Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Sep 22.
Published in final edited form as: Int J Behav Dev. 2007 May 1;31(3):274–283. doi: 10.1177/0165025407076440

Peer social preference and depressive symptoms of children in Italy and the United States

Jennifer E Lansford a, Cristina Capanna b, Kenneth A Dodge a, Gian Vittorio Caprara b, John E Bates c, Gregory S Pettit d, Concetta Pastorelli b
PMCID: PMC2748932  NIHMSID: NIHMS139547  PMID: 19777082

Abstract

This study examined the role of low social preference in relation to subsequent depressive symptoms, with particular attention to prior depressive symptoms, prior and concurrent aggression, mutual friendships, and peer victimization. Italian children (N = 288) were followed from grade 6 through grade 8, and American children (N = 585) were followed from kindergarten through grade 12. Analyses demonstrate that low social preference contributes to later depressive symptoms. The effects are not accounted for by depressive symptoms or aggression experienced prior to low social preference but are mostly accounted for by the co-occurrence of depressive symptoms with concurrent aggressive behavior; gender, mutual friendships, and peer victimization generally did not moderate these associations. We conclude that peer relationship problems do predict later depressive symptoms, and a possible mechanism through which this effect occurs is through the effect of poor peer relationships on increasing aggressive behavior, which is associated with depressive symptoms.

Keywords: aggressive behavior, depressive symptoms, peer rejection, social preference

A large body of research shows that children who are rejected by their peers are more likely than non-rejected children to show social, emotional, and behavioral problems. It has become apparent that peer rejection is correlated with a range of problems such as antisocial behavior (Bierman & Wargo, 1995), social withdrawal (Rubin & Mills, 1988), loneliness (Asher, Parkhurst, Hyrnel, & Williams, 1990), and delinquency and dropping out of school (Parker & Asher, 1987), either because peer rejection is an early marker of underlying problems that lead to subsequent maladjustment or because peer rejection plays a causal role in the development of such problems (see Newcomb, Bukowski, & Pattee, 1993; Parker & Asher, 1987, for reviews). Using data from two independent multi-site community samples followed prospectively for up to five years, Dodge et al. (2003) found that children rejected by their peers early in elementary school were more likely to be aggressive later in elementary school according to their teachers, even after controlling for prior aggression and other factors. These findings suggest that peer rejection serves not only as an early marker of problems that become manifest as subsequent aggression, but that peer rejection itself contributes to an increase in aggressive behavior over time.

Although these findings implicate peer rejection as a causal factor in the development of children's subsequent externalizing behavior problems, the findings with regard to the role of peer rejection in relation to internalizing problems are less clear (Parker & Asher, 1987). Several studies have found significant associations between problematic peer relationships and depressive symptoms (e.g., Patterson & Stoolmiller, 1991). However, these studies often have not taken into account potentially important contributions of prior internalizing problems or concurrent externalizing problems in examining these links. Although at face value it makes sense to think that being rejected by peers could cause children to become depressed, because of the social stigma of rejection or because of lack of opportunities for fun and companionship at school, the mechanisms might not actually be that direct. Thus, the overarching goal of this study was to clarify the role of peer rejection in relation to the development of children's depressive symptoms, with particular attention to children's prior depressive symptoms and concurrent aggression.

There is evidence that peer rejection both predicts and is predicted by social withdrawal (e.g., Rubin, Le Mare, & Lollis, 1990). Because of evidence that depressed children are at risk of being rejected by their peers (e.g., Blechman, McEnroe, Carella, & Audette, 1986; Levendosky, Okun, & Parker, 1995), one important question is the extent to which peer rejection causes subsequent depressive symptoms above and beyond depressive symptoms that may have resulted in peer rejection in the first place. In other words, if peer rejection remains a statistically significant predictor of subsequent depressive symptoms after controlling for depressive symptoms that temporally precede peer rejection, this suggests that depressed children are not merely being rejected by their peers but that peer rejection itself may contribute to an increase in depressive symptoms over time.

Not only is it important to consider the role of prior depressive symptoms, it is also important to consider the role of prior and concurrent aggressive behavior. Because aggression and depression often co-occur (e.g., Angold & Costello, 1993), only studies that include both constructs are able to assess whether peer rejection is associated specifically with depression or whether peer rejection is associated with maladaptation (including both internalizing and externalizing components) more generally. This is an important question because mechanisms through which peer rejection affects children's adjustment might be different if children's adjustment is affected uniformly versus affected in specific behavioral domains but not others.

Several cross-sectional studies have differentiated aggressive-rejected from nonaggressive-rejected youth. Boivin, Poulin, and Vitaro (1994) found in a sample of 214 fourth-grade children that although withdrawn-rejected, aggressive-rejected, and aggressive-withdrawn-rejected children had higher scores on the Children's Depression Inventory (CDI) than did nonrejected children, rejected children who were neither aggressive nor withdrawn did not differ from non-rejected children on the CDI, suggesting that withdrawal and aggression rather than rejection were the key predictors of depressive symptoms. Hecht, Inderbitzen, and Bukowski (1998) found in a cross-sectional sample of 1,687 students in grades 4, 6, 7, 8, 9, and 11 that although there were no sociometric status differences on global depression assessed by the CDI, aggressive-rejected children reported more interpersonal problems and feelings of ineffectiveness on this measure whereas neglected and submissive-rejected children reported more anhedonia, again suggesting that rejection per se was not the main predictor of depressive symptoms. In a cross-sectional study of 1,464 fourth-graders, Cole and Carpentieri (1990) found that rejected children had higher self-, peer-, and teacher-reported depression scores than did popular, neglected, or average status children. However, the overlap between depression and conduct disorder was high; children who exhibited depression with no conduct disorder were not more likely to be socially rejected, whereas children who exhibited depression in conjunction with conduct disorder and children with conduct disorder alone were both more likely to be socially rejected (Cole & Carpentieri, 1990). Taken together, these cross-sectional studies suggest that depressive symptoms are associated with problems related to peer rejection (such as aggression and conduct disorder) rather than to peer rejection per se.

