Attachment representations and characteristics of friendship relations during adolescence

https://doi.org/10.1016/j.jecp.2004.02.002Get rights and content

Abstract

Attachment theory proposes that experiences with the primary caregivers are an important basis for the development of close social relationships outside the parent–child relationship. This study examined the association between representations of attachment, as assessed with the Adult Attachment Interview (AAI), representations of friendship and peer relations, as assessed with an interview in a sample of 43 adolescents. Secure attachment representations were significantly related to interview-based assessments of close friendships, friendship concept, integration in a peer group, and emotion regulation within close friendships. Attachment experiences reported during the AAI, their integration, and their coherency were related to friendship quality and friendship concept. Results show the close associations between attachment representations and friendship relationships during adolescence. The associations between peer relations and attachment representations differed depending on whether an interview approach or a questionnaire approach was used.

Introduction

One of the major developmental changes in friendship quality during adolescence is the growing importance of reciprocity, trust, and emotional closeness within these relationships (Selman, 1980; Youniss & Smollar, 1985). Compared with popularity, friendship typically is characterized by the importance of affection, intimacy, reliable alliance, and instrumental and emotional support (Berndt, 2002; Bukowski & Hoza, 1989). During development from preadolescence to adolescence, intimacy and self-disclosure become defining features of a close or best friendship (Berndt, 2002; Bukowski, 1987). Empathy and consensual validation of opinions about oneself and current life experiences are salient characteristics of friendships at this age that promote the development of self-worth and more realistic interpretations of experiences (Bigelow, 1977; Youniss & Smollar, 1985).

Despite this normative developmental trend, there is empirical evidence that not all adolescents have friendships that are characterized by mutuality and trust. Adolescents differ in their degrees of self-disclosure and can also have friendships that are characterized by disengagement (Berndt & Hanna, 1995; Shulman, 1995). In addition, Keller and Wood (1989) found that the quality of the friendship concept (i.e., the expectations regarding a best friend and trust) develops with age but shows tremendous differences within each age group as well as interindividual differences during developmental change. Possible reasons for interindividual differences in friendship quality at a given age can be diverse (e.g., recent move to a different city, neighborhood characteristics), so that adolescents at test time have not yet been able to develop friendships at a level that would be age appropriate. The parent–adolescent relationship, with both direct and indirect influences, is another factor (Ladd, 1992). Parents directly structure and select their children’s peer contacts, and parents indirectly influence norms and beliefs about appropriate social behavior and the relationship models based on attachment experiences. Buhrmester and Furman (1990) suggested that friendships, with growing age, develop characteristics that formerly have been typical for the parent–child relationship. Differences in the quality of close friendships during adolescence, then, could be explained at least partly by differences of experiences in the parent–child relationships.

Attachment theory assumes that a child’s emotional bond with his or her parents is strongly predictive of other close relationships later in life (Bowlby, 1973; Waters & Cummings, 2000) if these relationships have similarity to aspects of an attachment relationship (Ainsworth, 1990). The salient characteristic of the attachment system is its security-regulating function. A child seeks emotional support from specific caregivers when he or she feels distressed and does not feel capable of autonomous emotion regulation. Attachment theory proposes that, based on experiences of effective or inefficient emotion regulation by the caregivers, a child develops internal working models of the self and of the caregivers (Bowlby, 1973; Bretherton & Munholland, 1999). The transfer of attachment to parents to other close relationships is based on this concept of internal working models (Bowlby, 1973; Sroufe & Fleeson, 1986). Internal working models influence information processing (e.g., perception, interpretation, expectations) as well as emotion and behavior regulation (e.g., proximity seeking, coping) (Zimmermann, 1999). There is empirical evidence that infant–child attachment patterns are associated with expectations toward friends (i.e., friendship concept) and interpretations of friends’ behavior and interaction patterns.

Insecure attachment organization has been associated with a more hostile interpretation of hypothetical conflict situations between peers during preschool (Suess, Grossmann, & Sroufe, 1992) and with a more rigid and negative interpretation of hypothetical social rejection situations during adolescence (Zimmermann, 1999). Elicker, Englund, and Sroufe (1992) found significant associations between infant–mother attachment and preadolescents’ friendship concepts (i.e., expectations of specific behaviors and rules within friendships).

Studies observing peer interactions showed that secure infant–mother attachment quality was significantly associated with social competence, low aggressiveness, and more cooperative friendships (Shulman, Elicker, & Sroufe, 1994; Sroufe, 1989; Suess et al., 1992). In a more recent meta-analysis, the average effect size of the influence of attachment on friendship from preschool to preadolescence was r=.24, whereas the mean effect size for studies assessing wider peer relations was lower at r=.14 (Schneider, Atkinson, & Tardif, 2001). Thus, there is empirical evidence of a significant association between attachment and especially close friendship quality during childhood.

