart, 2004). For example, inhibited children, and those high in negative emotionality tend to display withdrawal frompeers (Calkins & Fox, 1992; Kagan, 1989), while children with easy temperaments, or who tend to be open to novelty,tend to be social with their peers (Skarpness & Carson, 1986; Stocker & Dunn, 1990). These associations betweentemperament and social competence with peers have been consistently documented from the preschool years onward,and the present study aims to extend this line of investigation to the infant and toddler years.

1.3. The present study

The present study aims to work toward a new conceptualization of social competence with peers that is both age-appropriate to infants and toddlers, and identifies individual differences in social interactions with peers that can be usedto predict developmental outcomes. This conceptualization builds from the measure of social competence developedby Howes (1988).

In the present study several hypotheses were developed. It is expected that (1) factors of social competence with peersin infancy will include global dimensions of sociability and withdrawal. This hypothesis is consistent with previousfindings highlighting the relative immaturity of early peer interactions (Brownell & Brown, 1992; Eckerman et al., 1979;Rauh, 1987; Rubin et al., 1998). It is further hypothesized that (2) negative and aggressive interactive behaviors will be


included in these sociability and withdrawal factors rather than comprising their own unique social dimension. Thesefactors of social competence are expected to (3) demonstrate both convergent and predictive validity with generalizedsocial competence. It is further hypothesized that (4) infants will exhibit individual differences in these factors ofsocial competence with peers. Results from investigations of older children (e.g. Fitzgerald et al., 2003; Lamb, 1998;NICHD ECCRN, 2002, 2001) guide the hypothesis that (5) patterns of peer sociability will be predicted by familydemographics, experiences in out-of-home child care experiences, and individual temperament. Since little evidenceis available to suggest a relationship between out-of-home child care experiences and social withdrawal, however, (6)infant’s patterns of withdrawal from peer interaction are hypothesized to be explained by family demographics andindividual temperament rather than by child care experiences.

2. Method

2.1. Participants

Participants were drawn from a larger study examining the contribution of childcare and home environments oninfant development. The current sample includes 36 children (53% male) between the ages of 12 and 17 months ofage, with a mean age of 13.00 months. Families were recruited from four classrooms within three center-based infantchild care programs in medium sized city in Northern California. All of the children were attending child care at thetime of data collection. Children entered these center-based child care arrangements at varying ages, ranging from 2to 16 months of age, with a mean of 7.3 months.

The sample was primarily White, middle-class families (69% white, 97% from two-parent families, 71%US$ 50,000+ income). Other ethnic groups represented in this sample include Latino (6%), Asian/Pacific Islander(3%), and mixed or other ethnicity (23%). The annual income of 29% of the families fell below US$ 50,000, which islow enough to qualify for child care assistance programs for most family sizes in this area.

2.2. Procedure

Each of the three child care centers were visited by a researcher for three consecutive mornings, for approximatelyone and one half hours each day. Caregivers’ behavior during free play sessions was videotaped, but the videotapingprocedure focused on individual children to assess each child’s experience with his or her caregivers. Videotaping uti-lized a time sampling procedure in which one child was videotaped for three consecutive minutes, including caregiversand other children in proximity to that child. Then the next participating child was then videotaped for 3 min in thesame manner, until all the children had been videotaped for 3 min. This cycle was repeated three times during each ofthe 3 days, for a total of 27 min of videotaped free play time per child.

During the first visit to each classroom the primary caregiver for each child received a questionnaire packet tocomplete and return to the researcher the following week. Children’s parents also completed a packet of questionnairesconcerning family demographics and children’s temperamental characteristics. Approximately 6 months later theprimary caregiver for each child completed the questionnaire packet again. Four of the original 36 children had leftthe child care center by the time of this follow up. Data from the follow-up assessment are available for 32 of the 36children.

2.3. Measures

2.3.1. Family socioeconomic statusMothers completed a questionnaire regarding family demographic characteristics including family’s gross annual

household income and mother’s education level. Due to the high correlation between annual household income andmaternal education (r = .56, p < .001), these two scores were aggregated into a composite socioeconomic status variable.

2.3.2. Social competence with peersA modified version of the teacher-rated peer competence scale (Howes, 1988) was utilized in the present study. This

scale, originally derived from rating scales used with older children (Baumrind, 1968; Greenwood, Walker, Todd, &Hops, 1979), asks teachers to rate each child’s behavior with peers, relative to age mates. The version of the scale used


by Howes (1988) was designed for use with children of varying ages. Conclusions suggested that the scale may nothave been adequate for all ages (Howes, 1988). Therefore, modifications for infants and toddlers were made for usein the present study based on findings from observational studies of infant and toddler peer interactions (Brownell &Brown, 1992; Eckerman & Didow, 1996; Hay & Ross, 1982; Odom & Ogawa, 1992; Rauh, 1987; Vandell & Wilson,1987). The final scale included 12 items, each rated on a five point Likert scale, and included items such as “imitatethe actions or expressions of peers”, “take toys from peers”, and “watch rather than participate in peer interaction”.