Short-term longitudinal studies have also examined links among peer rejection, internalizing, and externalizing behaviors. In a study of adolescents assessed in grades 8 and 10, French, Conrad, and Turner (1995) found that rejected-antisocial adolescents had elevated levels of depression, whereas rejected-nonantisocial adolescents did not; in comparisons with antisocial and nonantisocial accepted peers, antisocial behavior rather than rejected status appeared more predictive of internalizing problems. In a sample of 521 children in grades 3, 4, and 5 who were assessed three times over the course of one year, Panak and Garber (1992) found that although peer rejection at time 1 did not predict depressive symptoms at time 2, increases in peer rejection between time 1 and time 2 were related to depressive symptoms at time 2, controlling for depressive symptoms at time 1. Furthermore, the link between increases in aggression between times 1 and 2 and depressive symptoms at time 2 was partially mediated by increases in peer rejection. However, depressive symptoms at time 1 were the only significant predictor of time 3 depressive symptoms (Panak & Garber, 1992). Taken together, these findings suggest that children who are rejected by their peers may be depressed not because of peer rejection per se, but because of other factors that are linked to both peer rejection and depression (e.g., aggression, withdrawal).

Friendships and peer victimization

In the literature, researchers have made the theoretical and empirical distinction between children's dyadic relationships and relationships with peers in the broader peer group (Asher, Parker, & Walker, 1996; Bukowski & Hoza, 1989; Ladd, 1999; Parker & Asher, 1993). For example, acceptance by the peer group can provide adolescents with a sense of inclusion or belonging to a larger group, whereas friendships, which are inherently dyadic, can provide reciprocal affection, intimate disclosure, validation, and support between two individuals (Asher et al., 1996; Bukowski & Hoza, 1989; Ladd, Kochenderfer, & Coleman, 1997). In addition to encompassing both similar and distinct relational provisions or features (Ladd, 1999), dyadic and group relationship domains have been found to make unique and overlapping contributions in the prediction of child adjustment (Ladd et al., 1997; Parker & Asher, 1993; Vandell & Hembree, 1994). Parker and Asher (1993) found that friendship, friendship quality, and peer acceptance each explained a unique percentage of variance in the prediction of children's loneliness. In contrast, Ladd and his colleagues (1997) found unique and redundant linkages between three forms of children's relationships (i.e., friendship, peer acceptance, and peer victimization) and child adjustment.

Whereas mutual friendships are important dyadic aspects of peer relationships, peer victimization is a group-level aspect of peer relationships that provides additional information regarding children's social adjustment beyond looking only at peer rejection. Peer victimization is characterized by physical or verbal abuse within the peer group, which is not necessarily part of rejected children's experience. Victimization and social preference, then, may make independent contributions to children's adjustment. For example, Schwartz, McFadyen-Ketchum, Dodge, Pettit, and Bates (1998) found that children who were victimized by the peer group in grade 3 or 4 showed more externalizing, attention dysregulation, and immature/dependent behavior two years later, incrementing the prediction afforded by knowledge of peer rejection alone.

Thus, the growing evidence seems to point toward shared and unshared relational provisions or features of children's different peer relationships as well as their linkages to child adjustment. In the context of the present study, mutual friendships and peer victimization were examined as possible moderators of the link between social preference and subsequent internalizing problems. One could hypothesize, for example, that children low in social preference may be at particular risk for subsequent depressed affect if they also lack mutual friendships or are victimized by peers. On the other hand, children low in social preference may be protected from subsequent depressed affect if they have mutual friendships (perhaps because key social needs could be met in this dyadic rather than a group context). Indeed, Kiesner, Poulin, and Nicotra (2003) found in a study of Italian 12-year-olds that being nominated as being a member of a social network after school served a protective function in offsetting negative effects of low inclusion within in-school networks on depressive symptoms. Simultaneous consideration of different aspects of peer relationships will have the potential to inform understanding of additional risk or protective factors that children may encounter within their peer relationships.

Perceptions of different reporters

The issue of who the respondents are is also an important one. For example, in a cross-sectional study of 96 adolescent psychiatric inpatients with suicidal ideation, higher levels of perceived peer rejection were associated with more depressive symptoms and severe suicidal ideation, but because the adolescents reported both on their perceptions of peer rejection and on their psychiatric symptoms, the possibility that the symptoms affected the adolescents' perceptions of peer rejection cannot be eliminated (Prinstein, Boergers, Spirito, Little, & Grapentine, 2000). In a direct comparison of results from different reporters, Reinherz, Giaconia, Hauf, Wasserman, and Paradis (2000) found that although self-reported low peer acceptance at age 9 predicted depression assessed through the Diagnostic Interview Schedule at age 21, teacher- and parent-reported peer acceptance was unrelated to subsequent depression. Boivin, Hymel, and Bukowski (1995) reported that elementary school children's perceptions of peer rejection mediate the association between peer-reported rejection and children's depressed mood (see also Panak & Garber, 1992). Because of these differences in the perceptions of different reporters, it is important to include multiple informants in studies of links between peer rejection and subsequent depressive symptoms.

The role of culture

Most research on links between peer social preference and children's adjustment has been conducted using North American samples. The present study included comparable data from long-term longitudinal studies in the United States and Italy to extend the findings to a cultural context that differs in some important ways from the North American contexts represented in the majority of previous research. For example, because Italian students spend the entire school day with the same peers throughout middle school (grades 6, 7, and 8) rather than moving to different peer contexts as they change classes within and across school years in American middle schools, the importance of peer rejection as a predictor of children's adjustment during middle school may be more pronounced in Italy than in the United States (Kiesner, 2002). On the other hand, because of the greater emphasis placed on the importance of the family in Italy compared to the United States (see Attili, Vermigli, & Schneider, 1997), peer rejection may be less related to children's subsequent adjustment in Italy than in the United States. For example, Young and Ferguson (1981) found that a number of adolescent boys in Italy expressed a preference for playing alone rather than with peers, whereas almost no Italian immigrant boys in Boston expressed this preference; one-third of the parents in Palermo, Italy indicated mat they preferred to have their sons spend time at home rather than with friends.

A small number of studies have examined correlates of peer rejection in Italian samples. For example, as in North American samples described above, peer rejection has been linked to aggressive behavior and social withdrawal in Italian samples (Attili et al., 1997; Caprara & Pastorelli, 1993; Casiglia, Lo Coco, & Zappulla, 1998). In a study that focused specifically on peer rejection as a predictor of subsequent depressive symptoms, Kiesner (2002) found in a sample of 215 Italian students in grades 6 and 7 who were followed for two years that peer-rated sociometric status at time 1 significantly-predicted self-reported depressive symptoms one year later, controlling for time 1 depressive symptoms and teacher-reported disruptive and argumentative behavior.