Studies on attachment measures and friendship relations during preadolescence and adolescence have found that self-reported or parent-reported attachment security is related to friendship quality that is characterized by trust, closeness, and mutuality (Kerns, Klepac, & Cole, 1996; Lieberman, Doyle, & Markeiwicz, 1999; Shulman, 1995). The growth in the importance of trust and reciprocal self-disclosure during adolescence makes friendship more similar to relationships with attachment characteristics. Differences in attachment, therefore, might be more likely to be paralleled by differences in friendship quality at that age. There is substantial evidence that attachment is related to social competence during adolescence as well (Allen & Land, 1999). However, studies focusing on the relation between attachment and close friendships during adolescence are rare.

Self-reported attachment does not take into account the issue of idealization in the description of attachment relationships. This is considered by studies using the Adult Attachment Interview (AAI) (George, Kaplan, & Main, 1985), which assesses the state of mind regarding attachment (Main, 1991). Using the AAI, participants can be classified into one of four patterns that parallel the patterns of attachment quality during infancy (but are not identical to them). Individuals classified as “secure” or “free to evaluate” coherently report about their past attachment experiences independent of their quality. They show signs of integration of these experiences in their representations of their selves in relationships and value attachment relationships. Individuals with “dismissing” attachment representations show an incoherent discourse by idealizing their caregivers, insisting on being unable to recall attachment-related episodes, and/or dismissing the effect of nonsupportive experiences. Individuals classified as enmeshed or “preoccupied” with attachment show anger when talking about their attachment relationships. They have difficulties in separating past and current relationships with parents, often oscillate between positive and negative appraisals of attachment experiences, and/or show a passive discourse style. In addition to these three major patterns, participants can receive a classification as unresolved/disorganized based on lapses in monitoring of reasoning or discourse when reporting about loss or abuse (Hesse, 1999). Besides the classification, mega-items, assessing the formal discourse quality, validly reported attachment experiences, and processing of attachment experiences (e.g., valuing of attachment) allow additional analyses (Kobak & Sceery, 1988; Zimmermann, 1999).

Kobak and Sceery (1988) reported that late adolescents with insecure attachment representations are more hostile (especially if classified as dismissing) and more anxious (especially if classified as preoccupied) than are adolescents with secure attachment representations, as characterized by their friends. Zimmermann, Gliwitzky, and Becker-Stoll (1996) replicated this finding for 16-year-olds. In addition, they found a higher concordance of self and friend descriptions by means of the California Adult Q-sort (Block, 1978) for adolescents with secure attachment representations. These adolescents present themselves to their friends as they experience themselves, so friends know each other quite well (Van Lieshout, van Aken, & van Seyen, 1990). Cooperation of friends when experiencing negative feelings also depends on adolescents’ attachment representations (Zimmermann, Maier, Winter, & Grossmann, 2001).

Because individuals with insecure attachment representations, especially with dismissing attachment organization, tend to idealize themselves and close relationships, it should be controlled in the assessment. This can be accomplished with an interview approach such as the AAI. Furman, Simon, Shaffer, and Bouchey (2001) used the AAI to assess working models of parents, friends, and romantic partners. The quality of attachment representation was significantly associated with the respective working model of friendship but not in all respects to working models of romantic relationships. Applying the basic principle of the AAI to the assessment of peer and romantic relationships helps to control for idealization. However, it leaves unanswered the question of whether the associations found may be explained methodologically as an overall concordance of coherence instead of the relationship characteristics. The AAI is not classified on the basis of reported experiences; rather, it is classified on the basis of the coherence or style of incoherence of individuals’ answers during the interview, even if the individuals experienced intense rejection (Hesse, 1999). If the same principle is applied to the assessment of the representation of other close relationships (e.g., friendships, romantic relationships), the associations between attachment security and security of other close relationships might reflect individuals’ ability to talk coherently about close relationships in general, even if the experienced relationship quality indeed was not supportive. However, attachment theory would suggest that there is a transfer of supportive experiences with parents to mutually supportive close relationships with peers or partners (Bowlby, 1973; Sroufe & Fleeson, 1986). Thus, the concept of coherence is important to control for idealization, but the theoretical assumptions that must be tested are whether relationship patterns are transferred from parent–child relationships to friendships.

The current study was designed to assess the associations between adolescents’ attachment representations and concurrent representations of adolescents’ friendship and peer relations and emotion regulation patterns. It was expected that adolescents with secure attachment representations would report closer friendships, a higher level of friendship concept, more integration in wider peer relations, and the ability to engage in cooperative conflict resolution with low rates of hostility and social anxiety. In addition, the assessment of peer relations by questionnaire might not reveal such associations due to idealization of the dismissing group. Finally, the study attempted to test whether concordance between the AAI and friendship is based solely on the associations with the coherence of the AAI or is also found with the reported attachment experiences, the reported attachment behavior, and the processing of attachment experiences.