2.3.3. Overall social competenceChild care providers completed the Brief Infant Toddler Social Emotional Assessment (Briggs-Gowan & Carter,

2001) at both the first and second measurement occasions. The BITSEA consists of 60 items drawn from the fullInfant Toddler Social Emotional Assessment (ITSEA) (Carter & Briggs-Gowan, 2000), a comprehensive measureof social–emotional problems and competencies in children 1–2 years of age. The BITSEA items were rated on athree point Likert scale ranging from “not true/rarely” to “very true/often” and comprise two separate Problem andCompetence scales. The complete Problem subscale measures both typical social problems (e.g. hits, bites, or kicks you;has trouble calming down when upset) and symptoms of more serious developmental disturbances (hurts him/herselfon purpose; is not afraid when should be). For the purposes of the present study only the typical social problemswere included in the problem subscale (Cronbach’s alpha = .96, .80). The Competence subscale includes items such as“imitates playful sounds” and “follows rules” (Cronbach’s alpha = .73, .83).

2.3.4. Child temperamentParents completed the Revised Infant Behavior Questionnaire (Garstein and Rothbart, 2003). This 191 item ques-

tionnaire asks parents how often their infant engaged in behaviors during the past week, with a seven point Likert-typeresponse scale ranging from “never” to “always”. Items are aggregated into sub-scales that assess approach, vocalreactivity, high intensity pleasure, activity level, perceptual sensitivity, sadness, distress to limitations, fear, fallingreactivity, low intensity pleasure, cuddliness, duration of orienting, and soothability (Garstein and Rothbart, 2003).The average internal consistency for these subscales in the present study was alpha = .75, with a range from .58 to .89.These sub-scales collectively comprise three broad dimensions of infant temperament: Surgency/Extraversion (Cron-bach’s alpha = .94), Negative Affectivity (Cronbach’s alpha = .92), and Orienting/Regulation (Cronbach’s alpha = .79).

2.3.5. Caregiver interruption of peer interactionVideotapes of children and their caregivers during free play were coded for caregiver’s interruption of peer interac-

tion. Each interaction was coded for either occurrence or no occurrence of caregiver interruption. Criteria for interruptionincluded any behavior, vocalization, or expression by a caregiver that caused the focal child to shift attention awayfrom the peer(s) and toward the caregiver. Occurrences of caregiver interruption were summed and then divided by thetotal number of minutes each child was videotaped, producing a score of the rate of interruptions per minute. Codingwas completed by two undergraduate students who were blind to the study hypotheses and who demonstrated adequateinter-rated reliability (89% agreement).

2.4. Analysis plan

An exploratory factor analysis was conducted to examine the dimensions of social competence with peers. Solutionscontaining one through four factors were compared through examination of the scree plot (Cattell, 1966), and evaluationof the patterns of factor loadings and cross-loadings (Gorsuch, 1983). Each of these three dimensions were thenexamined for convergent validity, through the use of zero-order correlations with the Brief Infant Toddler SocialEmotional Assessment (BITSEA). Predictive validity was examined through the use of Hierarchical Linear Modeling(HLM) (Raudenbush & Bryk, 2002) to examine the extent to which each of the social competence dimensions predictedindividual differences on the BITSEA Assessment 6 months later. The HLM procedure is the preferred method ofanalysis for these data because it accounts for the clustering of children within classrooms, whereas traditional regressionprocedures assume observations to be independent from one another. HLM also allows the numbers of children in eachchild care classroom to vary, whereas traditional tests for group differences, such as the ANOVA, assume an equalnumber (Raudenbush & Bryk, 2002).


Fig. 1. Scree plot for the exploratory factor analysis of infant and toddler social competence with peers.

Once convergent and predictive validity were established, patterns of stability and change in each dimension ofpeer competence were examined through the use of zero-order correlations, Analysis of Variance, and changes instandard deviations over time. Finally, we employed a second set of HLM analyses to examine antecedents of each ofthe dimensions of social competence with peers, including child, family, and child care characteristics. In each of theseanalyses, scores on the peer competence dimensions from the first measurement occasion were included as controlvariables to examine the contributions of child, family, and child care characteristics to changes in children’s socialcompetence with peers over a 6-month period of time.

