Children’s social competence within close friendship: The role of self-perception and attachment orientations

Children who experience difficulties in their close friendships are at risk for poor school adjustment and subsequent deviant behavior, whereas high-quality friendships buffer children from various harmful effects (Hodges, Boivin, Vitaro, & Bukowski, 1999). It is therefore important to learn more about the specific antecedents and qualities required to form and maintain successful friendships.

Friendships are considered vital for the acquisition of skills and competencies essential to the child’s social, cognitive, and emotional development (Newcomb & Bagwell, 1996). Friendships help children develop and maintain their perception of themselves as competent and worthwhile. They are also important contexts for intimacy and affection, safe-havens in times of difficulties, and practice arenas for later relationships (Parker, Rubin, Price, & DeRosier, 2006). Additionally, friends provide one another with assistance, mutual enjoyment, and joint activities (Parker, Rubin, Price, & DeRosier, 2006), and loneliness is perceived by children as a negative experience (Liepins & Cline, 2011).

Although researchers have acknowledged the importance of friendships for children’s well-being and adjustment (Bukowski, Newcomb, & Hartup, 1996), relatively little is known about the competencies children need to be successful in friendships. Most previous research focused on competencies associated with peer acceptance (Rubin, Bukowski, & Parker, 1998). However, the skills children need to be successful in their friendships are distinct from the skills they need to be well liked by peers (Rose & Asher, 1999).

Children’s ability to form and maintain friendships involves a range of social tasks such as initiation of interaction, being an enjoyable and resourceful companion, expressing care and concern, helping a friend with a problem, being a reliable partner, forgiving a friend despite feeling hurt, and managing conflict (Rose & Asher, 1999, 2000, 2004). Rose and Asher (1999, 2004) focused initially on the competency of dealing with the conflict of interest between a child’s own needs and those of the friend or the relationship, because it is central for the survival of the friendship. They found that children’s strategies in response to friendship conflict situations were related to the number of best friends children had, and were predictive of the quality of their best friendships. These findings held even when controlling for children’s general peer acceptance.

Children’s ability to turn to their friends in times of stress and need, and their willingness to provide support to their friends when the latter are in need of attention, are also important competencies. Rose and Asher (2004) found that children distancing themselves from their friends, whether with diplomacy or hostility, had fewer best friends than others. Notably, children’s strategies were related to the quality of the friendship only within the context of providing support to a friend, and not in the context of seeking support, and their endorsed strategies (especially the negative strategies) predicted changes in friendship adjustment (Glick & Rose, 2011).

Children’s ability and willingness to seek and provide social support in stressful circumstances and their conflict management competencies might vary and can be a derivative of their personal attributes and experiences. Children who experience empathic care learn to understand and recognize feelings in themselves and others, and to value closeness in the relationships as well as seek to preserve them. This ability serves as a foundation for interpreting social situations and responding prosocially to peers. Children who experienced more connected parent-child relationships showed higher levels of harmony and lower levels of conflict in friendships (Clark & Ladd, 2000).

Attachment, self-perception, and friendship

Based on experiences with their caregivers, children develop generalized expectations regarding the way in which parents will take care of them, and the extent to which the world may be seen as a secure and trustworthy environment (Bowlby, 1980). Children who experience sensitive and responsive care (secure attachment) will generalize their representational models and trust themselves and their social partners (Berlin, Cassidy, & Appleyard, 2008). By contrast, children who experience insensitive care will not develop trust in their environment or in themselves, and will perceive social partners as rejecting (avoidant attachment) or unpredictable (ambivalent attachment).

The behavioral style learned within attachment relationships may be carried over to other close relationships (Berlin et al., 2008). Indeed, security scores (with mothers) of 4th and 5th graders were positively associated with companionship and validation, and negatively associated with conflict and betrayal in their best friendships (Kerns, 1994). The interactions of pairs in which both children were securely attached were more responsive than those of secure-insecure friend dyads (Kerns, 1996). By contrast, children who experienced rejection from a caregiver were more likely to attach hostile attributes to peers’ ambiguous actions. Thus, cumulative findings support the notion that secure attachment fosters relational competence, especially in close friendships (Dwyer et al., 2010), and that insecure attachment is related to social difficulties. However, many studies were conducted with infants, toddlers, and preschoolers.

