403482 ot and MayselessJournal of Early Adolescence © The Author(s) 2012

JEA32410.1177/0272431611403482Gran

Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

Representations of Mother-Child Attachment Relationships and Social-Information Processing of Peer Relationships in Early Adolescence

Journal of Early Adolescence 32(4) 537­–564 © The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0272431611403482 http://jea.sagepub.com

David Granot1 and Ofra Mayseless2

Abstract We examined the concurrent associations between early adolescents’ representations of mother-child attachment relationships and how they process Attachment representations social information in their peer relationships.   were examined in a normative sample of 97 males and 88 females (mean age = 10.35 years), using an adaptation of the Attachment Doll Story Completion Task. Structural Equation Modeling was used to assess general latent structures of social-information processing (SIP) orientations across the different SIP steps. As expected, secure mother-child attachment representations were positively associated with prosocial SIP orientation and negatively with antisocial SIP orientation. Avoidant attachment was associated negatively with prosocial and distress expression SIP orientations. Ambivalent attachment was positively associated with distress expression SIP orientation. Disorganized attachment was positively associated with SIP distress expression orientation 1

Oranim, Academic College of Education, Tivon, Isreal University of Haifa, Haifa, Israel

2

Corresponding Author: Ofra Mayseless, Faculty of Education, University of Haifa, Haifa, 31905 Israel Email: [email protected]

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

538

Journal of Early Adolescence 32(4)

and with antisocial SIP orientation. Results are discussed as reflecting a generalization of social knowledge and regulation strategies from the attachment system to the affiliative system. Keywords Attachment, adolescence, peer relationship, social-information processing, Doll Story Completion Task

A key proposition of attachment theory is that the quality of early attachment relationships affects and guides the child’s quality of relations with peers later in life (Cassidy, 2008). Specifically, attachment representations formed following repeated interactions with attachment figures are assumed to subsequently guide the child’s perceptions and cognitions (social-information processing) with peers (Kerns, 1996; Parke & O’Neil, 1999). In the present study, we investigated this assumption by examining the associations between early adolescents’ representations of mother-child attachment relationships and how they process social information in their peer relationships. According to attachment theory, all infants have an innate, biologically based tendency to form an attachment bond to stronger and wiser individuals, who become their attachment figures (Bowlby, 1969/1982). Through repeated daily experiences with these attachment figures, children develop cognitiveaffective attachment representations of themselves as being (un)worthy of care and protection and of the tendencies of these figures to be (un)available, (un)responsive, and (in)sensitive to their need for protection (Cassidy, 2008). These representations are revealed in cognitive processes such as selective perception, memory retrieval and assimilation of divergent pieces of information into a coherent whole, as well as in emotion-regulation strategies (Bretherton & Munholland, 2008). These representations guide the processing of information on attachment relationships with these figures, but are also expected to guide social relationships in general, including relationships with other social partners such as peers (Kerns, 1996; Parke & O’Neil, 1999). For example, a child who has a sensitive attachment figure develops positive expectations and attitudes about relationships, as well as a flexible, open, and resilient emotion-regulation strategy. These expectations and attitudes, as well as the relevant emotion-regulation strategies, are expected to generalize to other contexts and to be carried forward into other relationships. Specifically, attachment researchers have suggested such a transfer from past experiences in attachment relationships—a close and hierarchical relationship with an

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

539

adult caregiver—to peer relationships—a symmetrical affiliation with same-age partners (Bretherton & Munholland, 2008). Such a postulated transfer between two different social systems—attachment and affiliation, each reflecting distinct evolutionary predispositions (Furman & Simon, 1999)—is not necessarily self-evident and needs empirical confirmation.

Social-Information Processing in Relations With Peers A central model for social-information processing, which can help elucidate individual differences in information processing in relations with peers, is the multistep model for Social-information processing (SIP) suggested by Dodge and colleagues (Crick & Dodge, 1994; Dodge & Price, 1994). These researchers assumed that variations in children’s social-information processing are implicated in children’s social problems (aggressiveness, passivity, withdrawal) and prosocial behavior. They proposed several steps that individuals undergo when they encounter a social situation: (a) encoding is perceiving and encoding situational and social cues; (b) interpretation is forming a mental representation and interpretation of the situation; (c) clarification of goals is selecting a goal or desired outcome for interaction; (d) response generation is recalling or constructing possible reactions to the situation; (e) response evaluation is evaluating these reactions; and (f) Enactment is initiating what is expected to be an adequate action while monitoring its effects and regulating behavior accordingly. According to the model, variations in children’s socioemotional functioning derive from differences in such social-information processes, leading to a diverse selection of behavioral reactions. The reason for the variation lies in the personal social “data base,” a collection of memories and insights that is based on experience and crystallizes into internalized representations and social schemes. These different internalized representations and social schemes guide children to different interpretations, hence to different social reactions. Crick and Dodge (1994) assumed that a child’s attachment representations are one of the central components of the social data base, which guide these processes of interpretations of social information in the peer arena. A large number of studies present strong empirical support for the association between specific SIP steps and children’s social adjustment (for review see Crick & Dodge, 1994). Dodge (1993) further postulated the existence of overall latent structures of social-information processing, reflecting major SIP orientations that are observed across the different steps. These SIP orientations reflect an overall tendency in social-information processing that can be identified across the different SIP steps, including the final enactment step (Dodge, 1993).

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

540

Journal of Early Adolescence 32(4)

Crick and Dodge (1994) postulated the existence of three main alternative response orientations in problematic social situations to “move toward” the other (prosocial orientation), to “move against” the other (antisocial-aggressive orientation), and to “move away” from the other (withdrawal/avoiding orientation). Von Salisch’s (2001) findings suggested an additional alternative orientation as a common response of children in problematic social situations to express their distress as a way to achieve their goals (distress expression orientation). In the present study, we assessed all four orientations as overall latent SIP orientations, and expected the different patterns of attachment representations to be distinctly associated with the different SIP orientations.

Individual Differences in Attachment Representations and Social-Information Processing Research in the attachment paradigm has identified four major patterns of attachment behavior and representations, one secure and three insecure: avoidant, ambivalent and disorganized (Cassidy, 2008). Secure children are characterized by a balance between self-reliance and reliance on the attachment figure in times of distress, and a flexible and prosocial emotion regulation strategy. Consequently, with regard to the processing of social information, secure children are likely to develop open, flexible, and nondefensive information-processing orientations, with a high tendency to be prosocial and to cooperate with others, and low tendency to be antisocial and aggressive toward others. The three insecure patterns are conspicuous in their imbalance, showing an emphasis on self-reliance at the expense of close relationships (avoidant), an emphasis on dependency at the expense of self-reliance and exploration (ambivalent), or no consistent pattern (disorganized). Avoidant children, who are assumed to be raised by a rejecting parent, are expected to develop a deactivating emotion-regulation strategy that minimizes the expression of negative affect (sadness, distress) as a defensive exclusion of threatening information (Cassidy, 2008; Mikulincer & Shaver, 2008). Based on these internalizations, avoidant children are likely to develop a distancing orientation to information processing in social situations, with low tendencies to openly express their need for help by the expression of distress or to promote prosocial interactions with others, especially when feeling rejected. Ambivalent children, raised by a parent who is usually inconsistent and unreliably