3. Results

3.1. Factor structure of peer questionnaire

Exploratory factor analyses of the peer interaction scale revealed three components. As illustrated by the screeplot (Cattell, 1966) (Fig. 1), the first three factors account for a much larger percentage of the total variance inscores on this scale than do additions of subsequent factors. Converging evidence for the three factor solution wasfound through examination of the factor structure for solutions with varying numbers of factors (Gorsuch, 1983).The three factor solution appeared to represent the data best, with each factor containing at least three salientloadings and other loadings near zero, each item loading on at least one factor, and few items cross-loading onmore than one factor (Gorsuch, 1983) (see Table 1). The three factor solution of peer interaction contained dimen-

Table 1Factor structure of the peer competence scale

Peer sociability Active peer refusal Passive peer avoidance

Smile at peers .811 −.078 −.364Initiate play with peers .829 .212 −.029Respond to peer’s attempts to initiate play by joining them .928 .019 .134Imitate the actions or expressions of peers .546 .044 .230Take toys from peers .453 .249 −.211Withdraw from peer activity .263 .867 .154Refuse peer’s attempts to play .004 .750 −.129Turn his or her back toward or move away from peers .077 .745 .321Watch rather than participate −.121 −.353 .723Act as if he/she does not notice peer’s attempts to initiate play .356 .198 .612Cry or fuss easily around peers −.004 .173 .415Hit, push, or in other ways hurt other children (reverse scored) .003 −.094 .683

Note: Loadings in bold type face were retained as indicators of the factor.


Table 2Correlations among study variables

1 2 3 4 5 6 7 8

1. Peer social 1 12. Peer refuse 1 .25 13. Peer withdraw 1 .16 .16 14. BITSEA competence .45** .17 .08 15. BITSEA problem .38* .45** .44** .25 16. Peer social 2 .54** −.06 .18 .21 .01 17. Peer refuse 2 .18 .31+ .12 .17 .36+ −.17 18. Peer withdraw 2 −.36* −.37+ .02 −.05 −.03 −.24 −.13 1

+ p < .10.* p < .05.

** p < .01.

sions termed “peer sociability” (Cronbach’s alpha = .78; e.g. smile at peer, initiate play, imitate, take toys frompeers), “active peer refusal” (Cronbach’s alpha = .76; e.g. refuse peers’ attempts to play, turn back to peers, moveaway from peers), and “passive peer avoidance” (Cronbach’s alpha = .52; e.g. watch rather than participate, actas if he/she does not notice peer’s attempts to initiate play). Although the use of an oblique rotational procedure(Promax) allowed the three factors to correlate with each other, they were not significantly related to each other(Table 2).

3.2. Convergent validity

Correlations between each of the three dimensions of peer competence and the two subscales of the BriefInfant Toddler Social Emotional Assessment (BITSEA) were conducted to examine the convergent validity of thepeer competence scale. “Peer Sociability” correlated with both the competence (r = .45, p < .01) and the prob-lem (r = .38, p < .05) subscales of the BITSEA (Table 2). Both active peer refusal and passive peer withdrawalcorrelated with the problem subscale of the BITSEA (r = .45, p < .01, and r = .44, p < .01, respectively). Collec-tively this set of correlations indicates adequate convergent validity of the peer competence scale for infants andtoddlers.

3.3. Stability of individual differences

Correlations between individual scores from the first and second measurement occasions suggested that infants’rank order in peer sociability within the sample remained moderately stable during this 6-month period (r = .54, p < .01)(Table 2). On the other hand, rank order in active peer refusal remained less stable over time, while the rank order ofpassive peer avoidance did not remain stable at all.

3.4. Predictive validity

Hierarchical Linear Modeling (HLM) was then used to predict the competence and problem subscales ofthe BITSEA from the three dimensions of social competence with peers from the first measurement occasion(n = 32). Following convention in multi-level modeling, estimates are presented in unstandardized form, provid-ing substantive meaning about the inter-relationships between variables within and across levels (Raudenbush& Bryk, 2002). Peer sociability positively predicted later social competence (B = .04, p < .05), whereas activepeer refusal negatively predicted later social competence (Table 3) (B = −.09, p < .05). Passive peer avoidancedid not predict BITSEA competence later on (Table 3) (B = −.02, p = .59) Similarly, both passive avoidanceand active peer refusal predict higher scores on the problem subscale of the BITSEA at a later point intime (Table 3). Scores from the peer sociability scale did not predict later problem scores either (Table 3)(B = 00, p = .73).


Table 3Predictive validity of the peer competence dimensions

Social competence Social problems

Intercept 1.32** (.08) .15* (.04)Family SES .19* (.07) −.02 (.02)Infant peer sociability .04* (.01) .00 (.01)

Intercept 1.34** (.08) .15* (.04)Family SES .15+ (.08) −.01 (.02)Infant peer refusal −.09* (.04) .03* (.01)

Intercept 1.32** (.09) .15* (.03)Family SES .22* (.08) −.03 (.02)Infant peer avoidance −.03 (.05) .03* (.01)

Note: Due to small sample sizes three separate Hierarchical Linear Models were conducted, one with each factor of peer competence as a predictorvariable.Unstandardized beta coefficients are shown, with standard errors in parentheses.