Another important personal attribute which might be associated with friendship competencies is self perception. From an early age children construct a personal theory about themselves based on their experiences and their interpretation of these experiences. Reactions and feedback from significant others influence and shape children’s emotions and feelings about themselves (Harter, 1999). Positive self-perception might facilitate the development of competence by motivating the child to attempt new and challenging social tasks. A child with a positive perception may have greater social confidence and may better handle the rebuffs that all children occasionally face (Goetz & Dweck, 1980).

In middle childhood, the peer group is especially important and children spend much time with their peers. The cognitive and emotional transformations change the behavioral expressions of attachment during this period. Children rely less on their attachment figures and turn more to others outside their family, though parents’ availability remains essential (Kerns, Tomich, & Kim, 2006). Friendships are based on loyalty, support, and common interests. During this period, self-perceptions become more differentiated (Cillessen & Bellmore, 1999), and children tend to use others as source of comparison when making self evaluations. Social competence is an aspect of self perception, and social-emotional competence and academic achievement are highly related (Del Prette, Del Prette, De Oliveira, Gresham, & Vance, 2012). Furthermore, at-risk children who experienced more frequent difficulties with friendships exhibited a subsequent increased likelihood of problem behaviors (Morrison, You, Sharkey, Felix, & Griffiths, 2012).

Thus, the first objective of this study was to explore how self-perception and attachment orientations relate to children’s competencies within friendships and to their social adjustment in class. It was expected that positive self-perception and secure attachment orientation would be associated with better social competencies. Further, in line with the depiction of competencies in friendships as varied and diverse (Rose & Asher, 2004) it was expected that children might reveal different levels of social skills when handling different social tasks. Additionally, their social competence in class, as perceived by their homeroom teachers, was examined.

Several scholars have argued that the genders differ in the quality and nature of their relationships with others (Maccoby, 1990). Rose and Asher (1999) found that girls were more likely to report willingness to accommodate and compromise, while boys were more likely to report that they would assert their own self-interest and behave in a hostile manner. They also found (2004) that girls were more likely than boys to report that they would offer support to a friend with a problem by sympathizing. Boys were more likely to report that they would avoid the friend with the problem, or blame him for having the problem (Rose & Asher, 2000). Nevertheless, Rose and Rudolph (2006) concluded that although girls generally perceive their friendships more positively than boys do pertaining to closeness, security, validation, and nurturance, boys are equally satisfied by their friendships. I expected to replicate Rose and Asher’s findings, indicating that girls would endorse more relation-oriented strategies than boys would in the various contexts.

Researchers have suggested that self-perception may serve as a mediator between attachment security and peer social competence (Booth-LaForce et al., 2006; Verschueren and Marcoen, 2005). Therefore, I also examined whether self-perception mediated the relationships between attachment and friendship competencies. Similarly, I examined whether gender, attachment, and self-perception moderated the relations between children’s characteristics and their social competencies.

In sum, the main objective of the study was to examine the associations between self-perception and attachment orientations and children’s social competencies: Strategies in managing conflicts, seeking and giving support within close friendships, and social competence in class. I also examined whether boys and girls differed in the strategies they endorsed within friendships, and in their social adjustment in class. Finally, I explored whether the relations between children’s characteristics and their social competence differ by gender, attachment orientations or self-perception, and whether self-perception mediates the link between attachment orientations and children’s competencies.

Method

Participants and procedure

Participants were 260 4th- and 5th-grade Jewish students (126 boys and 134 girls) from ten different classes in three schools in the northern part of Israel. The schools’ student populations were from lower middle class neighborhoods. After receiving permission from the Ministry of Education, consent letters were sent to parents. Eight children did not receive parental permission.