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

541

responsive, are assumed to develop a “heightening emotion-regulation strategy” in which they heighten expressions of anger, distress, or dependency to coerce the caregiver to respond to their needs (Cassidy, 2008; Mikulincer & Shaver, 2008). Based on these representations, ambivalent children are likely to develop a hypervigilant and anxious orientation, with a tendency to use expressions of distress in response to challenging peer interactions. Lyons-Ruth, Bronfman, and Atwood (1999) suggested that the disorganized pattern is rooted in an improper parent-child hierarchy in which the parent’s behaviors (frightening or frightened) prevent reciprocity in the relationships, leaving the child with two conceivable alternatives: to counteract (be hostile) or to give up (be helpless). Based on this experience, disorganized children are likely to develop two parallel social-information processing orientations: antisocial, which involves hostility and aggressiveness, and helplessness, which may be enacted by expressing distress. Several previous studies have examined the associations between parentchild attachment and social-information processing of relationships with the peer group. Suess, Grossmann, and Sroufe (1992) showed that German children who had been secure as infants (observed by the Strange Situation at 12 months) showed a lower tendency to hostile attribution of peer behavior in problematic social situations than children who had been avoidant as infants. Similarly, assessing 6- to 7-year-olds, Cassidy, Kirsh, Scolton, and Parke (1996) found that secure children assessed by a reunion procedure had a higher tendency to attribute positive intentions to peers. A similar finding was shown by Belsky, Spritz, and Crnic (1996) in a sample of American boys whereby boys with secure attachment histories (assessed by the Strange Situation at 12 months), when assessed again at 3 years of age, remembered positive events more accurately than negative events. The reverse was true of children with insecure attachment histories. This prosocial and positive propensity of secure children was also observed in the generation and evaluation of reactions to peers. Cassidy et al. (1996) found that American preschool children with secure and ambivalent patterns (assessed by the Strange Situation at 12 months) generated more positive reactions for problematic social situations than did avoidant children. Additionally, avoidant preschool children generated fewer (half as many) solutions for problematic social situations than children with secure and ambivalent patterns. Similarly, in an American sample of 6- to 7-year-olds, secure children assessed by a reunion procedure concurrently generated more positive reactions than others in processing social events. Secure attachment in Israeli 7- to 8-year-old children (assessed by the Strange Situation at 12 months) was also associated with higher tendency to evaluate

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

542

Journal of Early Adolescence 32(4)

competent-dialog responses as having a positive social outcome (Ziv, Oppenheim, & Sagi-Schwartz, 2004). Finally, Bauminger and Kimhi-Kind (2008), in a sample of 10- to 12-year-old Israeli boys (50% of them with learning disabilities), found that perceived security in attachment relationships with mother, as assessed through a self-report scale, added significant explained variance to these children’s social-information processing. The highest contribution was obtained for children’s ability to generate a competent solution. In sum, previous studies, mostly with early childhood samples, focused on the secure/insecure division and mainly showed associations between secure attachment pattern and social-information processing, characterized by a low tendency to hostile attribution and a high tendency to generate prosocial responses. In line with previous studies, in the present study we too focused on attachment to mother, who is considered the most significant attachment figure in nonclinical samples (Schneider, Atkinson, & Tardif, 2001). We further extended these studies in various ways. First, we focused on early adolescence (10- to 12-year olds), an age when the arena of peer relationships is highly central and important. Compared with early childhood, children in early adolescence have higher flexibility in thinking, greater social knowledge, and higher ability to understand peer social responses; their relationships with peers are more intense, complex, and important to them (for a review see Bukowski, Newcomb, & Hartup, 1996; Kerns, 2008). Hence, examination of social-information processing in the peer arena at this age may allow the identification of a multifaceted and complex picture of individual differences, and help us find the expected orientations of socialinformation processing for all four patterns of attachment representations. Second, to have sufficient statistical power to go beyond the secure/insecure division and examine all four patterns of attachment and their associations with the four SIP orientations, we assembled a sample larger than prior studies (Belsky et al., 1996; Cassidy et al., 1996; Suess et al., 1992; Ziv et al., 2004), which included children of both sexes. The number of participants allowed us to elucidate and identify the four latent SIP orientations using Structural Equation Modeling (SEM). Third, most SIP studies focused on hostile and nonadaptive SIP characteristics (hostile attribution bias, aggressive SIP reactions, or depressive-withdrawn SIP reactions), within at-risk populations (Crick & Dodge, 1994; Nelson & Crick, 1999). To broaden our understanding of the significance of SIP in the general population, as well as to allow us to elucidate individual differences in prosocial SIP orientation, which is rare among at-risk populations, we relied on a normative rather than at-risk sample.

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

543

Hypotheses Hypothesis 1: Secure children, with their open, flexible, and nondefensive emotion-regulation strategy, were expected to be characterized by a high tendency to cooperate with others and a low tendency to confront them; specifically, in problematic social situations with peers, they were expected to show a high tendency to prosocial SIP orientation and a low tendency to antisocial-aggressive SIP orientation. Hypothesis 2: Avoidant children, with their deactivating emotionregulation strategy, were expected to be characterized by a tendency to minimize responses in general, and especially responses promoting interactions with others; namely, they were expected to generate fewer responses, and to show a low tendency to prosocial and distress expression SIP orientations and a high tendency to withdrawal/ avoiding SIP orientation. Hypothesis 3: Ambivalent children, with their hypervigilant and anxious emotion-regulation, were expected to be characterized by a heightening of expressions of distress or dependency in order to coerce others to respond to their needs; in their social-information processing they were expected to show high tendency to use a distress expression SIP orientation. Hypothesis 4: Disorganized children, with their incoherent bipolar (hostile-helpless) emotion-regulation strategy, were expected to be characterized by a high tendency to coerce others to respond to their needs by hostile means, and simultaneously to yield helplessly when that approach was conceived as unrealistic; in their social-information processing they were expected to show high tendency to use an antisocial-aggressive SIP orientation reflecting the hostile aspect, together with a distress-expression SIP orientation reflecting the helpless aspect.

Method Participants The sample comprised 185 children in fourth and fifth grades (107 in fourth grade and 78 in fifth grade), 97 males and 88 females. Their mean age was 10.35 years (range = 9.00-11.7 years, 73% were 10-year-olds or older, with all others, except seven, between 9.5 and 10 years of age. The sample was recruited from the two elementary schools that serve one middle-class

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

544

Journal of Early Adolescence 32(4)

suburban area in Israel, and this suburban area is quite typical of middle-class families in Israel. Specifically, in line with the neighborhoods where the schools are located, the Israel Ministry of Education normative SES index of the two schools was 2.6 and 3, representing nonrisk middle and high-middle class student composition. This normative index supports decisions regarding extra funding of schools and reflects the SES level of the families whose children attend each school. It is expressed in percentiles ranging from 1 to 10 with lower scores (higher percentiles) reflecting better SES. All the fourth and fifth-grade classes in these schools were sampled. Parental permission was obtained for participation and compliance was very high at 90%. In line with the middle-class neighborhoods of these schools, parents were largely well educated (54% of the fathers and 53% of the mothers had at least a college education); 48% of the families were of Western origin, and their living quarters were of moderately low density (for 72% of the families there was one person or less per room)

Procedure The study involved gathering data at two points in time during one school year as part of a larger research project. At Time 1 (during the first trimester of the school year), each child was escorted from his or her homeroom to an adjacent quiet room where he or she was administered the Attachment Doll Story Completion Task for Middle Childhood—ADSCT for m-c (Granot & Mayseless, 2001); the procedure was videotaped. Two and a half months later (Time 2), each child individually completed a background questionnaire and was administered the Social-Information Processing Interview—SIPI (Erdley & Asher, 1996; Quiggle, Graber, Panak, & Dodge, 1992), one of the most widely accepted models for social-information processing (Crick & Dodge, 1994). The northern districts of the Israel Ministry of Education administered standardized reading comprehension, written expression, and math proficiency regional tests five times during that school year (September to June) to all the elementary schools in these regions. At the end of the year, the schools calculated a standardized score for each child averaging across the five test scores in each subject. For the present study, at the end of the school year, we obtained these standardized scores for all the children who participated in the study from the school records.