+ p < .10.* p < .05.

** p < .01.

3.5. Individual and contextual antecedents

Assessments of social competence with peers from the second measurement occasion were examined for predictionsfrom children’s child care experiences, family socio-economic status, and individual temperament, each measuredduring the first occasion. Since these analyses examine theoretical contributions of antecedent variables to infants’later social competence scores, each analysis includes social competence scores from the first occasion as controlvariables.

3.5.1. Child care factorsSince children were naturally grouped within three child care centers it was helpful to examine the amount of

variability in each social competence dimension that was accounted for by these groupings prior to including predictorvariables representing children’s individual child care experiences. This provides an overview of how much variability ineach dimension of social competence could be attributed to characteristics of the care environment. This partitioning ofvariance through computation of intraclass correlation coefficients is an important first step when conducting theoreticalhypothesis testing in which individuals are nested within a group, such as child care classrooms (Raudenbush & Bryk,2002). Findings indicated that sociability demonstrated the greatest variability across child care classrooms (52% ofthe total variance), p < .10. In comparison, 37% (p < .10) of the variance in passive avoidance and 12% (p = .50) of thevariance in active refusal was attributed to differences between classrooms.

Consistent with model building procedures for HLM analyses (Raudenbush & Bryk, 2002), the independent vari-able (caregiver interruption of peer interaction) was then added to the model and was allowed to predict this childcare classroom-level variance (Table 4). Results indicated that the frequency of caregiver interruption from the firstmeasurement occasion negatively predicted peer sociability at the second occasion, controlling for peer sociability at

Table 4Longitudinal contributions of family and child care to peer competence


Intercept 11.46** (.52) 4.50** (.22) 9.30** (.39)Outcome at previous occasion .44** (.13) −.02 (.10) .00 (.11)Family SES .22 (.47) .55* (.22) .05 (.20)Caregiver interruption of peer interaction −5.35* (2.50) −.56 (1.04) −1.16 (.95)

Note: Hierarchical Linear Model with unstandardized beta coefficients. Standard errors in parentheses.* p < .05.

** p < .01.


Table 5Longitudinal contributions of family and infant characteristics to peer competence


Intercept 11.20** (.56) 4.47** (.22) 9.16** (.34)Outcome at previous occasion .57** (.14) −.01 (.11) −.01 (.11)Family SES .44 (.49) .57* (.22) .10 (.19)Infant surgency/extraversion 1.48 (1.19) .20 (.50) .41 (.48)

Intercept 11.28** (.56) 4.50** (.22) 9.24** (.40)Outcome at previous occasion .53** (.14) −.01 (.10) .03 (.10)Family SES .43 (.50) .57* (.21) .13 (.17)Infant negative affectivity −.51 (1.39) −.52 (.55) 1.22* (.46)

Intercept 11.27** (.56) 4.51** (.22) 9.21** (.35)Outcome at previous occasion .52** (.14) −.02 (.10) .01 (.11)Family SES .40 (.53) .49* (.22) .01 (.20)Infant regulation & orienting .16 (1.22) .69+ (.45) .59 (.44)

Note: Due to small sample sizes three separate Hierarchical Linear Models were conducted, one with each dimension of infant temperament as apredictor variable.Unstandardized beta coefficients are shown, with standard errors in parentheses.

+ p < .10.* p < .05.

** p < .01.

the first occasion and family socioeconomic status (B = −5.25, p < .05). This effect also accounted for all significantvariance between child care classrooms in peer sociability. Caregiver interruption did not predict children’s scores onactive peer refusal (B = −.56, p = .60) or passive peer avoidance (B = −1.16, p = .20).

3.5.2. Individual temperamentFurther HLM analyses employed three facets of individual temperament to predict individual differences in social

competence with peers (Table 5). Results indicated that infants’ temperamental patterns of negative affectivity fromthe first measurement occasion predicted their future scores on the passive avoidance dimension of peer competence(B = 1.22, p < .05), while orienting and regulation marginally predicted future scores on the active peer refusal dimensionof peer competence (B = .69, p < .10). These associations held while controlling for earlier scores of peer competenceand family socioeconomic status. Individual differences in peer sociability were not predicted by earlier patterns ofinfant temperament (unstandardized beta coefficients ranged from .44 to 1.40, all non-significant).

3.5.3. Family socio-economic statusPatterns of active peer refusal were further predicted by family socioeconomic status (Tables 4 and 5). Specifically,

higher SES at the first measurement occasion predicted higher rates of active peer refusal, while controlling for earlierpeer refusal. There were no significant differences in any of the three dimensions of peer competence between genders(unstandardized beta coefficients ranged from .17 to .56, all non-significant).