Questionnaires were administered in the school setting during two 70-minute sessions (two weeks apart). In the second session, due to rehearsals for the end-of year party and illness among the children, questionnaires were collect from only 228 children. In the first session the children completed the attachment questionnaire and the seeking and giving social support within a friendship questionnaire (260 children). In the second session they completed the questionnaire on self-perception and strategies in response to conflicts within a friendship (228 children, 108 boys and 120 girls). There were no differences between children who participated in both sessions and those who participated in the first session only. In addition, homeroom teachers reported on children’s social adjustment.

Measures

The Self Perception Profile for Children (Harter, 1985) is 36-item questionnaire tapping scholastic competence, athletic competence, peer likeability, physical appearance, behavioral conduct, and global self-worth; Cronbach’s alphas were 0.78, 0.73, 0.73, 0.78, 0.65, and 0.69 respectively. Children were asked to select the type of child they most resembled and to indicate whether that description was true for them. This measure has been widely used and shows good psychometric qualities (Cillessen & Bellmore, 1999). The global scale was used, combining the different domains (α = 0.90).

Attachment orientations regarding peers (Pinzi, Har-Even, Wizeman, Tiano, & Schnit, 1996) is an adaptation of the Hebrew version of Hazan and Shaver’s (1987) attachment styles questionnaire for middle childhood. Children were asked to report the extent to which 15 statements described them on a 1 to 5 Likert scale. It includes three scales: Secure (e.g. ‘I connect easily with other children”; Avoidant. (e.g. ‘Sometimes someone gets too close to me’); and Ambivalent/anxious (e.g. ‘Sometimes I feel that others don’t want to be as good friends with me as I would like’). Because Cronbach alphas of the Secure and the Avoidant scales were low, the two scales were combined (after reversing the avoidant scale) to reflect security (α = 0.65). The alpha of the ambivalent scale was 0.83. The test–retest and concurrent validity of the measure was demonstrated in a previous study (Pinzi et al., 1996).

Children’s strategies in response to conflicts within a friendship (Rose & Asher, 1999). The questionnaire contained hypothetical situations, and the study focused on the context of managing disagreements over resources, as this was expected to reveal more individual differences than other less challenging contexts. Following the presentation of conflict-of-interest situations, children were presented with six strategies options. For example, ‘You are shooting basketball by yourself on the playground during recess. Your friend comes over and says that he wants the ball. You tell him that you just got the ball from someone else and you want to play with it, but your friend still says that he wants the ball. Children rated how likely they would be to adopt each strategy on a 5-point Likert scale (see Supplemental Material; Table 1 for items, scales, composite scales and reliabilities). I attempted to reduce the number of variables without losing meaningful distinctions among them and used the composite scales constructed by Rose and Asher (1999). I kept the compromise and the accommodation strategies separate because the association between the variables in the current sample was low (r = 0.08).

Children’s strategies in response to seeking and giving support within a friendship (Rose & Asher, 2004). Children were presented with 12 hypothetical situations. Six represented contexts in which a child personally encounters challenging situations with peers, and the other six represented contexts where a child’s friend encounters challenging situations with peers. For example: ‘One day, right before recess, your class is supposed to be doing work at their desks, and some children in your class start teasing you by knocking your papers off your desk’. A context in which children could give social support is the following: ‘One day your best friend has to make a presentation in front of the class, and when he gets up in front of the class he seems to forget what he was going to say and he does very poorly at making the presentation. All during his presentation you see a couple of children whispering and laughing at him. You see him looking at the children who are whispering and laughing at him. When he goes back to his seat after the presentation, they keep laughing and talking and start pointing at him’. Children rated how likely they would be to adopt each strategy on a 5-point Likert scale (see Supplemental Material; Table 1 for a detailed description of scale items). Composite scales were constructed, where possible, to reduce the number of variables (see also Glick & Rose, 2011). Background data such as birth order, number of children in the family were also gathered.

Teacher–Child Rating Scale (T-CRS) (Hightower et al., 1986). This social skills scale was used to assess teachers’ perceptions of children’s social competence in school. Teachers were asked to state how well the items described the child on a 5-point Likert scale (5 items, ‘friendly towards peers’, alpha = −0.90).

