Perceptions of Social Behavior and Peer Acceptance in Kindergarten

Journal of Research in Childhood Education, Fall-Winter, 1999 
by Leslie C. Phillipsen, Sara K. Bridges, T. Gayle McLemore, Lisa A. Saponaro

 

Abstract. Forty kindergartners participated in a study using social behavior ratings from observers, teachers, and parents to predict the children's perceptions of peer acceptance. Gender and race differences in social behaviors and peer acceptance also were investigated. Perceptions of peer acceptance were gathered through individual child interviews, sociometric ratings, and teacher and parent ratings. In regression analyses, friendship skill predicted peer acceptance as reported by children and parents. Shyness / withdrawal inversely predicted teacher-reported peer acceptance. Surprisingly, aggression did not predict peer acceptance in any of the regression analyses. Analyses of group differences revealed that girls were rated as more prosocial than boys, and boys were rated as more aggressive than girls. No statistically significant race differences were found. The findings suggest that teachers and parents should encourage child-child interaction in order to promote friendship skills and inhibit shyness / withdrawal.

The peer relations literature is replete with studies showing that children who demonstrate certain kinds of social behaviors while refraining from other types of behaviors tend to be liked by their peers. For example, children who play cooperatively and show leadership abilities usually enjoy high peer acceptance (Hatzichristou & Hopf, 1996; Ladd, Price, & Hart, 1988). In contrast, children who display high levels of aggressive behavior or who interact with their peers in argumentative, disruptive, and socially inappropriate ways are often rejected by their peers (Coie & Dodge, 1988; Dodge, 1983; Dodge, Coie, Pettit, & Price, 1990; Hatzichristou & Hopf, 1996; Ladd et al., 1988; Lemerise, 1997). Shy and withdrawn behavior, such as not playing interactively with peers, watching peers play rather than joining in, and wandering around a classroom or playground, also tends to be associated with low peer acceptance (Lemerise, 1997).

This study was designed to isolate the types of social behaviors that predict kindergarten children's peer acceptance when considering several types of social behavior simultaneously. The outcome of this question is important to help parents, teachers, and others who work with young children understand what social skills to specifically foster and promote in order to enhance children's perceptions of their peer acceptance. Previous research has discovered developmental differences in the associations between social behaviors and peer acceptance. Aggression, for example, is linked with problematic peer relations from early childhood through adolescence, while socially withdrawn behavior begins to be associated with low peer acceptance in middle and late childhood (Rubin, Bukowski, & Parker, 1998; Snyder, Horsch, & Childs, 1997).

Researchers have frequently used sociometric techniques, such as peer nominations or ratings, to measure peer acceptance. While sociometry is a solid, objective measure of children's standing within the peer group, it provides only one perspective of children's peer acceptance. How children themselves feel about their peer acceptance is important, because these perceptions may actually guide their behaviors with peers.

Adult perceptions of children's confidence in their own peer acceptance also may influence their social behaviors. Adults who believe children are not confident about their peer acceptance might provide more opportunities to help these children develop play and friendship skills that could, in turn, lead to more confidence in their peer acceptance. For example, a teacher who believes a child lacks confidence in his or her peer acceptance might pair the child with another child who is confident about her peer acceptance, in order to provide a model of social behavior. In addition, a parent wanting to help a child gain confidence in peer acceptance might provide play dates with familiar peers or enroll the child in extracurricular activities to broaden the child's social network.

It is also important to examine the associations between children's social behaviors and perceptions of peer acceptance with the demographic variables of gender and race. If gender and racial difference are evident, teachers and parents might want to examine their teaching and child-rearing practices, respectively. Gender and racial variations might be intentional to promote children's growth across developmental areas (e.g., extra reinforcement for girls to play with blocks to encourage the development of spatial reasoning, while not providing such encouragement to boys who are likely to be self-motivated toward block play and/or receiving reinforcement from peers for doing so). However, some socialization practices may unintentionally promote limiting stereotypes (e.g., "Big boys don't cry," "Girls, please play nicely").

Research consistently has documented the different social interaction styles displayed by girls and boys. While boys frequently are found to engage in more overt aggression than girls (Coie & Dodge, 1998), the socialization of girls tends to emphasize prosocial behavior more than does the socialization of boys (Eisenberg & Fabes, 1998). Moreover, in children's close relationships, girls appear to be interested in intimacy and affiliation, while boys focus on status (Ladd, 1990; Maccoby, 1990).