Summary

To summarize, our main research questions were as follows: (a) Do children who are low in peer social preference exhibit subsequent depressive symptoms up to several years later?; (b) Do longitudinal associations between peer social preference and children's subsequent depressive symptoms remain after controlling for prior depressive symptoms and prior and concurrent aggression?; (c) Do American and Italian children differ in the way that social preference relates to their subsequent depressive symptoms? (d) Do mutual friendships or peer victimization alter the role that peer social preference plays in the development of children's depressive symptoms?

Study 1

Method

Participants

Participants included 288 children (50% boys, 50% girls) from the ongoing Genzano Study who provided data when they were in grade 8 (age 13 years). When participants were in grade 6 (age 11 years) they were recruited in two cohorts from two middle schools in Genzano, a community near Rome, Italy. Informed consent was obtained from 92% of the families who were recruited (consenting N = 411). Follow-up assessments were conducted annually through grade 8 (age 13 years), with 70% of the original sample providing grade 13 data on the variables examined in the present study. Compared to the original sample of 411, the 288 participants who provided data for the present study did not differ by gender, peer social preference, peer evaluation of depression, and self evaluation of depressive symptoms in grade 6. Participating children's parents were socio economically diverse (14% professional or managerial, 36% merchants, 15% skilled workers, 33% unskilled workers, 2% retired).

Procedure and measures

Social preference

To assess peer liking, we asked children's peers in grades 6 and 7 to nominate three classmates with whom they would like to play and three classmates with whom they would like to study or do homework. To assess peer disliking, we asked children's classroom peers to nominate three classmates with whom they would not like to play and three classmates with whom they would not like to study or do homework. Nominations on these four sociometric items were standardized within classrooms, and a standardized social preference score was obtained for each child by subtracting the average of the two standardized dislike items from the average of the two standardized like items. A composite score was created by averaging the social preference scores across the two years (r = .64, p < .001).

Depressive symptoms and aggression

Measures of children's depressive symptoms were obtained from their teachers, mothers, peers, and from the children themselves. In grade 8, teacher reports were correlated .24, .22, and .20 with mother, peer, and self reports, respectively; mother reports were correlated .15 and .32 with peer and self reports, respectively; peer reports were correlated .28 with self reports. For each child, grade 6 and 8 teachers completed the 18-item anxiety/depression scale of the Teacher Report Form (TRF) of the Child Behavior Checklist (Achenbach, 1991a). For each item, teachers indicated whether the statement describing a particular behavior (e.g., crying, being anxious) was not true (0), somewhat or sometimes true (1), or very or often true (2) of the child. Teachers also completed a 10-item scale that measures children's depressive mood, disconsolateness, and hopelessness (e.g., how often during the last week the child appeared sad or hopeless; 1 = never, 2 = a few times, 3 = very often; Bandura, Pastorelli, Barbaranelli, & Caprara, 1999). When children were in grades 6 and 8, mothers completed the anxiety/depression scale of the CBCL (Achenbach, 1991b), comparable to the TRF described above, and a 10-item scale comparable to the teacher-report version assessing children's depressive mood, disconsolateness, and hopelessness (1 = never, 2 = a few times, 3 = very often). Alphas for the teacher- and mother-reported depressive symptoms ranged from .85 to .93 across respondents and years. Grade 6 and 8 peers were asked to nominate three classmates who are often sad, who are often discouraged, and who often appear unhappy; nominations were standardized, averaged, and aggregated to obtain the peer measure of depressed affect.

In grades 6 and 8, children completed a standard Italian version of the 27-item Children's Depression Inventory (Camuffo, Cerutti, Lucarelli, & Mayer, 1988; Kovacs, 1985), which measures feelings of despondency, hopelessness, loss of appetite and interest in activities, self-deprecation, and suicidal ideation. Children rated whether they had experienced each symptom on a 3-point scale (0 = sometimes, 1 = many times, 2 = always); items were averaged to create a scale (α = .83 and .90 in grades 6 and 8, respectively). In grades 6 and 8, children also completed the Youth Self-Report (Achenbach, 1991c), comparable to the TRF described above; alphas ranged from .81 to .86.

Teacher-, mother-, and youth-reported aggression was measured using the Achenbach (1991a, 1991b, 1991c) Teacher Report Form, Child Behavior Checklist, and Youth Self-Report described above, respectively, when children were in grades 6 and 8. Alphas for the aggression scales across reporters and years ranged from .88 to .95. Peers in grades 6 and 8 nominated classmates who kick and punch others, hurt others, and abuse others; scores were standardized, averaged, and aggregated to create scales reflecting peer-reported aggression.

Results

As a first step in examining links between peer social preference and subsequent depressive symptoms, we first calculated bivariate correlations between the continuous social preference scores children received from peers in grades 6 and 7 (averaged across years) and children's depressive symptoms in grade 8 (r = −.22, .10, −.38, −.23 for teacher, mother, peer, and self reports, respectively; mother report ns, others p < .001). However, because it is possible that children who demonstrate depressive symptoms are subsequently less preferred by their peers, and that depressive symptoms are therefore a precursor rather than a consequence of low social preference, and because aggression and depressive symptoms are moderately correlated with one another and often co-occur, links between low social preference and depressive symptoms could reflect the importance of problem behaviors more generally rather than depressive symptoms specifically.

To evaluate whether the previously found link between social preference and subsequent depressive symptoms remained after controlling for prior depressive symptoms and prior aggression, we conducted a series of regressions in which effects of depressive symptoms and aggression in grade 6 were entered at step 1 and social preference in grades 6 and 7 was entered at step 2 to predict depressive symptoms in grade 8, separately for the different reporters. Preliminary analyses examining whether gender moderated the association between social preference and subsequent depressive symptoms revealed no significant gender by social preference interactions; therefore the results reported below do not include gender in the models. As shown in Table 1, social preference significantly predicted teacher-, peer-, and self-reported (but not mother-reported) depressive symptoms in grade 8, controlling for prior depressive symptoms and prior aggression. We then re-ran the regressions, controlling for concurrent aggression (assessed in grade 8) instead of prior aggression; correlations between prior and concurrent teacher-, mother-, peer-, and self-reported aggression were .36, .69, .66, and .56, respectively. As shown in Table 2, social preference predicted mother- and peer-reported (but not teacher- or self-reported) subsequent depressive symptoms.

Table 1. Regressions predicting depressive symptoms in grade 8, controlling for prior depressive symptoms and prior aggression: Italian sample.