Results

The association between self perception, attachment orientations, and children’s competencies

Pearson correlations were performed to examine the association between self-perception and attachment orientations, and strategies in close friendship controlling for gender. Secure attachment orientation and positive self-perception were negatively associated with hostile strategies in managing conflicts (see Supplemental Material, Table 2). Self-perception was positively associated with the compromise strategy. Finally, the accommodation strategy in managing conflict situations was positively associated with the ambivalent attachment orientation and negatively associated with teachers’ reports regarding social competence.

With regard to the seeking support context, positive self-perception, secure orientation, and teachers’ reports regarding social competence were negatively associated with excluding friend strategies, while ambivalent orientation was positively associated with excluding friend strategy.

With regard to giving support, ambivalent attachment orientation and self perception were positively associated with verbal support strategies. Avoidance was negatively associated with secure orientation, self-perception, and social competence, and positively associated with ambivalent attachment. In summary, attachment security and positive self-perception were associated with prosocial goals and strategies, and negatively associated with disengaging strategies. By contrast, ambivalent attachment was associated with accommodation strategy and disengaging strategies in the context of seeking and giving support.

Gender

Looking at gender in the context of managing conflict, girls endorse less hostility strategy and endorse more compromise strategy than boys. In the context of support-seeking, girls endorse more disclosure/advice seeking, are less likely to endorse excluding friend and distraction/denial strategies. Finally, in the context of support giving, girls show higher levels of verbal support and lower levels of avoidance, dismissing, and distraction/denial strategies.

Prediction of friendship competencies–mediation and moderation

Mediation

All the predicting variables were first centered. To explore the possible mediation of self-perception, hierarchical regression analyses were conducted. In the first step gender (boys = 0; girls = 1), ambivalent and secure attachment orientations were entered; and in the second step self-perception was entered. Self-image was not found to be a mediator in any of the cases.

Moderation

To explore the possible moderation of gender, attachment orientation, and self perception to children’s competencies, hierarchical regression analyses were conducted. In the first step gender, ambivalent and secure attachment orientations, and self-perception were entered. In the second step the various interactions among gender, attachment ambivalence, attachment security, and self-perception were entered. Consequently, moderation was found in predicting three competencies.

The strategy of excluding friend in the context of support seeking, gender, attachment security, and the interaction between gender and self-image predicts 14% of explained variance (See Supplemental Materials, Table 3). The relations between self-image and excluding friend are 0.01 among boys and −0.39 among girls.

In predicting the strategy of advice seeking/disclosure in the context of support seeking, self perception and the interaction between gender and ambivalent attachment, and security and self perception explains 13% variance. Among boys the correlation between ambivalent attachment and verbal support seeking is −0.13, whereas among girls it is 0.20. Among children with low levels of attachment security there is no association between self-image and verbal support seeking (r = 0.00), whereas among children with high levels of attachment security the correlation is 0.30. Thus, gender and attachment security moderated the link in verbal support seeking.

Finally, in predicting teachers’ reports regarding social competence, gender, attachment security, self perception and the interaction between gender and attachment security explains 12% of variance. Among boys the correlation between attachment security and social competence is 0.32 whereas among girls it is 0.09. Thus, among boys attachment security is more prominent in predicting teachers’ perception of social competence.

Limitations

The current study does not indicate the directionality of the link between self-perception/attachment orientations and social competencies. Positive self- and other perceptions may contribute to children’s success in their friendships, although competency in friendships may also influence children’s perceptions regarding the self and others. Children with positive self-perception and positive orientations regarding others approach others more confidently, interpret others’ actions more accurately and more positively and, as a consequence, are able to manage conflicts and seek and provide support more competently than children with less positive perceptions. Nevertheless, children’s success in managing conflicts and seeking and providing friends’ support might promote positive perceptions regarding the self and others. This also has implications for intervention programs, and an important goal of future studies remains to explore the causal mechanisms involved in this association between self-perception/attachment orientation and children’s competence (Cillessen & Bellmore, 1999).