                                                                                                     

There are fewer investigations of racial and ethnic preferences among children; furthermore, those that exist show less consistent findings than those examining gender differences. Children do appear to prefer to play with partners of the same race or ethnicity (Finklestein & Haskens, 1983; Fishbein & Imai, 1993; Lederberg, Chapin, Rosenblatt, & Vandell, 1986). Engaging in bullying behaviors appears to be ethnically "blind," however, with one study finding White and Asian bullies to be equally prevalent (Boulton, 1995). Another study found that children's self-esteem was positively affected by cross-race relationships (Taylor, 1991).

In summary, this study investigated the associations between aggression, shyness/withdrawal, prosocial behavior, friendship skill, and social behavior problems and peer acceptance in kindergarten. Children's own feelings of peer acceptance, sociometric ratings from peers, and teacher and parent perceptions of children's confidence in their peer acceptance were included in the regression analysis to isolate the social behaviors that predict kindergartners' peer acceptance across informants. The present study also investigated differences in social behaviors and peer acceptance among children of different genders and varied ethnic backgrounds in a diverse school and community.

Method

Participants

All of the 53 kindergarten children (16 girls, 37 boys) who were enrolled at a university affiliated laboratory early childhood school over a two-year period participated in the study. Before enrolling their children in the laboratory school, parents are asked to sign a blanket consent form for their children to participate in research projects. While notification of specific studies is still required, individual consent forms for each study are not. The procedures of this study (including that children would be observed and interviewed, and that teachers and parents would be asked to provide questionnaire data) were described in a letter sent to all parents. No parent, after receiving a description of the study, requested that his or her child be excluded.

The average age of the kindergartners in November (when the observations began) of each year of data collection was 5 years, 8 months (range = 5 years, 1 month to 6 years, 8 months). Sixty percent of the school's children were European American, 30% African American, and 10% other (primarily Asian American or mixed race). The majority of the children at the school were from middle-class households.

The children were enrolled in one of two classrooms. The first was a kindergarten classroom that used whole language as its overarching teaching philosophy. Forty children (12 girls) were enrolled in this classroom over the two years of the study. The second classroom used the Montessori educational philosophy and was mixed-aged (preschool through kindergarten). Thirteen kindergartners (4 girls) were enrolled in the Montessori classroom.

All analyses were conducted using the 40 children, for whom there were complete and reliable data. There were incomplete data for eight children, primarily due to missing parent questionnaires. As detailed in the Measures section, five children did not have reliable observation data. The children included in the analyses did not differ in gender, race, or classroom enrollment significantly from those not included in the final sample.

Procedures

Observations of children's social behaviors took place over several months. Children were observed by researchers as they engaged in work and play inside and outside their classrooms during the course of the regular school day. In the winter of each data collection year, teachers and parents completed questionnaires about children's social skills and feelings of competence. In the spring of each data collection year, the authors individually interviewed the children to obtain their sociometric ratings and self-perceptions of competence.

Measures

Observations. 

Each observer recorded anecdotal notes of a child's social behavior for several hours over several days before beginning to use the Preschool Behavior Q-set (Baumrind, 1967) to assess children's social behavior. Then, the observer rated each of the 72 Q-set items by placing them in a rectangular distribution: 8 items in each of 9 piles from least characteristic (1) to most characteristic (9). Before coding independently, each observer established adequate reliability (average kappa scores across Q-set items = .70) with the principal investigator. At least two observers independently rated each child's social behaviors. The observers' scores were averaged for each child. Kappa scores were used to assess the reliability of observers rating children from observations at different times. The average kappa score was .70 (range .43-.89). When adequate kappa scores (at least .40; Howes & Hamilton, 1992) could not be obtained from two observers' Q-sets on a child, a third observer rated the child. For five children, adequate kappa scores could not be obtained with the three observers, therefore, these children were not included in the analyses. Four social behavior style dimensions are derived from the Q-set (Howes & Phillipsen, 1992): prosocial, gregarious, aggressive, and withdrawn. This report included three of the dimensions: prosocial, aggressive, and withdrawn. Cronbach alphas were .82, .91, and .81, respectively. The gregarious dimension was not included, due to lack of a counterpart in the teacher measures.

Teacher questionnaire. 