4. Discussion

Findings from the present study support the hypothesis that peer sociability and peer withdrawal are importantdimensions of infant and toddler peer interactions. Withdrawal included one factor termed active refusal and anotherlabeled passive avoidance. Each of the three dimensions of peer competence demonstrated adequate convergent andpredictive validity. As expected, negative and aggressive behaviors failed to represent a unique dimension of socialcompetence with peers. Also consistent with the proposed hypotheses in this study, peer sociability was predictedby child, family, and child care characteristics, whereas components of peer withdrawal (refusal and avoidance) werepredicted only by child and family characteristics. Collectively, the findings from the present study provide an importantstep toward developing a conceptualization of social competence that is appropriate for infants and toddlers, that iscapable of predicting subsequent patterns of social development, and that can be explained by antecedents from multiplecontexts.


The present results suggest that, whereas negative or aggressive interactive behavior may be an important indicatorof social functioning in the preschool and school-age years (Biovin et al., 2005; Coie & Dodge, 1998; Rubin et al.,1998), peer sociability may assume an important role throughout development. Peer sociability demonstrated stabilityin individual differences over time, and predicted social competence 6 months later. While previous research hasestablished peer sociability to be an important component of social success in the preschool years and beyond (Howes,1988; Rubin et al., 1998), the present findings extend this relationship to infants and toddlers.

The current pattern of findings stands in partial contrast to previous patterns found for children through 6 years ofage (Howes, 1988), which categorized peer behavior into social, hesitant and difficult dimensions. The current resultssuggest that the period of infancy and toddlerhood represents a unique period of development in peer competence. Inparticular, aggressive, negative behavior did not emerge as a unique dimension of peer competence from any of the fourfactor solutions considered. Instead, items depicting negative or aggressive interactive behavior loaded on the factorsof peer sociability, active avoidance of peers, and passive withdrawal from peer interaction. This pattern supportsrecent theories suggesting that aggression and other types of negative behavior may represent something qualitativelydifferent in the infant and toddler years, when it is first developing, than it does later in childhood (Brownell & Hazen,1999; NICHD ECCRN, 2001).

More specifically, the current findings suggest that these aggressive and negative behaviors represent early signs ofpeer sociability, avoidance, and refusal. For instance, taking another child’s toy appears to be an indicator of sociabilitywith peers. The item “Hit, bit, or hurt other children”, on the other hand, loaded negatively on the avoidance factor,seemingly representing an active avoidance strategy. Refusing peers’ attempts to interact loaded together with behaviorstypically considered withdrawal behaviors, to comprise a factor of peer refusal. Finally, crying easily around peersappears to signify another avoidance strategy. In sum, certain behaviors typically considered aggressive or negative,and indicative of poor social competence across childhood (Campbell et al., 2000; Howes, 1988; NICHD ECCRN,2001; for a review see Odom & Ogawa, 1992), may signify normative attempts to interact with peers, or to refuse andavoid social interaction, in infancy and toddlerhood.

Further examination of these dimension of peer sociability reflected adequate convergent validity with contem-poraneous measures of general social–emotional competence and social–emotional problems. Additionally, thesedimensions capture substantial individual variability that can be used to predict future social competence. Specifically,individual differences in all three dimensions predicted patterns of social–emotional competence and problem behav-iors 6 months later. Moreover, active peer refusal not only explained variance in later social–emotional problems, butalso predicted change (increase) in social–emotional problems over time. Therefore, even though social competencewith peers involves different dimensions during varying phases of the life span, individual differences in peer com-petence from one time may be used to predict later patterns of peer competence. Termed “heterotypic continuity”(Caspi & Roberts, 2001), changes within the structure of a behavioral construct with age is a common developmen-tal phenomenon; however, continuity in the construct itself may remain. The current study suggests that heterotypiccontinuity may apply to the study of social competence with peers.

These findings have important implications for early identification of children at risk for later social difficulties. Whilethe current results indicate some change patterns of social interaction between infancy and toddlerhood, there is also sub-stantial stability and predictive validity. Examination of potential sources of individual differences revealed that facetsof children’s experiences in child care, family demographics, and individual temperaments may be important contribu-tors to peer competence. Of particular interest in the current study is the finding that the rate at which caregivers interruptinfants early peer interactions appears to have lasting negative effects on peer sociability 6 months later. Future researchshould employ longer term longitudinal designs to examine predictive validity into early childhood, as well as to exploredifferential impacts of various forms of caregiver interruption, such as peer-related discourse and distraction from peers.