Predictors and test statistics Teacher Mother Peer Self
Step 1
 Prior depressive symptoms .39*** .57*** .53*** .47***
 Prior aggression .08 −.04 −.04 .01
 F 22.03*** 22.28*** 55.50*** 42.27***
 R2 .16*** .31*** .28*** .23***
Step 2
 Prior depressive symptoms .35*** .58*** .40*** .45***
 Prior aggression .03 −.01 −.23*** .01
 Social preference −.15* .16 −.36*** −.11*
 F 16.44*** 16.43*** 53.09*** 29.89***
 R2 .02* .02 .08*** .01*
Total R2 .18 .33 .36 .24
Open in a new tab

Note. Numbers are standardized betas and test statistics from the regressions.

*

p < .05;

**

p < .01;

***

p < .001.

Table 2. Regressions predicting depressive symptoms in grade 8, controlling for prior depressive symptoms and concurrent aggression: Italian sample.

Predictors and test statistics Teacher Mother Peer Self
Step 1
 Prior depressive symptoms .37*** .48*** .52*** .35***
 Concurrent aggression .30*** .17 −.07 .42***
 F 36.74*** 24.27*** 56.50*** 90.73***
 R2 .25*** .33*** .29*** .39***
Step 2
 Prior depressive symptoms .35*** .50*** .41*** .33***
 Concurrent aggression .28*** .16 −.20*** .41***
 Social preference −.07 .17* −.31*** −.09
 F 24.90*** 18.25*** 52.20*** 62.00***
 R2 .00 .03* .07*** .01
Total R2 .25 .36 .36 .40
Open in a new tab

Note. Numbers are standardized betas and test statistics from the regressions.

*

p < .05;

**

p < .01;

***

p < .001.

Findings from the Genzano Study suggest that children low in peer social preference in grades 6 and 7 are more likely to be rated by peers as having depressive symptoms in grade 8, even controlling for prior depressive symptoms, prior aggression, and concurrent aggression. However, links between peer social preference and teacher-, mother-, and self-reported depressive symptoms could be accounted for by prior depressive symptoms and concurrent aggression. Boys and, girls did not differ with respect to how social preference was related to subsequent depressive symptoms. We next sought to replicate and extend these findings using data from an American sample.

Study 2

Method

Participants

Children in this study were participants in the ongoing Child Development Project, a multi-site longitudinal investigation of children's adjustment (see Dodge, Bates, & Pettit, 1990; Dodge et al., 2003). Participants were recruited in two cohorts when children entered kindergarten in 1987 or 1988 at three sites: Knoxville and Nashville, Tennessee and Bloomington, Indiana. Parents were approached at random during kindergarten pre-registration and asked if they would participate in a longitudinal study of child development. About 15% of children at the targeted schools did not pre-register. These participants were recruited on the first day of school or by letter or telephone. Of those asked, approximately 75% agreed to participate. The sample consisted of 585 families at the first assessment (52% boys, 48% girls; 81% European American, 17% African American, 2% other ethnicities). The families' Hollingshead (1979) index of socioeconomic status at age 5 ranged from 11 to 66 (M = 40.35, SD = 14.38). Follow-up assessments were conducted annually through grade 12, with 75% of the original sample providing grade 12 data. Compared to the original sample of 585, the 438 families who provided data in grade 12 were of slightly higher socioeconomic status [M = 40.21 vs. 37.50, F(l,568) = 4.03, p < .05], but participants and non-participants did not differ by race, gender, or mothers' or teachers' reports of children's depressive symptoms in kindergarten.

Procedure and measures

Social preference, mutual friendships, and peer victimization

During the winter of each school year when children were in kindergarten and grades 1 through 3, sociometric interviews following the protocol described by Coie, Dodge, and Coppotelli (1982) were conducted in all classrooms in which at least 70% of children's parents gave consent. Interviews were conducted individually and orally. Children named up to three peers they especially liked and up to three peers they especially disliked. A social preference score was created by taking the standardized difference between the standardized like most nomination score and the standardized dislike most nomination score. Correlations between social preference scores in kindergarten through grade 3 ranged from .40 to .49; a reliable composite measure was created by averaging social preference scores across these four years (α = .77). Children were classified as being rejected by peers if their social preference score was less than −1, standardized like most score was less than 0, and standardized like least score was greater than 0. Based on these criteria, 66 children (11% of the sample), 45 children (9%), 52 children (10%), and 55 children (12%) were classified as being rejected in kindergarten through grade 3, respectively. A composite variable was created to reflect the number of years children were rejected in kindergarten through grade 3.

In grade 4 (for cohort 1 only) and grade 3 (for cohort 2 only), three peer victimization items were added to the sociometric interviews. Children were asked to nominate up to three peers who get picked on, who get teased, and who get hit and punched. Nominations were standardized within classrooms and averaged across the three items to create a peer victimization score.

Children in kindergarten and grades 1 through 3 also rated each of their participating classmates on a 1–5 scale (except for a few classrooms that used a 1–3 scale). A score of 1 indicated “Do not like” and a score of 5 indicated “Like very much” (a score of 3 was used in the classrooms using the 1–3 scale). The data for each classroom were converted into matrices and matched so that it was possible to identify what score each child gave and received. Matching ratings (e.g., child X rated child Y as a 5 and child Y reciprocally rated child X as 5) were identified for each child for “like very much” ratings. The total number of these reciprocated ratings was tallied for each child to give a number of mutual friendships score for each year.

Depressive symptoms and aggression

Reports of depressive symptoms and aggression were obtained from teachers, mothers, and children themselves. Classroom teachers completed the Teacher Report Form described above (Achenbach, 1991a) for each participating child approximately six months after children's initial entry into the study and annually through grade 8. Because most children had multiple teachers by the time they reached middle school, school personnel were asked to nominate the teacher most familiar with the adolescent to complete this measure.

Mothers completed the well-validated Child Behavior Checklist (Achenbach, 1991b), and adolescents completed the comparable Youth Self Report version (Achenbach, 1991c). The anxiety/depression subscale and the aggression subscale comparable to the teacher-report version of these subscales described above were used in the present investigation. Alphas across reporters and years ranged from .80 to .95. Established norms indicate whether scores are in borderline or clinical ranges (see Achenbach, 1991a, 1991b, 1991c). Correlations among teacher-, mother-, and youth-reported anxiety/depression (i.e., the dependent variables) ranged from .08 to .19 in grades 4 and 7; the correlation between mother- and youth-reported anxiety/depression was .41 in grade 12.