Section snippets

Sample and procedure

In a follow-up of the Bielefeld Longitudinal Study (Grossmann, Grossmann, Huber, & Wartner, 1981), an ongoing study on attachment and socioemotional development, 44 adolescents (of the original 52 children) were visited at home. Because the assessment could not be fully completed for 1 participant, data analysis was reduced to 43 participants (21 girls and 22 boys). The assessment took place within 3 months of the participants’ 16th birthdays.

Two different researchers interviewed the

Preliminary analyses

Preliminary analyses investigated possible effects of gender on the attachment and peer relations scales and the interrelation between the peer interview and peer questionnaire measures. A series of t tests revealed that girls were rated as significantly more socially anxious, t(42)=2.1, p<.05, in the Friendship and Peer Relations Interview and reported better peer relations in the Offer Self-Image Questionnaire, t(42)=2.9, p<.01, as compared with boys. None of the other scales of the peer

Discussion

Although most adolescents develop friendships and friendship concepts that include reciprocal trust and support, not all of them do so. One major aim of this study was to examine whether differences in friendship and peer relations were associated with attachment representations. Adolescents with secure attachment representations, in contrast to adolescents with dismissing attachment representations, reported emotionally close friendships and revealed an elaborated friendship concept. In

Acknowledgements

This study was supported by the Deutsche Forschungsgemeinschaft. The author thanks Elisabeth Fremmer-Bombik, Gottfried Spangler, and Peter Merkl for their assistance in interview ratings, Gottfried Spangler for a critical review of an earlier version of this article, and two anonymous reviewers for their thoughtful and helpful comments. The author also expresses his gratitude to the families for their cooperative participation.

References (50)

  • P Zimmermann et al.

    Stability of attachment representations in adolescence: The influence of ego-identity status

    Journal of Adolescence

    (2002)
  • M.D.S Ainsworth

    Some considerations regarding theory and assessment relevant to attachments beyond infancy

  • J.P Allen et al.

    Attachment in adolescence

  • J.P Allen et al.

    Attachment and adolescent psychosocial functioning

    Child Development

    (1998)
  • T.J Berndt

    Friendship quality and social development

    Current Directions in Psychological Science

    (2002)
  • T.J Berndt et al.

    Intimacy and self-disclosure in friendships

  • B.J Bigelow

    Children’s friendship expectations: A cognitive developmental study

    Child Development

    (1977)
  • J Block

    The Q-sort method in personality assessment and psychiatric research

    (1978)
  • J Bowlby

    Attachment and loss, Vol. 2: Separation: Anxiety and anger

    (1973)
  • I Bretherton et al.

    Internal working models in attachment relationships: A construct revisited

  • D Buhrmester

    Intimacy of friendship, interpersonal competence, and adjustment during preadolescence and adolescence

    Child Development

    (1990)
  • D Buhrmester et al.

    The changing functions of friends in childhood: A neo-Sullivan perspective

  • W Bukowski

    Friendship conceptions among early adolescents: A longitudinal study of stability and change

    Journal of Early Adolescence

    (1987)
  • W Bukowski et al.

    Popularity and friendship: Issues in theory, measurement, and outcome

  • J.A Crowell et al.

    Measurement of individual differences in adolescent and adult attachment

  • J Elicker et al.

    Predicting peer competence and peer relationships in childhood from early parent–child relationships

  • W Furman et al.

    Adolescents’ working models and styles for relationships with parents, friends, and romantic partners

    Child Development

    (2001)
  • George, C., Kaplan, N., & Main, M. (1985). An Adult Attachment Interview. Unpublished manuscript, University of...
  • K.E Grossmann et al.

    German children’s behavior towards their mothers at 12 months and their fathers at 18 months in Ainsworth’s Strange Situation

    International Journal of Behavioral Development

    (1981)
  • K.E Grossmann et al.

    A wider view of attachment and exploration: Stability and change during the years of immaturity

  • E Hesse

    The Adult Attachment Interview. Historical and current perspectives

  • M Keller et al.

    Development of friendship reasoning: A study of inter-individual differences in intra-individual change

    Developmental Psychology

    (1989)
  • K Kerns et al.

    Peer relationships and preadolescents’ perceptions of security in the mother–child relationship

    Developmental Psychology

    (1996)
  • Kobak, R.R. (1993). The Adult Attachment Interview Q-sort. Unpublished manuscript, University of...
  • R.R Kobak et al.

    Attachment and emotion regulation during mother–teen problem solving: A control theory analysis

    Child Development

    (1993)
  • Cited by (124)

    View all citing articles on Scopus
    View full text