Teachers completed the Cassidy and Asher (1992) teacher assessment of behavior for each child in their classroom. The 12-item measure asks teachers to rate children's behaviors on a five-point scale, with scores ranging from not true (1) to very true (5). The items are averaged into four behavioral dimensions: prosocial, aggressive, disruptive, and shy/withdrawn behavior. This study used three of these dimensions (Cronbach alphas are in parentheses): prosocial behavior (.89), aggression (.89), and shyness/withdrawal (.69). The disruptive dimension was not included, due to lack of a counterpart in the observer measures. Teachers also completed an item about children's confidence in their peer acceptance (How confident is this child about other children wanting to play with him/her?). This question assessed teachers' awareness of children's confidence in their relationships with peers. This item was rated on a four-point scale, from not very confident (1) to very confident (4).

Parent questionnaire. 

Parents responded to questions about children's friendship skills (Profilet & Ladd, 1994) and social behavior problems (Elliott, Gresham, Freeman, & McCloskey, 1988). The four friendship skill questions were rated on a five-point scale, from not at all easily (1) to extremely easily (5) (sample item: How easily does your child make friends with children he or she doesn't know?). These items were averaged to create a friendship skill score (Cronbach alpha = .79). There were 17 social behavior problems (e.g., fights with others, acts sad or depressed); parents rated the frequency of their child showing each behavior from never (0) to very often (2). The sum of the 17 behaviors served as the social behavior problem score (Cronbach alpha = .87). Parents also completed the same item that teachers did about children's confidence in their peer acceptance.

Child interview. 

Children were interviewed individually by a female researcher in a room nearby, yet separate from, their classrooms. During the interview, children completed sociometric ratings and perceptions of competence.For the sociometric ratings, children were first presented with a picture of themselves and asked who was pictured. All of the children were able to recognize themselves. Children were then asked how much they liked to play by themselves, and were presented with three options: a lot (3), corresponding to the largest of three plastic bowls; a little (2), corresponding to a medium-sized plastic bowl; and not at all (1), corresponding to the smallest of the three plastic bowls. Children rated themselves first to ensure that they understood the procedure. After rating themselves, children were presented, one by one, with pictures of their classmates and asked to name each child, and then to place the picture in one of three bowls according to how much they liked to play with each child. If they liked to play with a child a lot (3), they placed the picture in the largest bowl; if they liked to play with a child a little (2), they placed the picture in the medium-size bowl; if they liked to play with a child not a t all (1), they placed the picture in the smallest bowl. The sociometric ratings that each child received from all peers were averaged.

For the perceptions of competence, children answered 16 questions (a subsample of items from the Pictorial Scale of Perceived Competence and Acceptance of Young Children [Harter & Pike, 1980]). Pilot testing indicated that time constraints necessitated reducing the measure. Omitted items were those with which children might have had limited experience (e.g., eating dinner or sleeping over at a friend's house). Children completed one sample item before answering the 16 study items. Each item was presented in two steps. First, children were asked which of two pictures of children engaging in similar activities most closely represented them. For example, for a peer acceptance item, children were shown two pictures presented side by side. One picture showed a group of several children standing together on a playground, with the target child (identified by a pictured arrow, as well as being pointed to by the interviewer) in the middle of the group. The second picture showed a group of children standing together in the foreground of a playground, but the target child was standing away from the group by him/herself. Children were told, "This boy/girl usually gets asked to play with the other kids (interviewer points to target child in first picture), but this boy/girl gets lonely sometimes because the other kids don't ask him/her to play" (interviewer points to target child in second picture). After the child chose one of the two pictures as being like him/herself, he or she was asked whether the picture was "a lot like you" or "a little like you." This procedure gave children simple choices, while enabling their responses to be coded on a 4-point scale. Cronbach alpha for the peer acceptance sub scale (4 items) was .68.

Data Reduction

To reduce overlap among constructs, the observer and teacher variables were averaged to create composite variables of prosocial, aggressive, and shy/withdrawn behavior. That is, the observers' ratings of children's prosocial behavior and the teacher's prosocial ratings were averaged; the observers' ratings of children's aggressive behavior and the teacher's aggressive ratings were averaged; and the observers' ratings of children's withdrawn behavior and the teacher's shy/withdrawn ratings were averaged. The Cronbach alphas for these variables were as follows: prosocial behavior = .86, aggressive behavior = .90, and shy/withdrawn behavior = .68. It must be noted that the alpha for the shyness/withdrawal composite was quite modest, particularly in comparison to the alphas for the prosocial and aggressive behavior composites.