Results

As a first step in examining links between low social preference and subsequent depressive symptoms, we first calculated bivariate correlations between the continuous social preference scores children received from peers in kindergarten through grade 3 (averaged across years) and children's depressive symptoms in grade 4 (r = −.21, r = −.16 for teacher and mother reports, respectively, p < .01), grade 7 (r = −.35, p < .001; r = −.16, p < .01; and r = −.04, ns for teacher, mother, and self reports, respectively), and grade 12 (r = −.20, p < .001; r = −.08, ns for mother and self reports, respectively). The number of years the child was rejected was significantly (p < .05 or better) related to depressive symptoms in the borderline or clinical range in grade 4, χ2(2) = 7.68 (teacher report; mother reported depressive symptoms not significant); grade 7, χ2(2) = 19.06 and 8.84 (teacher report and mother report, respectively; self reported depressive symptoms not significant); and grade 12, χ2(2) = 8.15 (mother report; self reported depressive symptoms not significant). Across years and reporters, approximately 5% of the scores fell into the borderline or clinical range of depressive symptoms. However, for children who were rejected in two or more years during elementary school, 19% to 25% had subsequent teacher- and mother-reported depressive symptoms in the borderline or clinical range.

As in the Italian study, we next sought to investigate the role of prior depressive symptoms and prior and concurrent aggression in the link between social preference and subsequent depressive symptoms. Preliminary analyses examining whether gender moderated the association between social preference and subsequent depressive symptoms revealed no significant gender by social preference interactions; therefore the results reported below do not include gender in the models. In the first set of regressions, depressive symptoms and aggression from kindergarten were entered at step 1, and the social preference score was entered at step 2 in the prediction of depressive symptoms in each of three grades (grade 4, 7, or 12). As shown in Table 3, at step 2, social preference incrementally predicted depressive symptoms above and beyond prior depressive symptoms and prior aggression in six of the seven models (the single exception was self-reported depressive symptoms in grade 12).

Table 3. Regressions predicting depressive symptoms, controlling for prior depressive symptoms and prior aggression: American sample.

Predictors and test statistics Grade 4 Grade 7 Grade 12
Teacher Mother Teacher Mother Self Mother Self
Step 1
 Prior depressive symptoms .13** .36*** .20*** .24*** .10 .18** .09
 Prior aggression .17*** .12* .18*** .23*** −.04 .17** .04
 F 12.01*** 47.54*** 17.99*** 46.33*** 2.06 21.91*** 2.12
 R2 .05*** .19*** .08*** .17*** .01 .10*** .01
Step 2
 Prior depressive symptoms .12** .38*** .20*** .27*** .10* .21*** .09
 Prior aggression .08 .07 .02 .18** −.10 .12* .00
 Social preference −.20*** −.14** −.38*** −.14** −.13* −.17*** −.10
 F 13.72*** 35.28*** 33.48*** 35.11*** 3.30* 19.17*** 2.59*
 R2 .03*** .02** .12*** .02** .01* .03*** .01
Total R2 .08 .21 .20 .19 .02 .13 .02
Open in a new tab

Note. Numbers are betas and test statistics from the regressions.

*

p < .05;

**

p < .01;

***

p < .001.

We then re-ran the regressions substituting aggression measured concurrently with depressive symptoms (rather than kindergarten aggression; correlations between concurrent aggression and aggression in kindergarten ranged from .35 to .55 across reporters and years). As shown in Table 4, in these models, social preference did not incrementally predict subsequent depressive symptoms above and beyond prior depressive symptoms and concurrent aggression, with a single exception (teacher-reported depressive symptoms in grade 7). Thus, social preference contributed incrementally to subsequent depressive symptoms, even after taking into account prior aggression; however, the development of depressive symptoms was not independent of the development of concurrent aggression.

Table 4. Regressions predicting depressive symptoms, controlling for prior depressive symptoms and concurrent aggression: American sample.

Predictors and test statistics Grade 4 Grade 7 Grade 12
Teacher Mother Teacher Mother Self Mother Self
Step 1
 Prior depressive symptoms .15*** .29*** .20*** .23*** .08 .17*** .11**
 Concurrent aggression .40*** .52*** .29*** .58*** .48*** .61*** .55***
 F 50.39*** 151.22*** 31.89*** 188.11*** 63.65*** 155.71*** 95.97***
 R2 .18*** .43*** .13*** .46*** .24*** .43*** .32***
Step 2
 Prior depressive symptoms .15*** .28*** .19*** .23*** .08 .17*** .10*
 Concurrent aggression .36*** .52*** .19*** .58*** .48*** .61*** .55***
 Social preference −.08 .02 −.32*** −.01 .01 .01 −.03
 F 34.70*** 100.69*** 40.84*** 125.18*** 42.35*** 103.57*** 64.09***
 R2 .01 .00 .10*** .00 .00 .00 .00
Total R2 .19 .43 .23 .46 .24 .43 .32
Open in a new tab

Note. Numbers are betas and test statistics from the regressions.

*

p < .05;

**

p < .01;

***

p < .001.

The role of mutual friendships and peer victimization

Our next goal was to investigate whether mutual friendships or peer victimization moderated the link between social preference and depressive symptoms, controlling for prior depressive symptoms and concurrent aggression. We conducted regression analyses separately for each of these peer domains (i.e., mutual friendships and peer victimization). Each regression included main effects of prior depressive symptoms, concurrent aggression, social preference in a given year, and either the number of mutual friendships or degree of peer victimization assessed in the same year as the included social preference variable, as well as a social preference by mutual friendships or peer victimization interaction term.

Overall, these analyses provided little evidence that mutual friendships or peer victimization moderated the link between social preference and depressive symptoms. Two of the 28 mutual friendships × social preference interaction terms were significant: (a) Kindergarten social preference interacted with number of mutual friendships in kindergarten in the prediction of teacher-rated depressive symptoms in grade 7 (β = .16, p < .01, ΔR2 = .02; Slopes were −.513 (p < .05), −.5131 (ns), and −.5133 (ns) at low, mean, and high levels of friendships in kindergarten, respectively), and (b) Grade 1 social preference interacted with number of mutual friendships in grade 1 in the prediction of youth-reported depressive symptoms in grade 12 (β = −.10, p < .05, ΔR2 =.01; Slopes were .69 (p < .01), −.06 (ns), and −.80 (ns) at low, mean, and high levels of friendships in grade 1, respectively). This means that for children with average or high numbers of friendships in kindergarten, there was no link between kindergarten social preference and teacher-rated depressive symptoms in grade 7; however, for those children with few friendships in kindergarten, being higher in kindergarten social preference was associated with fewer teacher-reported depressive symptoms in grade 7. Similarly, for children with average or high numbers of friendships in grade 1, there was no link between social preference in grade 1 and self-reported depressive symptoms in grade 12; however, for children with few friendships in grade 1, being higher in grade 1 social preference was associated with more self-reported depressive symptoms in grade 12, a pattern opposite of the findings related to teacher-reported depressive symptoms in grade 7.