Measures Attachment measure.The ADSCT for m-c (Granot & Mayseless, 2001) assesses the child’s representations of attachment to the parent in terms of

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

545

security, avoidance, ambivalence, and disorganization. The measure is an adaptation of Bretherton, Ridgeway, and Cassidy’s (1990) story-stem technique, in which an interviewer begins to tell a story and the child uses dolls and props to complete the story. The target child (the protagonist), one parent (in the present study, the mother), and siblings are presented as characters in the story but are assigned different names. Five stories that had been used with younger children (3- to 6-year-olds) were adapted to make stories appropriate for children 9-12 years of age. The themes of the stories are (a) spilled juice: while the family is seated at the dinner table the child accidentally spills juice on the floor; (b) hurt knee: the child tumbles off a high rock and hurts his or her knee and it bleeds; (c) scary figure: the child is sent to bed and cries out that there is a scary figure in his or her bedroom; (d) departure story: the mother leaves for a 3-day trip and a babysitter stays with the children (older siblings are taken out); (e) reunion story: the babysitter sees the caregiver as he or she returns from the 3-day trip and announces his or her arrival to the children. The interviewer introduces each story with a short script and encourages the child to enact the story completion: “Show me what happened next.” Open-ended questions and prompts are provided, when necessary, to encourage the child to play. The procedure takes about 45 to 60 minutes and is videotaped. Following extensive pretesting with 30 children, several modifications were made for middle childhood and early adolescence. These included changes in story content, in the administration procedure, and in the criteria for security. They also included a classification into four-attachment prototypes (by coding the videotaped interview). The changes in content were planned to adjust the stories’ stress level and content to middle-childhood and early adolescence. (a) The monster in the Monster in the Bedroom story stem was changed to scary figure to make it more appropriate for older children; (b) in the original Hurt Knee story stem the child, crying after falling from the rock, says, “I’ve hurt my knee.” The sentence “I am bleeding” was added to make it more stress inducing for older children; (c) in the Departure Story, the separation was lengthened from 1 night to 3 days to make it more stress inducing for older children. In addition, the older siblings were removed so that they would not act as replacement caregivers. The changes to the administration procedure were planned so that the child’s story completions would include specific details about the events, both verbally and procedurally. For example, the interviewer asks: “So what exactly happened when . . .?” “How exactly does he or she do that?” “Show me how this happens,” and so forth. The request to enact the details of the story completion with puppets enhances the possibility of invoking the child’s actual personal experience based on the procedural and the episodic memories (partly without his or her conscious awareness) and less on general semantic knowledge about the maternal role based on

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

546

Journal of Early Adolescence 32(4)

cultural scripted knowledge. Finally, minor changes were also made in criteria for security in specific stories; for example, the child did not have to enact “looking for the parents” to be classified as secure in the Departure Story, in accordance with developmental expectation for older children. A coding system for each story stem, and an overall coding of the five stories, were used to analyze and classify the child into one of four attachment prototypes (analogous to the four attachment patterns). This was accomplished in two stages: (a) in the prototype overall ratings each participant was rated, on a 1-5 Likert-type scale from 1 (not typical at all) to 5 (very typical), for how characteristic his or her stories were of each of the four prototypes; (b) in the prototype-based classification, the participant was assigned to one prototype. Prototype ratings and prototype-based classifications were both used in the statistical analyses. The descriptions of the four prototypes cover four aspects: (a) emotional expression, (b) relationship with caregiver, (c) protagonist worldview, (d) narrative structure. These aspects could be reflected in the story completions in verbal and nonverbal (physical contact, moving away) means of communication between the protagonist and the caregiver figures. Children were classified into the secure prototype if the emotional expression of the child in the story was spontaneous and situation appropriate, the relationship with the caregiver was depicted as warm, accessible, stable, supporting, trusting, and mutual, and, in the protagonist’s worldview, threatening situations that might arise were resolved positively by the adult’s responsible and reliable intervention. In addition, the story’s narrative structure was diverse, logical, and integrated. Children were classified into the avoidant prototype when the emotional expression of the child in the story was shallow, the relationship with the caregiver was characterized by distance, lack of emotional accessibility, diminished significance, and instrumental behavior, and the protagonist’s worldview was characterized by a neutral attitude to threatening situations, and a perception that one must rely on oneself only and not expect help from others. In addition, the story’s narrative structure was minimalist, informative, and routine. Children were classified into the ambivalent prototype when the emotional expression of the child in the story was exaggerated, inappropriate, and inconsistent, the relationship with the caregiver was characterized by lack of accessibility on his or her part and overdependence and feelings of rejection, with the child blaming the caregiver, and, according to the protagonist’s worldview, coercive action was necessary to ensure appropriate caregiver action in threatening situations. In addition, the story’s narrative structure may be somewhat diverse, open to unpredictable changes, and overloaded with irrelevant details that describe scenes not related to the situation. Children were

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

547

classified into the disorganized prototype when the emotional expression of the child in the story was somewhat incoherent, the relationship with the caregiver was characterized as unbalanced, the child was seen as hostile, tending to control the adult, or helpless in relations with the caregiver, and, in the protagonist’s worldview, reality was frightening, and no one (including the caregiver and the protagonist) could be trusted as events were typically dangerous, and escape from danger was only temporary. In addition, the story’s narrative structure was bizarre; frequently lacking logical continuity and plausibility, and the overall atmosphere was chaotic. Granot and Mayseless (2001) showed good interjudge agreement (80% category agreement, scale r ranged from .78 to .85) and high test-retest reliability (94%) over a 3-month interval in another Israeli sample including both genders. The ADSCT (for m-c) was also significantly associated with teacher’s ratings of school adjustment and peer nominations, with children classified as secure showing better adjustment than children classified as avoidant or disorganized. Attachment classifications were not related to logical reasoning and language test scores showing discriminate validity. In the present study, two coders (D. G. and I. K., who did not participate in the data-collection phase) rated 52 interviews after evaluating 20 trial cases from a different sample. There was good agreement between the two coders on the classifications of the children into the four prototype-based classifications in 50 of the 52 cases (Kappa = .92). The two additional cases were resolved through joint discussion. Both judges (D. G. and I. K.) proceeded individually to rate all the remaining videotapes (62 each), transferring them to another rater whenever a classification was not clear cut. Ten more cases were resolved in this way. Interjudge reliabilities for the prototype ratings of attachment were high (r ranged from .80 to .85). As another partial validation of the ADSCT for m-c in the present study, attachment prototypes ratings were moderately associated, as expected, with self-report measures of attachment: The Attachment Security Scale—ASS (Kerns, Klepac, & Cole, 1996) and the Preoccupied and Avoidant Coping Strategies Questionnaire (Finnegan, Hodges, & Perry, 1996). In line with previous studies, we expected only modest associations when different methods were used to assess attachment dimensions (Raikes & Thompson, 2005; van IJzendoorn, Vereijken, Bakermans, & Riksen, 2004). Self-report of security of attachment was positively associated (r = .23; p < .001) with the rating of the secure prototype and negatively associated (r = –.27; p < .001) with the rating of the avoidant prototype. Self-report of avoidant coping strategies was positively associated (r = .24; p < .01) with the rating of the avoidant prototype. Self-report of preoccupied coping strategies was positively, but only