In sum, five social behavior variables were retained as independent variables in the analyses: prosocial behavior (composite of observer and teacher ratings), aggression (composite of observer and teacher ratings), shyness/withdrawal (composite of observer and teacher ratings), friendship skill (rated by parents), and social behavior problems (rated by parents). Four perceptions of peer acceptance were included as dependent variables in the analyses: the child's self-report of feelings of peer acceptance, the average sociometric rating received from all classmates, the teacher's report of the child's confidence in peer acceptance, and the parent's report of the child's confidence in peer acceptance.

 Results

Analysis Plan

Preliminary analyses indicated that there were no significant differences between children enrolled in the two classrooms on any of the five social behavior variables described above ("Data Reduction" section); thus, no classroom distinctions were made in the analyses. Three sets of analyses were conducted. The correlational and group difference analyses were conducted first, because the results of these analyses might have had an impact on the variables included in the regression analyses. Correlational analyses were completed on interrelations among the five social behavior ratings and the four views of peer acceptance. Second, group differences (gender and race) in the social behavior and peer acceptance variables were compared. Finally, a series of simultaneous multiple regression analyses were conducted to discover which social behavior variables best predicted each of the peer acceptance measures.

Correlations

Correlations were first computed among the five social behaviors: prosocial behavior, aggression, shyness/withdrawal, friendship skill, and social behavior problems. The social behavior ratings were modestly inter-correlated, with an average r of .32 (see Table 1). Prosocial behavior and aggression were strongly (inversely) related. Aggression and shyness/withdrawal were also inversely related. Parent-rated friendship skill and social behavior problems were inversely associated as well. Across informants, only prosocial behavior was significantly associated (inversely) with social behavior problems.

Next, correlations were computed among the four perceptions of peer acceptance: the child's report of feelings of peer acceptance, the average sociometric rating received from all classmates, the parent's report of the child's confidence in peer acceptance, and the teacher's report of the child's confidence in peer acceptance. The average correlation among the four perceptions of peer acceptance was .20. As shown in Table 2, children's perceptions of their peer acceptance were not linked with the sociometric ratings, teacher perceptions, or parent perceptions. The sociometric ratings, teacher perceptions, and parent perceptions were positively correlated, although the observed correlations were not strong (rs ranged from .31 to .33,p [less than] .05).

Group Differences

T-tests were used to investigate gender differences in the social behavior variables. As shown in Table 3, two significant differences emerged. Girls were rated as more prosocial than boys were (t(38)=2.44, p [less than] .05), and boys were rated as more aggressive than girls (t(38)=1.76, p [less than] .10). No gender differences were found in the parent ratings of friendship skill or social behavior problems.

One-way analyses of variance (ANOVAs) were completed to examine differences among the social behavior ratings for European American African American, and other races. No statistically significant differences were found.

Predictions of Peer Acceptance

Four simultaneous multiple regression analyses were conducted, each with one of the four measures of peer acceptance as the dependent variable. All analyses included the five social behavior ratings as predictor variables.

The first regression, predicting children's own feelings of peer acceptance from the social behavior ratings, was significant (F (5,34)=2.60, p [less than] .05). The overall variance accounted for by the predictor variables was 28%. The results (displayed in Table 4) show that friendship skill was a significant positive predictor of children's perceptions of peer acceptance. Parent reports of high friendship skill were associated with child reports of high peer acceptance.

The second regression predicted sociometric ratings from the social behavior ratings. It was not statistically significant. In the third regression, teacher ratings of children's confidence in their peer acceptance served as the dependent variable. The overall R was significant (F (5,34)=2.78, p [less than] .05). The overall variance accounted for by the predictor variables was 29%. As Table 4 shows, shy-withdrawn behavior was the only significant predictor (inversely related). That is, children who observers and teachers thought were shy and withdrawn also were rated by teachers as lacking confidence in their peer acceptance.

The fourth regression predicted parent ratings of children's confidence in their peer acceptance. The results of this analysis were statistically significant (F(5,34)=2.86, p [less than] .05). The social behavior ratings accounted for 30% of the variance in parent ratings of children's confidence in their peer acceptance. As shown in Table 4, friendship skill was a significant positive predictor of parent ratings of children's confidence in their peer acceptance. That is, when parents thought their children had good friendship skills, they also thought their children felt accepted by peers.