This study was conducted in the Israeli cultural context, so the ability to generalize its findings to other contexts needs further to be explored. Definitions of competence are strongly shaped by culturally defined tasks, and models of success and preferred strategies for achieving social goals are also culturally defined (Obgu, 1981). For example, in the current sample accommodation is not associated with ‘positive qualities’, as it was in a US sample, which may reflect the importance of assertiveness in the Israeli culture (Roniger & Feige, 1993).

Most of the information gathered from participants in this study relies on self-report, and some additional information was gathered from homeroom teachers. Although most of these show good reliability, the secure attachment orientation shows only low internal reliability. Self-reports are open to various biases, for example, towards desirable self-representations. While children’s reports on their interpersonal relationships are a major and important source regarding their friendships, more can be learned when other perspectives are examined (Ladd, 2009). For example, the perceptions of children’s best friends could enrich our understanding. In addition, more can be learned by the use of other measures, such as observing children’s behaviors within friendships in experimental, as well as naturally occurring, situations.

This study employed a hypothetical situations methodology to allow controlled comparisons between children. Although a question arises regarding children’s actual behavior in real-life situations, their responses to hypothetical situations were found to relate to their actual behavior (Chung & Asher, 1996) and were confirmed in this research by homeroom teachers’ reports. Despite these caveats, the results attest to the important role of self-perception and attachment orientations in children’s competencies within friendships. This study focuses on some of the social tasks that are important in children’s friendships. Future research should focus on other friendship tasks to enrich our understanding of the various competencies needed within friendships and the inter-relations among them. This will help to identify children’s strengths and weaknesses, and to adapt intervention accordingly.

Discussion and implications

Self perception, attachment orientations and children’s competencies

As expected, both attachment security and positive self perception are related to pro-social and engaging strategies within friendship. These results suggest that children with positive perceptions of themselves and others in their social world are better equipped to cope with challenging situations within friendships; they refrain from hostility and do not avoid closeness. Children with positive self-perception are also willing to compromise in conflict situations and to give support to their friends when needed. As demonstrated in previous studies, abstaining from negative strategies is more important for the friendship than providing active support (Glick & Rose, 2011).

There were no associations between self-perception/attachment orientations and disclosure and advice seeking strategies, findings that are in line with Glick and Rose’s (2011) study. They found that the way in which children seek help has less impact on their friendships than the way in which they give help. This may be because children of this age hesitate to seek support, refrain from expressing vulnerability, and fear possible mockery from their peers (von Salisch, 2001). It may also reflect the cultural legitimacy attributed to working out problems or dealing with emotions on your own (Rose & Asher, 2004), or even the illegitimacy of seeking support.

The ambivalent orientation is associated with the accommodation strategy, and may reflect forgoing one’s own needs and wishes (at least in the Israeli culture) for the sake of preserving the friendship, frequently accompanied by feelings of dissatisfaction (Anders & Tucker, 2000) and anger at being exploited. Further, ambivalent attachment orientation is associated with excluding friends when actually needing help. Despite the need for help, individuals with preoccupied representations (ambivalent orientation) opted to seek support less than others (Scharf, Mayseless, & Kivenson-Baron, 2011). Though their need for help is intense, they are less inclined to request this, perhaps because of their mistrust of close others and their perception of support as insufficient to meet their needs (Kobak & Sceery, 1988). Note, however, that at the same time these children are inclined to give verbal support to their friends, and more inclined to distance themselves from friends (avoid or blame). These conflicting behaviors might confuse their partners and hamper the quality of their friendships.

Differences in social competencies between boys and girls

Boys showed higher levels of hostile strategies, and lower levels of compromise strategies, than girls in the context of managing conflicts. In the contexts of seeking and providing support, girls showed higher levels of engaging strategies than boys, whereas boys showed higher levels of disengaging strategies than girls (Glick & Rose, 2011; Rose & Asher, 2004).