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

548

Journal of Early Adolescence 32(4)

marginally, significantly associated (r = .13; p < .06) with the rating of the ambivalent prototype. Social-information processing measure.The SIPI for 8- to 12-year-olds (Erdley & Asher, 1996; Quiggle et al., 1992) was used, following some modifications, to tap children’s social-information processing in relations with peers. This measure is based on Crick and Dodge’s (1994) revised SIP model. The modified interview includes four short social vignettes, two peer entry stories (told from the viewpoint of the child attempting the entry), and two stories of provocation in which the child is the victim, one including intentional provocation and the other, ambiguous. For example, one of the peer entry vignettes is “Imagine that some kids you know are sitting at a table eating sandwiches. You can see that they are having a good time and you’d like to sit with them. You walk up to them ask them if they’d make room for you so you could sit down too. They tell you, No.” The interviewer read each of the four vignettes aloud to each child individually and then asked a series of question aimed at examining four steps in the child’s social-information processing: interpretation (attribution of hostile intent), clarification of goals, response generation, and response evaluation. In the attribution of hostile intent (interpretation) step, participant children were asked how much they thought what happened in the story was due to the deliberate malevolent intent of another (e.g., “How much do you think the kids who said ‘No’ were trying to be mean?”). They responded on a 4-point rating scale ranging from 1 (not at all) to 4 (very much). Scores were summed across the four stories resulting in a 4 to 16-point range (Cronbach’s α = .60). In the clarification of goals step, four general social goals were presented for each story as possible alternative ways of reaction to the situations: (a) prosocial (e.g., “Find out with the children why they had refused to accept you”); (b) distress expression (e.g., “Show them that they hurt you very much”); (c) antisocial-aggressive (e.g., “Retaliating physically or verbally”); and (d) withdrawal/avoidance (e.g., “Showing that you don’t need them”). Participants were asked to circle the reaction they believed they would “most likely” be their reaction. The number of times each goal alternative was circled “most likely” was summed across the four stories. A score range of 0 to 4 was possible for each social goal. In the response generation step, following the goal clarification step, the children reported what they thought they would do if this happened to them. They were probed to generate multiple responses (up to four per story). Each response was later coded into one of four mutually exclusive categories: (a) prosocial, (b) distress expression, (c) antisocial-aggressive, (d) withdrawal/ avoidance. Prosocial responses were those in which the child defends his or

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

549

her opinion or goal by negotiating, requesting explanation, bargaining, or persuading assertively. Responses were coded as distress expression if they involved begging, nervousness, helplessness, moodiness, and taking the blame. Antisocial-aggressive responses were those that involved an act of physical or verbal aggression, retaliation, or teasing. Withdrawal/avoidance were responses that involved looking for other possibilities/environments or walking away from the situation, deliberately ignoring or behaving as if nothing important had happened. Scores were calculated by summing the number of reactions coded into each category across the four stories. In the present study, the number of reactions ranged from 4 to 17 per participant. Intercoder reliability calculated for 20% of the responses chosen randomly yielded 92% agreement (disagreements were settled by joint discussion; Kappa = .88. One of the coders (S. A.) classified the remaining reactions. Following the response generation step, in the response evaluation step, participants were read four types of responses supposedly provided by other children: (a) prosocial, (b) distress expression, (c) antisocial-aggressive, (d) withdrawal/avoidance. Examples of each type of response for the entry story are (a) prosocial: “One kid said, ‘I think there would be room for me if you moved over a little. I’d really like to sit with you”; (b) distress expression: “One kid overreacted, behaving like he or she was very upset (distressed, insulted), or very angry (mad)”; (c) antisocial-aggressive: “One kid just pushed between the kids and sat down anyway without asking their permission”; (d) withdrawal/avoidance: “One kid just turned his or her eyes away and looked for someplace else to sit and eat.” Participants rated how likely it was that they would respond as described, from 1—“definitely would not” to 4—“definitely would”. In line with the procedure in former SIP studies conducted in Israel, as well as in other countries (e.g., Bauminger & Kimhi-Kind, 2008; Bryan, Burstein, & Ergul, 2004), scores were summed across the four stories for each type of alternative response. The children completed A Children’s Background Questionnaire that related to parents’ education and ethnic origin, number of rooms in the children’s home, and number of people in the household. These indicators are accepted in Israel as good indices of Socioeconomic Status (SES; Yair, 1991). Three SES indicators were used: (a) parent’s education, (b) ethnic origin, and (c) the ratio of people to rooms in the child’s home. The first two indicators were each coded into three categories. Parents’ education was coded into (a) did not complete high school, (b) completed high school, and (c) received an academic qualification; while the following categories were used for ethnic origin: both parents of Eastern origin, one parent of Eastern and the other of Western origin, both parents of Western origin. A higher level of education, Western ethnic

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

550

Journal of Early Adolescence 32(4)

origin and a low ratio of people to rooms are considered indicative of higher levels of socioeconomic status (Yair, 1991). The Israeli Ministry of Education’s Normative, Standardized, Regional Test was used to assess the children’s reading comprehension, written expression, and math proficiency. Scores were obtained from the school’s records at the end of the school year. These scores reflect the children’s abilities throughout the school year and are the mean of five different tests administered in each subject during the year. Expressive language skills and math proficiency are considered to reflect cognitive ability, which might affect the assessment of attachment representations and social-information processing (Parke & O’Neil, 1999). Collecting this information enabled us to statistically control this possible bias.

Results Preliminary Analyses The distribution of attachment prototypes was as follows: 66.5% (n = 123) of the children were assigned a secure classification, 14.1% (n = 26) an avoidant classification, 9.2% (n = 17) an ambivalent classification, and 10.3% (n = 19) a disorganized classification. There were no significant differences between children classified as secure, avoidant, ambivalent, or disorganized with regard to age, socioeconomic indicators, language skills (reading comprehension, written expression), and math proficiency. The effect of gender on secure _classification was significant, F(3,183) = 4.35; p < .05: compared _ to boys (X = 3.5; SD = 1.2) girls scored a higher secure rating (X = 3.9; SD = 1.2; Effect size ŋ2 = .02), gender effect was also significant _ for avoidant classification, F(3,183) = 9.33; p < .01:_compared to girls (X = 1.5; SD = .9) boys scored a higher avoidance rating (X = 2.0; SD = 1.2; Effect size ŋ2 = .05). Consequently the effect of gender was statistically controlled when secure and avoidant ratings were used.

Structural Equation Modeling (SEM) SEM utilizing the Amos 7.0 statistical program (Arbuckle, 2006) was used to test models of the hypothesized relations among attachment representation and SIP variables (for range, means, SD, and correlations among SEM variables see Table 1). Evaluations of the models’ fit were conducted using six indicators. First, we used the chi-square indicator, but because a large number of scholars

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

551

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

13

14

15

16 17

.08 .01

−.00 .06

−.17* .04

−.08

7.1 2.6 −.06

6.2 2.2 −.10

12.0 1.9 .06

4-16

6-16

SIP R-E antisocial

SIP R-E withdrawal

— —

.07

−.36** −.08 .40**

.24* −.07

.03

.06

.08

−.04 .20**

−.12

.09

.03

.17* −.27**

.24** −.04

.04

−.01 −.14

— — —

.39**

.35*

.11

−.04 −.12

.08

.09

.13 −.02

−.09 −.11

.07

.06 −.04

—

−.32**

−.05

.10

— —

−.06

.09

−.07

−.20**

−.09

−.25**

−.02

.07

.06

.13

.24** −.03

−.15*

—

−.09

—

−.04

.17*

    .23* —  

—

 

 

 

 

 

 

 

 

 

 

.26** .16* .10 −.09 —

.41** −.01

.15* −.00

−.10

.25** −.25** −.20** −.13

.07

−.08

.06 −.27** −.24*

−.60** −.19* −.14*

.21* .25* −.30**

17*

−.05 −.14

−.03 −.11

.03

.17* .23* −.08

−.05 −.13

.08

—

Note: A-P, attachment prototype (secure, avoidant, ambivalent, disorganized); SIP C-G, social information processing clarification of goals (prosocial, distress expression, antisocial, withdrawal); SIP R-G, social information processing response generation (prosocial, distress expression, antisocial, withdrawal); SIP R-E, social information processing response evaluation (prosocial, distress expression, antisocial, withdrawal). Effect of gender was partialled out for associations with secure and avoidance attachment ratings. *p < .05. **p < .01.