Discussion

The primary goal of this study was to isolate social behaviors that predicted perceptions of peer acceptance of kindergartners when considering several types of social behavior simultaneously. The multiple regression analyses revealed that both friendship skill and shy, withdrawn behavior predicted perceptions of kindergartners' peer acceptance. The ratings of friendship skill were parents' judgments of their children's abilities to form and maintain friendships with peers. The importance of friendship skills cannot be overestimated, as previous research has shown that having friends is extremely important for children's psychological wellbeing (Renshaw & Brown, 1993; Vandell & Hembree, 1994).

The findings related to shyness/withdrawal are particularly interesting. Previous research has suggested that aggression puts children at risk for peer relations difficulties across childhood, while shy/withdrawn behavior has been considered less problematic until children are older (i.e., in upper elementary school; see Rubin et al., 1998). The present findings suggest that shy and withdrawn behavior is problematic for kindergartners' peer relations, at least in this sample of children. However, the shyness/withdrawal composite variable showed fairly low internal consistency, so these findings must be interpreted with caution.

The practical implications of these findings are that parents and teachers should encourage and scaffold children's interaction with peers in the classroom and the playground. This one emphasis could address both of the social behaviors linked with perceptions of low peer acceptance: the lack of friendship skill and the presence of shy and withdrawn behavior. Fostering positive and appropriate child-child interactions could promote close relationships among children, while also discouraging the withdrawal of one or two children from the peer group.

To promote such interaction, teachers could organize the classroom materials, physical environment and daily schedule to encourage different kinds of groupings (e.g., pairs of children, small groups, and whole group) and interactions (e.g., play, peer tutoring, cooperative learning groups, discussions) throughout the day. Teachers also could implement a curriculum unit about friendship, using stories, games, artwork, and other activities to facilitate children's thinking and discussion about friendship skills (Ramsey, 1991). Parents could foster friendship skill and interactive behavior by encouraging their children to talk about the children they like and dislike (and why), organizing play dates between children of compatible behavioral styles, and modeling and talking about their own experiences with peers and friendship. Social skills training programs are also available for children with serious issues with aggression, withdrawal, or inappropriate behavior (Coie, 1985; Mize & Ladd, 1990).

Aggressive behavior, as rated by observers or teachers, did not predict peer acceptance in any of the regression analyses. These results are somewhat surprising, given the many studies that have found that aggressive behavior is negatively associated with peer acceptance (Dodge et al., 1990; Hatzichristou & Hopf, 1996; Ladd et al., 1988; Lemerise, 1997;Rubin et al., 1998). One interpretation of these findings is that the kindergarten classrooms in this study may have fostered a climate of acceptance of aggression. The fact that the majority (65 percent) of the classrooms' children were boys, who tend to be more aggressive than girls, lends credence to this possibility. Coie and Dodge (1998) describe contexts where aggression "works": "Environments that allow children to be exposed to aggression, to try out aggression, and to experience its positive instrumental consequences are likely to have children who develop aggressively" (p. 797). Classrooms with a large proportion of boys and a high degree of child in dependence (as both classrooms in this study could be described) might be environments that foster an aggressive climate.

An alternative explanation for the lack of prediction of aggression for peer acceptance is that measurement error related to the aggression measures contributed to the lack of results. This possibility seems slim, however. The aggression composite included observer and teacher ratings about children's propensity for mean and hurtful behavior, including bullying, starting fights, and general destructiveness. The associations among the measures and the group difference results revealed strong and consistent results for aggression (i.e., aggression was inversely related to prosocial behavior and positively related to social behavior problems, and aggression was rated more highly for boys than girls).

The analyses of group differences revealed results that are consistent with previous studies. Girls were rated as more prosocial than boys, and boys were rated as more aggressive than girls (Dodge et al., 1990; Hatzichristou & Hopf, 1996; Ladd et al., 1988; Lemerise, 1997). The lack of racial differences is also fairly convergent with a literature that has found few significant differences (Finklestein & Haskens, 1983). Since findings of racial differences often actually reflect socioeconomic disparities, the fact that the socioeconomic background of the sample was primarily middle class also makes these results not surprising.