One of the seven peer victimization × social preference interaction terms was significant; Grade 3 social preference interacted with peer victimization in the prediction of grade 7 teacher-rated depressive symptoms (β = −.14, p < .05, ΔR2 = .02; only seven regressions were conducted with victimization variables because victimization was assessed in only one year, whereas mutual friendships were each assessed in four years). Slopes were −.27 (ns), −.74 (p < .01), and −1.20 (p < .001) at low, mean, and high levels of peer victimization, respectively. This means that for children who were not victimized by their peers, there was essentially no relation between social preference and depressive symptoms; for children who were victimized by their peers, depressive symptoms decreased as social preference increased.

The findings with the American sample suggest that children low in social preference in early elementary school are rated by their teachers and mothers as having more subsequent depressive symptoms, even after taking into account prior depressive symptoms and prior aggression. However, controlling for concurrent aggression attenuated the associations between social preference and subsequent depressive symptoms. The findings did not differ for boys and girls. We found only minimal evidence that mutual friendships or peer victimization affected the links between social preference and subsequent depressive symptoms.

Discussion

Taken together, the longitudinal findings from the Italian Genzano Study and American Child Development Project demonstrate that low social preference contributes to the prediction of later depressive symptoms, with some caveats. For the Italian sample, links between social preference and subsequent teacher-, mother-, and self-reported depressive symptoms could be accounted for by a combination of prior depressive symptoms and concurrent aggression; however, low social preference remained a significant predictor of peer-reported depressive symptoms, even after taking into account prior depressive symptoms, prior aggression, and concurrent aggression. The effects were almost completely accounted for by the co-occurrence of depressive symptoms with aggressive behavior in the American sample. That is, when prior depressive symptoms and co-occurring aggression were controlled, the effect of low social preference on subsequent depressive symptoms no longer held in the American sample. The findings were consistent for both boys and girls in both samples, and there was little evidence that mutual friendships or peer victimization moderated these links.

Although prior studies (e.g., Dodge et al., 2003) have provided strong evidence that peer rejection increments the prediction of the development of externalizing behavior problems, the evidence in the present study does not warrant the same conclusion regarding the role of peer rejection in the development of depressive symptoms. These findings are consistent with Parker and Asher's (1987) conclusion that the evidence for links between peer difficulties and maladjustment are stronger for criminality and dropping out of school (arguably, forms of externalizing behaviors) than for internalizing aspects of psychopathology. Prior studies that have demonstrated links between problematic peer relationships and depressive symptoms often have not taken into account depressive symptoms that preceded the problematic peer relationships (Patterson & Stoolmiller, 1991) or concurrent externalizing problems (Burks, Dodge, & Price, 1995). Studies that have included measures of externalizing behaviors often find that externalizing problems (e.g., aggression) are more importantly related to depressive symptoms than is peer rejection (French et al., 1995; Panak & Garber, 1992).

We found only limited evidence that mutual friendships and peer victimization affect the relation between social preference and children's subsequent internalizing problems. Two of the significant moderating findings made good conceptual sense, however. That is, for children who were at low risk (by virtue of having an average or high number of friends in kindergarten or not being victimized by peers), low social preference was not related to higher levels of subsequent depressive symptoms. However, for children who were at higher risk (by virtue of having few or no friends in kindergarten or being victimized by peers), low social preference was related to higher levels of subsequent depressive symptoms. Future research will help elucidate other contexts in which low social preference may place children at risk for later depression.

In the present study, social preference scores were obtained through nominations made by children's classroom peers and can be considered objective measures of peers' views rather than subjective measures of what target children believe their peers' views to be. This is important because children's perceptions of peer rejection have been found to be more strongly related to internalizing problems than has peers' actual rejection (Reinherz et al., 2000). One possibility is that children who are rejected by peers become aware of this rejection, which contributes to subsequent depressive symptoms. Another possibility is that children who are depressed are more likely to perceive their peers as rejecting them, regardless of whether the peers actually are rejecting. We were not able to evaluate these different possibilities in the present study because we did not have data on children's perceptions of peer rejection, but this remains an area of inquiry for future research.

A strength of this study was our ability to capitalize on reports from several independent sources. Teachers were able to rate children's depressive symptoms in the classroom, mothers were able to rate children's depressive symptoms at home, and peers were able to rate children's depressive symptoms in social interactions; each of these sources had access to different visible manifestations of children's internal psychological processes (see Bandura et al., 1999). Because only children themselves have access to the range of their moods across diverse settings and situations and because they may not show outward manifestations of their internal psychological states, children may be regarded in some ways as the best reporters of their own depressive symptoms (Coie, Lochman, Terry, & Hyman, 1992). Some researchers, however, have expressed concerns that children's self-reports of their moods may not be the best indicators (Cairns & Cairns, 1984). Thus, it is useful to examine the perspectives of multiple raters of children's depressive symptoms.

Although we were able to compare findings conceptually across the two samples, we were not able to compare findings statistically because the measures and times of measurement were not identical for the American and Italian samples. The main difference in the results in the American sample and the Italian sample was with regard to the role of concurrent aggression. Low social preference may lead American children to behave more aggressively, and they may appear more depressed because depression is co-morbid with aggressive behavior problems. In the Italian sample, concurrent aggression accounted for the links between low social preference and subsequent teacher-reported, but not mother-, peer-, or self-reported, depressive symptoms. It is important to note that in our study, the measures of aggression captured overt or direct aggression rather than relational or indirect aggression (see Grotpeter & Crick, 1996); future research would benefit from examining the role of relational aggression.