.23**

11.4 2.2 .14* −.22** −.02

4-16

4-16

SIP R-E prosocial

SIP R-E distress expression

.03

1.6 1.7 −.16* −.04

3.4 2.0 −.01

0-16

0-16

.04

SIP R-G antisocial

.00 .09

−.25*

4.0 2.7 .13

0.4 0.8 −.06

0-16

0-16

SIP R-G prosocial

SIP R-G distress expression

SIP R-G withdrawal

.07

.13 .13

0.5 0.8 −.13

1.4 1.2 −.13

0-4

0-4

SIP C-G antisocial

SIP C-G withdrawal

−.02

−.04

−.10

0.22 0.4 .01

0-4

SIP C-G distress expression

.14*

0-4

−.07

8.9 2.3 −.08

1.52 1.2 .24** −.26* −.02

SIP hostile attribution 4-16

SIP C-G pro-social

.01

   

— .32**

1.6 0.9 −.46** −.28*

1.4 0.9 −.55**

12

1-5

11

1-5

10

A-P ambivalent

9

A-P disorganized

8

 

7

—

—

6

3.7 1.1

5

1.73 1.0 −.47**

4

1-5

3

1-5

2

A-P secure

1

A-P avoidant

SD

_ X

Variable

Range

Table 1. Range, Means, Standard Deviations, and Correlations Among Structural Equation Modeling variables

552

Journal of Early Adolescence 32(4)

(e.g., Schermelleh-Engel, Moosbrugger, & Muller, 2003) have discussed the shortcomings associated with this, we also added the relative chi-square (χ2/df ratio) that is considered less sensitive to sample size (Ullman, 2001). In addition, we used the Adjusted Goodness-of-Fit Index (AGFI), Comparative Fit Index (CFI), p of Close Fit (PCLOSE), and Root Mean Square of Error Approximation (RMSEA) fit indices. According to conventional guidelines, a nonsignificant chi-square is considered a good fit, a χ2/df ratio of 2-3 or less is a reasonably good indicator of model fit, AGFI greater than .90 is considered a good fit, CFI of .90 and above indicates an acceptable fit to the data, PCLOSE greater than .5 reflects that the fit of the model is “close” and RMSEA of .06 or less is considered to be a good fit (Hu & Bentler, 1999; Ullman, 2001).

Generating Latent Variables for SIP Overall Orientations Based on the suggestions of Crick and Dodge (1994) and Von Salisch (2001), we postulated the existence of four main SIP orientations of response in problematic social situations: Prosocial, antisocial, distress expression, and withdrawal/avoiding. These four orientations were assessed in three of the SIP steps assessed in this study: clarification of goals, response generation, and response evaluation. In contrast, the assessment of the first SIP step, which involves attribution of hostile intent, does not directly map into one specific SIP orientation. Dodge (1993) hypothesized that early experiences of insecure attachment relationships (among other things) lead the child to develop memory structures of the world as a hostile place. These knowledge structures predispose the child to be hypervigilant to hostile cues, which in turn contributes to a hostile attribution bias. Perception of the world as a hostile place could be coupled with two SIP orientations (Crittenden, 1992; Dodge, 1993; Von Salisch, 2001), one in which the hostile intent of others requires coercive behavior to achieve desired outcomes (antisocial SIP orientation) and the other in which the hostile intent of others requires exaggerated display of weakness and distress to appease the other (distress expression SIP orientation). Therefore, the present study examined the possible inclusion of attribution of hostile intent in each of these SIP orientations. In the SEM models examining the study hypotheses, all four SIP orientations (prosocial, antisocial, distress expression, and withdrawal/avoiding orientations) were simultaneously included, hence each observed variable could only be entered once. We therefore ran two different SEM measurement models of all four SIP orientations, one in which attribution of hostile intent was included in the distress SIP orientation latent variable and another in which it was included in the antisocial SIP orientation.

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

553

In these measurement models, the prosocial SIP orientation was approved as a latent construct with three observed variables: prosocial clarification of goals (β = .66), prosocial response generation (β = .63), and prosocial response evaluation (β = .29). The fit indicators for the withdrawal/avoiding SIP orientation measurement model were less than acceptable and none of the deletions of observed variables changed this. Therefore, we did not construct and evaluate the withdrawal/avoiding SIP orientation as a general orientation. We did, however, examine possible differences among attachment prototypes with regard to these observed variables of the withdrawal/ avoiding SIP orientation. The antisocial SIP orientation was also approved as a latent construct with three significant observed variables: Antisocial clarification of goals (β = .57), antisocial response generation (β = .70), and antisocial response evaluation (β = .59). For the distress expression SIP orientation, the estimate weight of correlations with the observed variable of distress expression—clarification of goals—was too low, therefore this observed variable was deleted from this SIP orientation. Attribution of hostile intent was more clearly included in the distress expression SIP orientation than in the antisocial SIP orientation. When hostile intent was included in the antisocial SIP orientation χ2 = 35.4 (p = .06), χ2/df ratio = 1.50, AGFI = .92, CFI = .94, RMSEA = .05, and PCLOSE = .45, and estimate weight of correlations = .31, whereas when hostile intent was included in the distress expression SIP orientation χ2 = 28.2 (p = .25), χ2/df ratio = 1.20, AGFI = .94, CFI = .98, RMSEA = .03, and PCLOSE = .75 and estimate weight of correlations = .61. Consequently, in the present study, attribution of hostile intent was used as an observed variable only as part of the estimation of the latent variable assessing the distress expression SIP orientation, and this SIP orientation was approved as a latent construct with three significant observed variables: Attribution of hostile intent (β = .61), distress expression response generation (β = .30), and distress expression response evaluation (β = .42). Following these analyses, three SIP orientations were simultaneously assessed, each including three indicators, with distress expression orientation including the attribution of hostile intent. In the measurement models, we also assessed the associations between the different SIP orientations as latent variables. The prosocial orientation was associated negatively with the antisocial orientation (r = –.67, p < .01), and the antisocial orientation was associated positively with the distress expression orientation (r = .43, p < .01). The prosocial and the distress expression orientations were not significantly associated (r = –.21, p = .20).

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

554

Journal of Early Adolescence 32(4)

.67 SIP Pro-social orientation .31

Gender

SIP Antisocial aggressive orientation

–.22

–.09

Clarification of goals-pro-social Response generation-pro-social

.31

Response evaluation-pro-social

.57

Clarification of goals-antisocial

–.67

.15 Secure attachment prototype rating

.60

.69 .59

Response generation-antisocial Response evaluation-antisocial

.43

SIP Distress expression orientation

.54

Attribution of hostile intent

.32

Response generation-distress expression

.48

Response evaluation-distress expression

Figure 1. SEM of secure prototype and SIP prosocial, antisocial, and distress expression orientations Note: Solid lines indicate statistically significant relations (p < .05); dashed lines indicate nonsignificant relations. Double-sided arrows connecting the latent variables are correlations among the errors of these variables. This model provided a good fit to the data: χ2 = 57.4, p = .04, χ2/df = 1.44, AGFI = .92, CFI = .92, RMSEA = .05, and PCLOSE = .51.