The correlational analyses of children's social behaviors were in line with expectations. Prosocial behavior was negatively associated with aggression and social behavior problems, aggression, and shyness/withdrawal were negatively linked, and social behavior problems and friendship skill were negatively associated. It is interesting that parents' ratings of children's friendship skills were not associated with the observer and teacher ratings of children's social behaviors, yet they were such a consistent predictor of children's peer acceptance. It might be surmised that parent ratings were not linked with the observer and teacher ratings because the children's behavior was seen in different contexts. Observers and teachers most likely base their information on children's behavior at school, while parents presumably rate their child's behavior outside of school. Yet, parent reports of peer acceptance were associated with teacher reports and sociometric ratings. Further examination of parent ratings of friendship skills seems warranted in order to understand the behaviors observed by parents and how they differ from children's behavior at school.

Kindergartners' perceptions of their own peer acceptance were strikingly divergent from that of their peer group, parents, and teachers. Clearly, young children's feelings about their social competence do not always match the view of others. These findings contrast with research investigating associations between subjective and objective views of peer acceptance in older children. Children in the 3rd through 5th grades have been found to assess their peer acceptance more similarly to others' views than did the kindergartners in this study (Hymel, Bowker, & Woody, 1993; Patterson, Kupersmidt, & Griesler, 1990). These results also call peer relations researchers to bear in mind the perspective of their informant in making conclusions about their data. A young child who is viewed as socially competent by others may not feel socially competent, a dissonance that may affect his or her ber bavior. Also, the perspective of the child should be considered when designing both informal and formal interventions of peer acceptance.

In this study, two social behaviors--the lack of friendship skills and the presence of shy and withdrawn behavior--were most strongly linked with children's and adults' perceptions of low peer acceptance. These results indicate that parents and teachers should guide and encourage children's interactions with peers in the classroom and the playground. This emphasis could promote close relationships among children while also discouraging the withdrawal of one or two children from the peer group. The findings from this study also emphasize the variations among informants' views of children's social behaviors and peer acceptance.

References

Baumrind, D. (1967). Child care practices anteceding three patterns of preschool behavior. Genetic Psychology Monographs, 75, 43-88.

Boulton, M. J. (1995). Patterns of bully/victim problems in mixed race groups of children. Social Development, 4, 277-293.

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

Coie, J. D. (1985). Fitting the social skills intervention to the target group. In B. Schneider, K. Rubin, & S. Ledingham (Eds.), Children's peer relations: Issues in assessment and intervention (pp. 141-156). New York: Springer-Verlag.

Coie, J. D., & Dodge, K. A. (1988). Multiple sources of data on social behavior and social status in the school: A cross-age comparison. Child Development, 59, 815-829.

Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In W. Damon (Ed. in Chief) & N. Eisenberg (Volume Ed.), Handbook of child psychology: Vol. 3. Social, emotional and personality development (5th ed., pp. 779-862). New York: Wiley.

Dodge, K. A. (1983). Behavioral antecedents of peer social status. Child Development, 54, 1386-1399.

Dodge, K. A., Coie, J. D., Pettit, G. S., & Price, J. M. (1990). Peer status and aggression in boys' groups: Developmental and contextual analyses. Child Development, 61, 1289-1309.

Eisenberg, N., & Fabes, R. A. (1998). Prosocial development. In W. Damon (Ed. in Chief) & N. Eisenberg (Volume Ed.), Handbook of child psychology: Vol. 3. Social, emotional and personality development (5th ed., pp. 701-778). New York: Wiley.

Elliott, S. N., Gresham, F. M., Freeman, T., & McCloskey, G. (1988). Teacher and observer ratings of children's social skills: Validation of the social skills rating scales. Journal of Psychoeducational Assessment, 6, 152-161.

Finklestein, N., & Haskens, R. (1983). Kindergarten children prefer same-color peers. Child Development, 45, 502-508.

Fishbein, H. D., & Imai, S. (1993). Preschoolers select playmates on the basis of gender and race. Journal of Applied Developmental Psychology, 14, 303-316.

Harter, S., & Pike, R. G. (1980). The pictorial scale of perceived competence and acceptance for young children. Denver, CO: University of Denver.

Hatzichristou, C., & Hopf, D. (1996). A multiperspective comparison of peer sociometric status groups in childhood and adolescence. Child Development, 67, 1085-1102.


Source: http://findarticles.com/p/articles/mi_hb1439/is_1_14/ai_n28745191/?tag=content;col1