One important caveat to the interpretation of these differences between countries involves the ages at which the American and Italian children were assessed. In the American sample, social preference was assessed in elementary school, whereas in the Italian sample, social preference was assessed in adolescence; thus, development and culture are confounded. Developmentally, the timing of experiences with the peer group might be important. For example, Rodkin, Farmer, Pearl, and Van Acker (2000) found two subtypes of well-liked boys in adolescence: popular-prosocial (“model”) and popular-antisocial (“tough”). Both groups were nominated as being “cool” and athletic. In addition, the “tough” group was nominated as being aggressive, whereas the “model” group was nominated as being non-aggressive and academically competent. In a longitudinal study that directly compared the behavioral correlates of attraction to peers in different age groups, Bukowski, Sippola, and Newcomb (2000) found that attraction to aggressive peers increased as students made the transition from elementary school to middle school, particularly in terms of girls being attracted to more aggressive boys. Taken together, these findings suggest that youth who are admired by their peers are a heterogeneous group, and characteristics that are valued by elementary school peers may not be the same characteristics as those valued by middle school peers. Furthermore, children in middle school may be able to assess more accurately their standing within the peer group than can children in elementary school, which may make adolescents more susceptible to subsequent adjustment problems. These speculations await future empirical tests.

The risk that low social preference presents for later depressive symptoms holds for both boys and girls in the United States and Italy. However, in the American sample, the effects of low social preference on later depressive symptoms were accounted for by co-occurring aggression. In the Italian sample, the effects of low social preference on later depressive symptoms were accounted for by prior depressive symptoms and co-occurring aggression, except for peer-reported depressive symptoms. Thus, we conclude that low social preference does predict later depressive symptoms, and a possible mechanism through which this effect occurs is through the effect of poor peer relationships on increasing aggressive behavior, which is associated with depressive symptoms.

Acknowledgments

The Child Development Project has been funded by grants MH42498, MH56961, MH57024, and MH57095 from the National Institute of Mental Health and HD30572 from the National Institute of Child Health and Human Development. The Genzano Study has been funded by grants from the W.T. Grant Foundation, Spencer Foundation, and Johann Jacobs Foundation.