Attachment Prototype Ratings and Social-Information Processing—SEM Analyses Consistent with Hypothesis 1, Model 1 (presented in Figure 1) in which gender was included as a control variable provided a good fit to the data, showing that the secure prototype rating was positively associated with the tendency to use the prosocial SIP orientation and negatively associated with the tendency to use the antisocial SIP orientation in problematic social situations. In contrast, the secure prototype rating was not associated with the tendency to use the distress expression SIP orientation. Consistent with Hypothesis 2, Model 2 (presented in Figure 2) in which gender was included as a control variable provided a satisfactory fit to the data, showing that the avoidant prototype rating was negatively associated with the tendency to use the prosocial SIP orientation and negatively associated with the tendency to use the distress expression SIP orientation. The avoidant prototype rating was not significantly associated with the tendency to use the antisocial SIP

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

555

Granot and Mayseless

.64 SIP Pro-social orientation

–.42

Gender

SIP Antisocial aggressive orientation

.05

–.20 (p=.07)

Clarification of goals-pro-social Response generation-pro-social

.33

Response evaluation-pro-social

.55

Clarification of goals-antisocial

–.67

–.20 Avoidant attachment prototype rating

.62

.70 .59

Response generation-antisocial Response evaluation-antisocial

.43

SIP Distress expression orientation

.49

Attribution of hostile intent

.27

Response generation-distress expression

.55

Response evaluation-distress expression

Figure 2. SEM of avoidant prototype and SIP prosocial, antisocial, and distress expression orientations Note: Solid lines indicate statistically significant relations (p < .05); dashed lines indicate nonsignificant relations. Double-sided arrows connecting the latent variables are correlations among the errors of these variables. This model provided satisfactory fit to the data: χ2 = 65.5, p = .007, χ2/df = 1.64, AGFI = .91, CFI = .90, RMSEA = .06, and PCLOSE = .27.

orientation. Consistent with Hypothesis 3, Model 3 (presented in Figure 3) provided a good fit to the data, showing that the ambivalent prototype rating was positively associated with the tendency to use the distress expression SIP orientation. Yet the ambivalent prototype rating was not associated with the tendencies to use the prosocial and the antisocial SIP orientations. Finally, consistent with Hypothesis 4, Model 4 (presented in Figure 4) provided a good fit to the data, showing that the disorganized prototype rating was positively associated with the tendency to use the antisocial SIP orientation and the tendency to use the distress expression SIP orientation. The disorganized prototype rating was not associated with the tendency to use the prosocial SIP orientation.

Attachment Prototypes and Social-Information Processing of the Withdrawal/Avoiding Indices Four ANOVAs were conducted with gender and the four attachment prototype groups serving as independent variables and clarification of goals— withdrawal/avoiding, response generation—withdrawal/avoiding, response

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

556

Journal of Early Adolescence 32(4)

.64 SIP Pro-social orientation

–.02

Ambivalent attachment prototype rating

.63

Response generation-pro-social

.30

Response evaluation-pro-social

.57

Clarification of goals-antisocial

–.67

SIP Antisocial aggressive orientation

.04

.68

.60 .36

Clarification of goals-pro-social

Response generation-antisocial Response evaluation-antisocial

.43

SIP Distress expression orientation

.46

Attribution of hostile intent

.27

Response generation-distress expression

.56

Response evaluation-distress expression

Figure 3. SEM of ambivalent prototype and SIP prosocial, antisocial, and distress expression orientations Note : Solid lines indicate statistically significant relations (p < .05); dashed lines indicate nonsignificant relations. Double-sided arrows connecting the latent variables are correlations among the errors of these variables. This model provided a good fit to the data: χ2 = 33.6, p = .34, χ2/df = 1.08, AGFI = .94, CFI = .99, RMSEA = .02, and PCLOSE = .86.

evaluation—withdrawal/avoiding, and response generation—total number of responses indices as dependent variables. None of the gender main effects and the interactions with attachment groups was significant. The effect of attachment groups on preference for withdrawal/avoidance _ goals was significant, F(3,183) = 2.88; p < .05) with the avoidant group_(X = 2.0, SD = 1.5), exhibiting higher preference than the secure group (X = 1.2, SD = 1.2; Effect size ŋ2= .05). Attachment effect was also significant, (F(3,183)_= 4.65; p < .01) for number of responses generated: the avoidant group (X_= 7.8, SD = 2.6) generated the smallest number of responses (For all others, X = 9.9, SD = 2.5; Effect size ŋ2 = .07).

Discussion Our findings demonstrated significant associations between attachment representations with mother and social-information processing in the peer arena in four SIP steps (attribution of hostile intent, preference of goals, generating possible responses, evaluating response alternatives). As expected, secure mother-child attachment representations were positively associated

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

557

Granot and Mayseless

.64 SIP Pro-social orientation –.08

Disorganized attachment prototype rating

Clarification of goals-pro-social Response generation-pro-social

.30

Response evaluation-pro-social

.55

Clarification of goals-antisocial

–.67 SIP Antisocial aggressive orientation

.20

.22 (p=.06)

.63

.71 .58

Response generation-antisocial Response evaluation-antisocial

.43

SIP Distress expression orientation

.54

Attribution of hostile intent

.35

Response generation-distress expression

.47

Response evaluation-distress expression

Figure 4. SEM of disorganized prototype and SIP prosocial, antisocial, and distress expression orientations. Note: Solid lines indicate statistically significant relations (p < .05); dashed lines indicate nonsignificant relations. Double-sided arrows connecting the latent variables are correlations among the errors of these variables. This model provided a good fit to the data: χ2 = 37.2, p = .21, χ2/df = 1.20, AGFI = .93, CFI = .97, RMSEA = .03, and PCLOSE = .76.

with social-information processing reflecting a prosocial orientation and negatively associated with antisocial orientation. These findings suggest that the goal-corrected partnership experienced by secure children in relationships with their mothers shapes a similar orientation to peers in socialinformation processing. As part of such orientation, secure children forgo instinctive antisocial reactions (such as retaliation) when these may harm the relationships, and instead focus on responses that promote social integration with peers. The profile of results regarding avoidant attachment representations was, to some extent, the converse of the profile with attachment security. Avoidant attachment, examined by a continuous rating scale, was negatively associated with prosocial and distress expression social-information processing orientations. In addition, adolescents in the avoidant category showed the highest preference for the withdrawal/avoidance goals in peer relations, and in line with their deactivating emotion-regulation strategy, also had the lowest number of generated responses of any kind. These findings accord with our expectations that avoidant early adolescents would evince an orientation away from others in their social-information processing in the peer arena. For the avoidant child, early experiences with a caregiver who