In addition to caregiver behaviors, individual and family characteristics also helped to explain variation in infants’ andtoddlers’ social competence with peers. Specifically, children with higher negative affectivity exhibited more passiveavoidance of their peers 6 months later. In addition, a marginally significant relationship between regulation/orientingand active peer refusal was also detected, such that children with higher levels of regulation and orienting tended torefuse their peers’ attempts to interact with them 6 months later. To help explain this finding we conducted post hocanalyses utilizing the subscales of the broad regulation and orienting temperament dimension1. Results suggested that

1 For more information regarding these post hoc analyses please contact the first author.


the subscales labeled low intensity pleasure and duration of orienting were the only two that significantly predicted anincrease in later active peer refusal (the soothability and cuddliness subscales had no relation to peer refusal). In essence,then, these findings suggest that children that enjoy activities low in stimulation (such as playing by themselves) andthat attend to a single object for extended periods of time when they are 1 year of age tend to refuse peer interactionmore often than do other children as they develop, perhaps because they prefer to play by themselves. Given the smallsample size and moderate significance of this effect, however, these findings should be replicated and investigatedfurther in future studies.

Finally, children from more advantageous socioeconomic backgrounds demonstrated higher levels of peer refusal.This finding was unexpected in light of previous research documenting that more economic resources facilitate socia-bility, in part because parents with more advantages set up play dates with other children and help structure early peerinteractions (for reviews see Park & Buriel, 1998; Rubin et al., 1998). However, recent work suggests that family SESmay relate differently to children’s social competence in classrooms or schools than it does in other settings (Schneider,Richard, Younger, & Freeman, 2000). In light of the present findings then, although children from families with higherSES may exhibit more sociability in home and neighborhood settings with small groups of peers, they may actuallyrefuse peer interaction more often than others in child care settings, perhaps because of the larger size of peer groupsand the absence of the parental scaffolding to which they have become accustomed.

Overall, the findings from the present study suggest that negative and aggressive interactive behavior in infancyand toddlerhood fails to represent a unique dimension of social competence. Rather, patterns of peer interaction atthese ages can be characterized by sociability with peers, active refusal of peer interaction, and passive avoidance ofinteraction with peers. Further, each of these three dimensions of peer competence appears to demonstrate its owndevelopmental antecedents. While peer sociability is sensitive to caregiver interruption of peer interaction, active peerrefusal and passive peer avoidance are more closely tied to individual temperament. Active peer refusal can also beexplained, in part, by family socioeconomic status.

While the present findings help to advance our understanding of early peer competence, the relatively small samplesize limits generalizeability and should be replicated with larger samples. Further longitudinal research with a largersample size is also needed to understand the developmental period in which negative and aggressive social behaviorsbegin to detract from social competence. Specifically, more systematic investigations of both aggressive/negativebehaviors and dimensions of social competence more relevant to infants and toddlers are necessary to fully understandthis important developmental course. Additionally, given the current findings suggesting that caregivers assume animportant role in shaping future peer behaviors, further research should examine antecedents at the classroom level,such as caregiver training and caregiver attitudes about child development. These studies may also find it beneficial torevise the subscale representing active peer avoidance to improve the internal consistency of this measure.

Acknowledgements

Support for this research has come from multiple sources including the University of California, Davis Jastro SheildsResearch Award and the University of California, Davis Cota Robles Fellowship, both awarded to the first author. Theauthors acknowledge the cooperation of the families, faculty, and staff of the Center for Child and Family Studies in theDepartment of Human and Community Development at the University of California, Davis in conducting this study.

References

Baumrind, D. (1968). Manual for the preschool behavior Q sort. Berkeley: University of California, Institute of Human Development.Bayer, C. L., Whaley, K. L., & May, S. E. (1995). Strategic assistance in toddler disputes: Sequences and patterns of teachers’ message strategies.

Early Education and Development, 6, 405–432.Biovin, M., Vitaro, F., & Poulin, F. (2005). Peer relationships and the development of aggressive behavior in early childhood. In R. E. Tremblay, W.

W. Hartup, & J. Archer (Eds.), Developmental origins of aggression. New York, NY: The Guilford Press.Briggs-Gowan, M. J., & Carter, A. S. (2001). The brief infant-toddler social and emotional assessment (BITSEA), unpublished measure.Bronson, W. C. (1981). Toddlers’ behaviors with agemates: Issues of interaction, cognition, and affect. Monographs on Infancy: Vol. 1, NJ: ABLEX.Brownell, C., & Brown, E. (1992). Peers and play in infants and toddlers. In V. Van Hasselt, & M. Hersen (Eds.), Handbook of social development

(pp. 183–200). NY: Plenum.Brownell, C. A., & Hazen, N. (1999). Early peer interaction: A research agenda. Early Education & Development, 10, 403–413.Burks, V. S., Dodge, K. A., & Price, J. M. (1995). Models of internalizing outcomes of early rejection. Development and Psychopathology, 7,

683–696.