References

  1. Achenbach TM. Manual for the Teacher's Report Form and 1991 profile. Burlington, VT: University of Vermont, Department of Psychiatry; 1991a. [Google Scholar]
  2. Achenbach TM. Manual for the Child Behavior Checklist and 1991 profile. Burlington, VT: University of Vermont, Department of Psychiatry; 1991b. [Google Scholar]
  3. Achenbach TM. Manual for the Youth Self Report Form and 1991 profile. Burlington, VT: University of Vermont, Department of Psychiatry; 1991c. [Google Scholar]
  4. Angold A, Costello EJ. Depressive comorbidity in children and adolescents: Empirical, theoretical and methodological issues. American Journal of Psychiatry. 1993;150:1779–1791. doi: 10.1176/ajp.150.12.1779. [DOI] [PubMed] [Google Scholar]
  5. Asher SR, Parker JG, Walker DL. Distinguishing friendship from acceptance: Implications for intervention and assessment. In: Bukowski WM, Newcomb AF, Hartup WW, editors. The company they keep: Friendships in childhood and adolescence. New York: Cambridge University Press; 1996. pp. 366–405. [Google Scholar]
  6. Asher SR, Parkhurst JT, Hymel S, Williams GA. Peer rejection and loneliness in childhood. In: Asher SR, Coie JD, editors. Peer rejection in childhood. New York: Cambridge University Press; 1990. pp. 253–273. [Google Scholar]
  7. Attili G, Vermigli P, Schneider BH. Peer acceptance and friendship patterns among Italian schoolchildren within a cross-cultural perspective. International Journal of Behavioral Development. 1997;21:277–288. [Google Scholar]
  8. Bandura A, Pastorelli C, Barbaranelli C, Caprara GV. Self-efficacy pathways to childhood depression. Journal of Personality and Social Psychology. 1999;76:258–269. doi: 10.1037//0022-3514.76.2.258. [DOI] [PubMed] [Google Scholar]
  9. Bierman KL, Wargo JB. Predicting the longitudinal course associated with aggressive-rejected, aggressive (nonrejected), and rejected (nonaggressive) status. Development and Psychopathology. 1995;7:669–682. [Google Scholar]
  10. Blechman EA, McEnroe MJ, Carelia ET, Audette DP. Childhood competence and depression. Journal of Abnormal Psychology. 1986;95:223–227. doi: 10.1037//0021-843x.95.3.223. [DOI] [PubMed] [Google Scholar]
  11. Boivin M, Hymel S, Bukowski WM. The roles of social withdrawal, peer rejection, and victimization by peers in predicting loneliness and depressed mood in childhood. Development and Psychopathology. 1995;7:765–785. [Google Scholar]
  12. Boivin M, Poulin F, Vitaro F. Depressed mood and peer rejection in childhood. Development and Psychopathology. 1994;6:483–498. [Google Scholar]
  13. Bukowski WM, Hoza B. Popularity and friendship: Issues in theory, measurement, and outcome. In: Berndt TJ, Ladd GW, editors. Peer relationships in child development. New York: Wiley; 1989. pp. 15–45. [Google Scholar]
  14. Bukowski WM, Sippola LK, Newcomb AF. Variations in patterns of attraction to same- and other-sex peers during early adolescence. Developmental Psychology. 2000;36:147–154. [PubMed] [Google Scholar]
  15. Burks VS, Dodge KA, Price JM. Models of internalizing outcomes of early rejection. Development and Psychopathology. 1995;7:683–695. [Google Scholar]
  16. Cairns RB, Cairns BD. Predicting aggressive patterns in girls and boys: A developmental study. Aggressive Behavior. 1984;10:227–242. [Google Scholar]
  17. Camuffo M, Cerutti R, Lucarelli L, Mayer R. Il test sulla depressione dei bambini. Firenze: Organizzazioni Speciali; 1988. [Google Scholar]
  18. Caprara GV, Pastorelli C. Relazione tra coetanei. La nomina dei pari come predittore delle forme di adattamento e disadattamento nella prima infanzia (Peer relations. Peer nominations as predictors of adjustment and maladjustment during infancy) Eta' Evolutiva. 1993;44:86–93. [Google Scholar]
  19. Casiglia AC, Lo Coco A, Zappulla C. Aspects of social reputation and peer relationships in Italian children: A cross-cultural perspective. Developmental Psychology. 1998;34:723–730. doi: 10.1037//0012-1649.34.4.723. [DOI] [PubMed] [Google Scholar]
  20. Coie JD, Dodge KA, Coppotelli HA. Dimensions and types of social Status: A cross-age perspective. Developmental Psychology. 1982;18:557–569. [Google Scholar]
  21. Coie JD, Lochman JE, Terry R, Hyman C. Predicting early adolescent disorder from childhood aggression and peer rejection. Journal of Consulting and Clinical Psychology. 1992;60:783–792. doi: 10.1037//0022-006x.60.5.783. [DOI] [PubMed] [Google Scholar]
  22. Cole DA, Carpentieri S. Social status and the comorbidity of child depression and conduct disorder. Journal of Consulting and Clinical Psychology. 1990;58:748–757. doi: 10.1037//0022-006x.58.6.748. [DOI] [PubMed] [Google Scholar]
  23. Dodge KA, Bates JE, Pettit GS. Mechanisms in the cycle of violence. Science. 1990;250:1678–1683. doi: 10.1126/science.2270481. [DOI] [PubMed] [Google Scholar]
  24. Dodge KA, Lansford JE, Burks VS, Bates JE, Pettit GS, Fontaine R, Price JM. Peer rejection and social information-processing factors in the development of aggressive behavior problems in children. Child Development. 2003;74:374–393. doi: 10.1111/1467-8624.7402004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. French DC, Conrad J, Turner TM. Adjustment of antisocial and nonantisocial rejected adolescents. Development and Psychopathology. 1995;7:857–874. [Google Scholar]
  26. Grotpeter JK, Crick NR. Relational aggression, overt aggression, and friendship. Child Development. 1996;67:2328–2338. [PubMed] [Google Scholar]
  27. Hecht DB, Inderbitzen HM, Bukowski AL. The relationship between peer status and depressive symptoms in children and adolescents. Journal of Abnormal Child Psychology. 1998;26:153–160. doi: 10.1023/a:1022626023239. [DOI] [PubMed] [Google Scholar]
  28. Hollingshead AA. Four-factor index of social status. New Haven, CT: 1979. Unpublished manuscript. [Google Scholar]
  29. Kiesner J. Depressive symptoms in early adolescence: Their relations with classroom problem behavior and peer status. Journal of Research on Adolescence. 2002;12:463–478. [Google Scholar]
  30. Kiesner J, Poulin F, Nicotra E. Peer relations across contexts: Individual-network homophily and network inclusion in and after school. Child Development. 2003;74:1328–1343. doi: 10.1111/1467-8624.00610. [DOI] [PubMed] [Google Scholar]
  31. Kovacs M. The Children's Depression Inventory (CDI) Psychopharmacology Bulletin. 1985;21:995–1124. [PubMed] [Google Scholar]
  32. Ladd GW. Peer relationships and social competence during early and middle childhood. Annual Review of Psychology. 1999;50:333–359. doi: 10.1146/annurev.psych.50.1.333. [DOI] [PubMed] [Google Scholar]
  33. Ladd GW, Kochenderfer BJ, Coleman CC. Classroom peer acceptance, friendship, and victimization: Distinct relational systems that contribute uniquely to children's school adjustment? Child Development. 1997;68:1181–1197. doi: 10.1111/j.1467-8624.1997.tb01993.x. [DOI] [PubMed] [Google Scholar]
  34. Levendosky AA, Okun A, Parker JG. Depression and maltreatment as predictors of social competence and social problem-solving skills in school-age children. Child Abuse and Neglect. 1995;19:1183–1195. doi: 10.1016/0145-2134(95)00086-n. [DOI] [PubMed] [Google Scholar]
  35. Newcomb AF, Bukowski WM, Pattee L. Children's peer relations: A meta-analytic review of popular, rejected, neglected, controversial and average sociometric status. Psychological Bulletin. 1993;113:99–128. doi: 10.1037/0033-2909.113.1.99. [DOI] [PubMed] [Google Scholar]
  36. Panak WF, Garber J. Role of aggression, rejection, and attributions in the prediction of depression in children. Development and Psychopathology. 1992;4:145–165. [Google Scholar]
  37. Parker JG, Asher SR. Peer relations and later personal adjustment: Are low accepted children at risk? Psychological Bulletin. 1987;102:357–389. doi: 10.1037//0033-2909.102.3.357. [DOI] [PubMed] [Google Scholar]
  38. Parker JG, Asher SR. Friendship and friendship quality in middle childhood: Links with group acceptance and feelings of loneliness and social dissatisfaction. Developmental Psychology. 1993;29:611–621. [Google Scholar]
  39. Patterson GR, Stoolmiller M. Replications of a dual failure model for boys' depressed mood. Journal of Consulting and Clinical Psychology. 1991;59:491–498. doi: 10.1037//0022-006x.59.4.491. [DOI] [PubMed] [Google Scholar]
  40. Prinstein MJ, Boergers J, Spirito A, Little TD, Grapentine WL. Peer functioning, family dysfunction, and psychological symptoms in a risk factor model for adolescent inpatients' suicidal ideation severity. Journal of Clinical Child Psychology. 2000;29:392–405. doi: 10.1207/S15374424JCCP2903_10. [DOI] [PubMed] [Google Scholar]
  41. Reinherz HZ, Giaconia RM, Hauf AMC, Wasserman MS, Paradis AD. General and specific childhood risk factors for depression and drug disorders by early adulthood. Journal of the American Academy of Child and Adolescent Psychiatry. 2000;39:223–231. doi: 10.1097/00004583-200002000-00023. [DOI] [PubMed] [Google Scholar]
  42. Rodkin PC, Farmer TW, Pearl R, Van Acker R. Heterogeneity of popular boys: Antisocial and prosocial configurations. Developmental Psychology. 2000;36:14–24. doi: 10.1037//0012-1649.36.1.14. [DOI] [PubMed] [Google Scholar]
  43. Rubin KH, LeMare L, Lollis S. Social withdrawal in childhood: Developmental pathways to peer rejection. In: Asher SR, Coie JD, editors. Peer rejection in childhood. New York: Cambridge University Press; 1990. pp. 217–249. [Google Scholar]
  44. Rubin KH, Mills RS. The many faces of social isolation in childhood. Journal of Consulting and Clinical Psychology. 1988;56:916–924. doi: 10.1037//0022-006x.56.6.916. [DOI] [PubMed] [Google Scholar]
  45. Schwartz D, McFadyen-Ketchum SA, Dodge KA, Pettit GS, Bates JE. Peer group victimization as a predictor of children's behavior problems at home and in school. Development and Psychopathology. 1998;10:87–99. doi: 10.1017/s095457949800131x. [DOI] [PubMed] [Google Scholar]
  46. Vandell DL, Hembree SE. Peer social status and friendship: Independent contributors to children's social and academic adjustment. Merrill-Palmer Quarterly. 1994;40:461–477. [Google Scholar]
  47. Young HB, Ferguson LR. Puberty to manhood in Italy and America. New York: Academic Press; 1981. [Google Scholar]

RESOURCES