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

558

Journal of Early Adolescence 32(4)

consistently ignored the child’s distress are expected to promote a tendency to minimize dealing with such experiences and to withdraw from close interactions (Cassidy, 2008). Interestingly, avoidant representations were not associated with higher antisocial SIP orientation, but rather with deactivation of the affect (i.e., low number of responses generated, low endorsement of distress exhibition) and with turning away from contact and closeness in the peer arena. As expected, ambivalent attachment representations were positively associated with SIP distress expression orientation. As expected, disorganized prototype ratings were associated positively with SIP distress expression and antisocial orientations, supporting the expected existence of dual emotion regulation strategies (helplessness and aggressiveness/hostility). In this nonclinical and generally well-functioning sample, where the open and intentional endorsement of antisocial SIP orientation may be less prevalent and less expected, the association between disorganized attachment representations and antisocial SIP orientation reflects a clear risk for the development of severe forms of aggressiveness and hostility in adolescents with disorganized attachment representations. In general, our findings empirically support the assertion of a clear association between attachment-related emotion-regulation strategy (at the level of representation) for maintaining mother’s accessibility in stressful situations, and the tendency to adopt a similar orientation for social-information processing when trying to react to problematic situations in early adolescent relationships with peers. This finding is consistent with conceptualizations of attachment theory (Bretherton & Munholland, 2008) and of social-information processing theory (Crick & Dodge, 1994), but it is not trivial. It indicates a generalization of social knowledge and regulation strategies from the attachment system, a close and hierarchical relationship with an adult caregiver with whom a unique and hardly replaceable bond has been established, to peer relationships, which are based on a symmetrical affiliation with sameage partners. In the vignettes provided in this study, the peers depicted were those in the general network of the peer group and not a preferred peer with whom a special or unique close relationship was established—a relationship which has been described as reflecting attachment dynamics (Mayseless, 2005). Hence our findings support a nontrivial association between two different social systems—attachment and affiliation, of differing evolutionary predisposition (Furman & Simon, 1999). What might be the underlying structure that facilitates such a transfer? Children in early adolescence, like their younger counterparts, possess a relationship-specific working model that represents attachment in particular relationships. By early adolescence, however, children have also developed

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

559

general working models that represent attachment state of mind across a variety of relationships and relationship contexts (Kerns, Schlegelmich, Morgan, & Abraham, 2005). Overall, Fletcher, and Friesen (2003) suggest that these specific and general working models are cognitively organized and structured as a multilevel network of attachment representations. The findings of the present study show an association between a specific relationship with the mother, which is part of the general familial domain, and a generalized pattern of a relationship in another, namely the peer domain. Based on Overall et al.’s (2003) conceptualization, we should expect only a small degree of association between two such different levels of relational contexts, as indeed was the case in our study. In contrast, and in line with Overall et al.’s (2003) proposed hierarchical model of representations, Kerns et al. (2005), using the Doll story adaptation, found moderately strong concordance in children’s attachment patterns with mother and with father, which are part of the same relational arena. Our results are also consistent with a model of transfer of expectations and emotion regulation from one domain to another based on learning processes that do not assume a hierarchy of cognitive schemas, but all in all may provide preliminary corroboration of the hierarchical model of attachment representations. For example, future studies might explore children’s attachment-relationship representations of friendships with peers (Furman, 2001) and their social-information processing in close friendships, and compare these with social-information processing in the general network of peers (as assessed in this study). Such studies could advance our understanding of the way the representations are organized, and shed light on their interconnections as well as the way the general schemas develop and possibly transfer from one arena to another. Another avenue for future research would be to examine emotion regulation strategies as mechanisms of transfer between the attachment and peer arenas. This study differs from previous studies that assessed the association between attachment and social-information processing in the peer domain in four major ways. First, this study is the first to associate the four patterns of early adolescent attachment representations with three different social-information processing general orientations in peer relationships; thereby providing a more complete understanding of the predispositions of adolescents with different attachment representations. Second, most social-information processing studies focused on the specific steps in the SIP model and examined their contribution to adjustment. In line with the suggestions of central researchers in this area (e.g., Dodge, 1993) the present study adopted a more general view of SIP and assessed four general SIP orientations postulated to be reflected in the different SIP steps. Such assessment, using SEM, provided clear support for the proposed latent structures underlying the different steps. Several

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

560

Journal of Early Adolescence 32(4)

difficulties in assessing these latent constructs call for further research. For example, the choice of distress expression as a goal in peer relations (clarification of goals step) was not included in the latent structure of the distress expression orientation. It might be that openly admitting to purposefully adopting such goal is not desirable or acceptable, and early adolescents may not perceive themselves as choosing such a goal though they may enact it in a predictive way. Further, we could only assess three general coherent orientations: prosocial, antisocial, and distress expression. The avoidance/withdrawal orientation did not appear coherent across the different SIP steps, although the goal clarification step was associated, as expected, with attachment avoidance. It is not clear whether this is specific to our sample or a general difficulty with the coherency of this orientation. Future research may need to examine the four SIP general orientations in different samples and different cultures and address this lacuna. Third, most social-information processing studies focused on at-risk populations (e.g., Nelson & Crick, 1999). The present study broadened our understanding of the significance of social-information processing in a nonclinical population and highlighted the importance of also examining prosocial social-information processing, which characterizes secure adolescents. Fourth, most social-information processing studies that focused on at-risk populations described the association between hostile attribution bias and aggressive SIP response (Crick & Dodge, 1994). These studies did not include the alternative of SIP distress expression orientation. The present study suggested the possibility that children’s hypervigilance to hostile cues, which advances a hostile attribution bias, would also be associated with distress expression SIP responses. This study, which replicated some of the previous findings regarding SIP in the United States and Germany, was conducted in the Israeli cultural context. The replication accords with a view that underscores the universal nature of social-information processing in social situations and their association with attachment representations. Nevertheless, such processes may be construed differently in diverse cultural contexts and future studies need to be sensitive to such possibilities. Several limitations of this study call for further research. First, the concurrent nature of the associations also accords with the possibility that socialization processes within the peer group affect the adolescent’s perception regarding his or her relationships with attachment figures (Bradford & Larson, 2009). Second, the concurrent association may also be related to a third factor, such as the adolescent’s temperament, which affects both contexts: peer and motherchild relations. In addition, some of the control variables were assessed following the major study variables; hence, studies that assess similar research questions using longitudinal designs are called for. Third, our study focused

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

561

on the mother-child attachment representations. Relationships with fathers may be another important context that affects peer relationships (Williams & Kelly, 2005). Furthermore, the use of a single informant for both attachment and SIP may increase the probability of shared method bias. Both issues might need to be addressed in future studies. Finally, in this study, girls scored a higher secure rating and boys scored a higher avoidant rating. A number of studies with school-age children (Barnett, Kidwell, & Leung, 1998; Park & Waters, 1989) indicated similar gender differences in the classification into attachment patterns. Research examining developmental processes related to socialization showed relevant gender differences, such as girls having a higher tendency to invest in, and foster, close and intimate relationships than boys (for a review, see Maccoby, 1998). Developmental processes related to socialization during early or middle childhood possibly operate and intervene to affect the quality of attachment of older children and adults of the two sexes differently. Future research may need to be sensitive to this possibility. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was partially supported by the Spencer Foundation.

References Arbuckle, J. L. (2006). AMOS (Version 7.0) [Computer software]. Chicago, IL: SPSS. Barnett, D., Kidwell, S. L., & Leung, K. H. (1998). Parenting and preschooler attachment among low-income urban African-American families. Child Development, 69, 1657-1671. Bauminger, N., & Kimhi-Kind, I. (2008). Social information processing, security of attachment and emotion regulation in children with learning disabilities. Journal of Learning Disabilities, 41, 315-332. Belsky, J., Spritz, B., & Crnic, K. (1996). Infant attachment security and affectivecognitive information processing at age 3. Psychological Science, 7, 111-114. Bowlby, J. (1982). Attachment and loss (Vol. 1): Attachment. New York, NY: Basic Books. Original published in 1969. Bradford, B. B., & Larson, J. (2009). Peer relationships in adolescence. In R. M. Lerner & L. Steinberg (Eds.), Handbook of adolescent psychology (Vol. 2): Contextual influences on adolescent development (3rd ed., pp. 74-103). Hoboken, NJ: Wiley.