Cairns, R. B., & Cairns, B. D. (2000). The natural history and developmental functions of aggression. In A. J. Sameroff, M. Lewis, & S. M. Miller(Eds.), Handbook of developmental psychopathology. New York, NY: Kluwer Academic.

Calkins, S., & Fox, N. A. (1992). The relations among temperament, security of attachment and behavioral inhibition at 24 months. Child Development,63, 1456–1472.

Campbell, S. B. (2002). Peer relationships and young children’s development. In Behavior problems in preschool children: Clinical and developmentalissues (pp. 172–202). New York: Guilford Press.

Campbell, J. J., Lamb, M. E., & Hwang, C. P. (2000). Early child-care experiences and children’s social competence between 1 1/2 and 15 years ofage. Applied Developmental Science, 4, 166–175.

Carter, A. S., & Briggs-Gowan, M. J. (2000). Infant-toddler social and emotional assessment (ITSEA)—Manual. New Haven, CT: Yale University,Department of Psychology.

Caspi, A., & Roberts, B. W. (2001). Personality development across the life course: The argument for change and continuity. Psychological Inquiry,12, 49–66.

Cattell, R. B. (1966). The screen test for the number of factors. Multivariate Behavioral Research Monographs, 72–81.Chen, D. W. (2003). Preventing violence by promoting the development of competence conflict resolution skills: Exploring roles and responsibilities.

Early Childhood Education Journal, 30, 203–208.Coie, J. K., & Dodge, K. A. (1998). Aggression and antisocial behavior. In W. Damon (Series Ed.), N. Eisenberg (Vol. Ed.), Handbook of child

psychology: Vol. 3. Social, emotional, and personality development (5th ed., pp. 779–862). NY: John Wiley & Sons.Coie, J. D., Dodge, K., & Kupersmidt, J. B. (1990). Peer group behavior and social status. In S. R. Asher, & J. D. Coie (Eds.), Peer rejection in

childhood (pp. 17–59). New York: Cambridge University Press.Coie, J. D., Terry, R., Lenox, K., Lochman, J., & Hyman, C. (1995). Childhood peer rejection and aggression as predictors of stable patterns of

adolescent disorder. Development and Psychopathology, 7, 697–714.Eckerman, C., & Didow, S. (1996). Nonverbal imitation and toddlers’ mastery of verbal means of achieving coordinated action. Developmental

Psychology, 32, 141–152.Eckerman, C. O., Whatley, J. L., & McGehee, L. J. (1979). Approaching and contacting the object another manipulates: A social skill of the

1-year-old. Developmental Psychology, 15, 585–593.Fitzgerald, Mann, Cabrera, & Wong, 2003. Diversity in caregiving contexts. In R. M. Lerner (Series Ed.), M. A. Easterbrooks (Vol. Ed.), Handbook

of psychology: Vol. 6. Developmental psychology (pp. 135–167). NY: Wiley & Sons.Garstein, M. A., & Rothbart, M. K. (2003). Studying infant temperament via the revised infant behavior questionnaire. Infant Behavior and

Development, 26, 64–86.Greenwood, C., Walker, H., Todd, N., & Hops, H. (1979). Selecting a cost effective screening measure for the assessment of preschool social

withdrawal. Journal of Applied Behavior Analysis, 12, 639–652.Goncu, A., & Cannella, V. (1996). The role of teacher assistance in children’s construction of intersubjectivity during conflict resolution. In M.

Killen (Ed.), Children’s autonomy, social competence, and interactions with adults and other children: Exploring connections and consequences(pp. 57–70). San Francisco, CA: Jossey-Bass Inc.

Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.Harper, L. V., & McClusky, K. S. (2003). Teacher-child and child-child interactions in inclusive preschool settings: Do adults inhibit peer interactions?

Early Childhood Research Quarterly, 18, 163–164.Hay, D. F. (1985). Learning to form relationships in infancy: Parallel attainments with parents and peers. Developmental Review, 5, 122–161.Hay, D., & Ross, H. (1982). The social nature of early conflict. Child Development, 53, 105–113.Howes, C. (1980). Peer Play Scale as an index of complexity of peer interaction. Developmental Psychology, 16, 371–372.Howes. (1988). Peer interaction in young children. Monographs of the Society for Research in Child Development, 53.Howes, C., Phillips, D. A., & Whitebook, M. (1992). Thresholds of quality: Implications for the social development of children in center-based

child care. Child Development, 63, 449–460.Kagan, J. (1989). Temperamental contributions to social behavior. American Psychology, 44, 668–674.Kagan, J. (1998). Biology and the child. In W. Damon (Series Ed.), N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional,

and personality development (5th ed. pp. 177–236). New York: Wiley.Kontos, S. (1999). Preschool teachers’ talk, roles, and activity settings during free play. Early Childhood Research Quarterly, 14, 363–382.Lamb, M. E. (1998). Nonparental child care: Context, quality, correlates, and consequences. In: W. Damon (Series Ed.), I. Sigel, K. A. Renninger

(Vol. Eds.), Handbook of child psychology: Child psychology in practice: Vol. 4 (pp. 73–133). New York: Wiley.Ladd, G. W. (1999). Peer relationships and social competence during early and middle childhood. Annual Review of Psychology, 50, 333–359.McDougall, Hymel, Vaillancourt, & Mercer. (2001). The consequences of early peer rejection. In M. Leary (Ed.), Interpersonal rejection (pp.