Girls’ friendships are characterized by greater focus on dyadic interaction, while boys’ friendships tend to be found in the context of a larger peer group (Rose & Asher, 2000). By practicing their social competencies in these distinct areas, girls and boys acquire different types of interpersonal expertise (Maccoby, 1990). Thus, girls are more interpersonally oriented than boys, whereas boys are more instrumentally oriented than girls (Crick & Dodge, 1994). Nevertheless, although girls reported higher levels of emotional closeness and other positive aspects (Bukowski et al., 1996), girls and boys reported similar levels of satisfaction with friendships (Rose & Rudolph, 2006).

Mediation and moderation

Self-perception does not mediate the link between attachment orientations and children’s competencies. Each of the above dimensions contributes separately to children’s competencies. Though these dimensions are inter-related, it appears that both more intra-personal (self perception) and more interpersonal dimensions (attachment orientations) make a unique contribution to children’s social competencies.

With regard to moderation, the results of the current study demonstrate that engaging and disengaging strategies may function differently for boys and girls, and may have different implications on functioning and adaptation (Rose & Rudolph, 2006). In predicting positive strategies within friendships, the relations between attachment orientation/self-perception and pro-social competencies (disclosure/advice seeking and social competence according to teachers) are stronger among boys than among girls. The inverse (girls’ prominence) is demonstrated with regard to the excluding friend strategy. Thus, the role of each gender is more prominent in its less familiar and skilled competence. In these tasks the advantages of higher levels of secure attachment orientation and higher self esteem are especially marked in the gender that endorses the particular strategy less.

With regard to the strategy of advice seeking/disclosure, boys with ambivalent attachment orientation endorse this strategy less, whereas girls with ambivalent orientation endorse this strategy more. Among less securely attached children there are no relations between self image and this strategy, whereas among more securely attached children the relations are high. Disclosure and seeking advice is more characteristic of adolescent friendships (Rubin, Coplan, Chen, Bowker, & McDonald, 2011). In this respect, it seems that girls and more securely attached children favor more mature intimate behaviors, even at this relatively early age.

Practical implications

It appears that early screening and identification of children who exhibit difficulties in developing and maintaining friendships could prove an effective strategy in enhancing children’s adjustment. Thus, observing children’s functioning within friendships could help school psychologists, counselors, and teachers identify at-risk children. Conducting intervention programs that target social-emotional competence can bolster these skills before the problems exacerbate, decrease the possible long-term negative impact, and promote children’s well-being and adjustment (McCabe & Altamura, 2011).

Children often need to cope with rebuffs, rejection, and conflicts in friendships. Children characterized by insecure attachment orientation and poor self perception lack the social competence, confidence, and emotional capacity needed to negotiate these relationships in a fulfilling way. Changing children’s self perception and attachment is a complex, and probably long-term, therapeutic goal that is less suited to the broad-school intervention. However, practicing pro-social and engaging competencies might be a more applicable and useful vehicle in creating a classroom environment that facilitates children’s positive interactions with each other (Witvliet, van Lier, Cuijpers, & Koot, 2009). Involving parents and educators in the intervention may further help children alter their maladaptive patterns (McCabe & Altamura, 2011). It is also important to identify children who may need more intensive intervention. In this case it could be especially beneficial to also include parents in the intervention.

Helping children with ambivalent attachment orientation be aware of the negative implications of their conflicting and confusing behavior in seeking help from, and providing it to their partners might help them to use more adaptive behaviors. In any case, I recommend utilizing programs that have been found effective, such as Second Step (Holsen, Smith, & Frey, 2008; Larsen & Samdal, 2012), pair counseling (Karcher, 1997; Selman & Schultz, 1990), PATHs curriculum (Domitrovich, Cortes, & Greenberg, 2007), and similar evidence-based interventions.

Although it is important to help children work out problems on their own, the findings of this study regarding little use of support-seeking, calls for the combined efforts of educators and educational psychologists in challenging and changing children’s attitudes regarding support-seeking. Finally, as the findings of this study reveal, it is necessary to adapt interventions to children’s needs and deficits. Instead of practicing competencies that children have already acquired and use efficiently, it will be more useful and beneficial to practice more intensively different social competencies. For example, boys may benefit from learning how to compromise, whilst girls may benefit from practicing their distraction competencies. Thus, targeting specific competencies will result in more successful consequences.