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

562

Journal of Early Adolescence 32(4)

Bretherton, I., & Munholland, K. A. (2008). Internal working models in attachment relationships: Elaborating a central construct in attachment theory. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research and clinical applications (2nd ed., pp. 122-127). New York, NY: Guilford. Bretherton, I., Ridgeway, D., & Cassidy, J. (1990) Assessing internal working models of the attachment relationship: An attachment story completion task for 3-yearolds. In M. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschool years: Theory, research, and intervention (pp. 273-308). Chicago, IL: University of Chicago Press. Bryan, T., Burstein, K., & Ergul, C. (2004). The social-emotional side of learning disabilities: A science-based presentation of the state of the art. Learning Disability Quarterly, 27, 45-51. Bukowski, W. M., Newcomb, A. F., & Hartup, W. W. (Eds.). (1996). The company they keep: Friendship in childhood and adolescence. Cambridge studies in social and emotional development (pp. 262-285). New York, NY: Cambridge University Press Cassidy, J. (2008). The nature of the child’s ties. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research and clinical applications (2nd ed., pp. 3-22). New York, NY: Guilford. Cassidy, J., Kirsh, S. J., Scolton, K. L., & Parke, R. D. (1996). Attachment and representations of peer relationships. Developmental Psychology, 32, 892-904. Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information-processing mechanisms in children’s social adjustment. Psychological Bulletin, 115, 74-101. Crittenden, P. M. (1992). Quality of attachment in the preschool years. Development and Psychopathology, 4, 209-241. Dodge, K. A. (1993). Social-cognitive mechanisms in the development of conduct disorder and depression. Annual Review of Psychology, 44,559-584. Dodge, K. A., & Price, J. M. (1994). On the relation between social information processing and socially competent behavior in early school-aged children. Child Development, 65, 1385-1397. Erdley, C. A., & Asher, S. R. (1996). Children’s social goals and self-efficacy perceptions as influences on their responses to ambiguous provocation. Child Development, 67, 1329-1344. Finnegan, R. A., Hodges, E., & Perry, D. G. (1996). Preoccupied and Avoidant coping during middle childhood. Child Development, 67, 1318-1328. Furman, W. (2001) Working models of friendships. Journal of Social and Personal Relationships, 18, 583-602. Furman, W., & Simon, V. A. (1999). Cognitive representations of adolescent romantic relationships. In W. Furman, B. B. Brown, & C. Feiring (Eds.), The development of romantic relationships (pp. 74-98). Cambridge, UK: Cambridge University Press. Granot, D., & Mayseless, O (2001). Attachment security and adjustment to school in middle childhood. International Journal of Behavioral Development, 25, 530-541.

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

Granot and Mayseless

563

Hu, L., & Bentler, P. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55. Kerns, K. A. (1996). Individual differences in friendship quality: Links to childmother attachment. In W. M. Bukowski, A. F. Newcomb, & W. W. Hartup (Eds.), The company they keep: Friendship in childhood and adolescence (pp. 137-157) New York, NY: Cambridge University Press. Kerns, K. A. (2008). Attachment in middle childhood. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research and clinical applications (2nd ed., pp. 366-382). New York, NY: Guilford. Kerns, K. A., Klepac, L., & Cole, A. (1996). Peer relationships and preadolescents’ perceptions of security in the child-mother relationship. Developmental Psychology, 32, 457-466. Kerns, K. A., Schlegelmich, A., Morgan, T. A., & Abraham, M. M. (2005). Assessing attachment in middle childhood. In K. A. Kerns & R. A. Richardson (Eds.), Attachment in middle childhood (pp 46-71), New York, NY: Guilford. Lyons-Ruth, K., Bronfman, E., & Atwood, G. (1999). A relational diathesis model of hostile-helpless states of mind. In J. Solomon & C. George (Eds.), Attachment disorganization (pp. 33-69). New York, NY: Guilford. Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge, MA: Harvard University Press. Mayseless, O. (2005). Ontogeny of attachment in middle childhood: Conceptualization of normative changes. In: K. A. Kerns & R. A. Richardson (Eds.), Attachment in middle childhood. New York, NY: Guilford. Mikulincer, M., & Shaver, P. R. (2008). Adult attachment and emotion regulation. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research and clinical applications (2nd ed., pp. 503-531). New York, NY: Guilford. Nelson, D. A., & Crick, N. R. (1999). Rose-colored glasses: Examining the social information-processing of pro-social young adolescents. Journal of Early Adolescence, 19, 17-38. Overall, C. O., Fletcher, G. J. O., & Friesen, M. D. (2003). Mapping the intimate relationships mind: Comparisons between three models of attachment representations. Personality and Social Psychology Bulletin, 20, 1479-1493. Park, K. A., & Waters, E. (1989). Security of attachment and preschool friendships. Child Development, 60, 1076-1081. Parke, R. D., & O’Neil, R. (1999). Social relationships across contexts: Family-Peer linkages. In W. A. Collins & B. Laursen (Eds.), Relationships as developmental contexts: The 30th Minnesota Symposium on Child Psychology (pp. 211-240). Hillsdale, NJ: Erlbaum. Quiggle, N., Garber, J., Panak, W., & Dodge, K. A. (1992). Social-information processing in aggressive and depressed children. Child Development, 63, 1305-1320.

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015

564

Journal of Early Adolescence 32(4)

Raikes, H. A. & Thompson, R. A. (2005). Relationships past, present, and future: Reflections on attachment in middle childhood. In K. A. Kerns & R. A. Richardson (Eds.), Attachment in middle childhood (pp 255-283). New York, NY: Guilford. Schermelleh-Engel, K., Moosbrugger, H., & Muller, H. (2003). Evaluating the fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of Psychological Research Online, 8, 23-74. Schneider, B. H., Atkinson, L., & Tardif, C. (2001). Child-parent attachment and children’s peer relations: A quantitative review. Developmental Psychology, 37, 86-100. Suess, G. J., Grossmann, K. E., & Sroufe, L. A. (1992). Effects of infant attachment to mother and father on quality of adaptation in preschool: From dyadic to individual organization of self. International Journal of Behavioral Development, 15, 43-65. Ullman, J. B. (2001). Structural equation modeling. In B. G. Tabachnick & L. S. Fidell (Eds.), Using multivariate statistics (4th ed; pp 653-771). Needham Heights, MA: Allyn & Bacon. Van Ijzendoorn, M. H., Vereijken, C. M. J. L., Bakermans, K. M. J., & Riksen, W. J. M. (2004). Assessing attachment security with the attachment Q Sort: Meta-analytic evidence for the validity of the observer AQS. Child Development, 75, 1188-1213. Von Salisch, M. (2001). Children’s emotional development: Challenges in their relationships to parents, peers, and friends. International Journal of Behavioral Development, 25, 310-319. Williams, S. K., & Kelly, F. D. (2005). Relationships among involvement, attachment and behavioral problems in adolescence: Examining father’s influence. Journal of Early Adolescence, 25, 168-196. Yair, G. (1991). The deprivation index: Paradoxes and empirical assessment, policy implications. Megamot, 34, 5-26. (in Hebrew) Ziv, Y., Oppenheim D., & Sagi-Schwartz, A. (2004). Social information processing in middle childhood: Relations to infant-mother attachment. Attachment & Human Development, 6, 327-348.

Bios David Granot is the head of the Special Education Department at Oranim, the Academic College of Education, Israel. His current research and practice focuses on the psychosocial functioning of children with learning disabilities and on teacherstudent relations using an attachment perspective. Ofra Mayseless is dean of the Faculty of Education at the University of Haifa, Israel. She is a professor of developmental psychology. Her current research focuses on the care-giving/nurturing motivational system as it is manifested in leadership, parenting, teaching, and role reversal, as well as in how it relates to the spiritual realm and spiritual development.

Downloaded from jea.sagepub.com at University of Haifa Library on November 18, 2015