213–247). New York: Oxford University Press.NICHD Early Child Care Research Network. (2001). Child care and children’s peer interaction at 24 and 36 months: The NICHD study of early

child care. Child Development, 72, 1478–1500.NICHD Early Child Care Research Network. (2002). Child-care structure, process, outcome: Direct and indirect effects of child-care quality on

young children’s development. Psychological Science, 13, 199–206.NICHD Early Child Care Research Network. (2004). Trajectories of physical aggression from toddlerhood to middle childhood. Monographs of the

Society for Research in Child Development, 69.Odom, S. L., & Ogawa, I. (1992). Direct observation of young children’s social interaction with peers: A review of methodology. Behavioral

Assessment, 14, 407–441.Park, R. D., & Buriel, R. (1998). Socialization in the family: Ethnic and ecological perspectives. In W. Damon (Series Ed.), N. Eisenberg (Vol. Ed.),

Handbook of child psychology: Vol. 3. Social, emotional, and personality development (5th ed. pp. 619–700). New York: Wiley.


Peisner-Feinberg, E. S., Burchinal, M. R., Clifford, R. M., Culkin, M. L., Howes, C., Kagan, S. L., et al. (2001). The relation of preschool child-carequality to children’s cognitive and social developmental trajectories through second grade. Child Development, 72, 1534–1553.

Raudenbush, S. W., & Bryk, A. S. (2002). Heirarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA:Sage.

Rauh, H. (1987). Social development in infant peers. In H. Rauh, & H.-Ch. Steinhausen (Eds.), Psychobiology and early development. North-Holland:Elsevier.

Rosenthal, M. K. (1991). The relation of peer interaction among infants and toddlers in family day care to characteristics of the child care environment.Journal of Reproductive and Infant Psychology, 9, 151–167.

Ross, H. R., & Lollis, S. P. (1987). Communication within infant social games. Developmental Psychology, 23, 241–248.Rubin, K. H., Bukowski, W., & Parker, J. G. (1998). Peer interactions, relationships, and groups. In W. Damon (Series Ed.), N. Eisenberg (Vol. Ed.),

Handbook of child psychology: Vol. 3. Social, emotional, and personality development (5th ed. pp. 619–700). New York: Wiley.Sanson, A., Hemphill, S. A., & Smart, D. (2004). Connections between temperament and social development: A review. Social Development, 13,

142–170.Schneider, B., Richard, J., Younger, A., & Freeman, P. (2000). A longitudinal exploration of children’s social participation and social withdrawal

across socioeconomic status levels and social settings. European Journal of Social Psychology, 30, 497–519.Shantz, C. U. (1987). Conflicts between children. Child Development, 58, 283–305.Skarpness, L. R., & Carson, D. K. (1986). Temperament, communicative competence and the psychological adjustment of kindergarten children.

Psychological Reports, 59, 1299–1306.Stocker, C., & Dunn, J. (1990). Sibling relationships in childhood: Links with friendships and peer relationships. British Journal of Developmental

Psychology, 8, 227–244.Suomi, S. J. (2000). A biobehavioral perspective on developmental psychopathology. In A. J. Sameroff, M. Lewis, & S. M. Miller (Eds.), Handbook

of developmental psychopathology. New York, NY: Kluwer Academic.Tremblay, R. E. (2000). The development of aggressive behavior during childhood: What have we learned in the past century? International Journal

of Behavioral Development, 24, 129–141.Tremblay, R. E., & Nagin, D. S. (2005). The developmental origins of physical aggression in humans. In R. E. Tremblay, W. W. Hartup, & J. Archer

(Eds.), Developmental origins of aggression. New York, NY: The Guilford Press.Vandell, D., & Wilson, K. (1987). Infants’ interactions with mother, sibling, and peer: Contrasts and relations between interaction systems. Child

Development, 58, 176–186.Vaughn, B. E., Vollenweider, M., Bost, K. K., Azria-Evans, M. R., & Snider, J. B. (2003). Negative interactions and social competence for preschool

children in two samples: Reconsidering the interpretation of aggressive behavior for young children. Merrill-Palmer Quarterly, 49, 245–327.

Documents

Reformulating infant and toddler social competence with peers