The Integrative Model of Multiple Attachment Relationships in Adolescence: Linkages to Bullying and Victimization

Abstract

The present study sought to examine the independent, hierarchical, and integrative models of multiple attachment relationships in a sample of Greek-Cypriots in middle adolescence and to test the distinct and interactive effects of these relationships on bullying involvement. A sample of Greek-Cypriot adolescents (N = 406, 55.4% females, M_age = 16.01, SD = 0.82) completed the Greek forms of the Revised Olweus Bully-Victim Questionnaire (BVQ-R), the Inventory of Parental and Peer Attachment (IPPA), and the Questionnaire on Teacher Interaction (QTI). Path models were used to test for the effect of multiple attachment relationships on bullying and victimization, using AMOS 24.0.The integrative model of multiple attachment relationships was supported and found to have significant effects on bullying involvement. The effect of the quality of the mother-adolescent relationship on bullying involvement was fully mediated by the quality of the attachment relationships with peers and the interpersonal relationship with teachers. The quality of adolescents’ relationship with teachers affected their interactions with peers, and not the opposite, though both continued to have distinct effects on bullying involvement.

Keywords

Multiple attachment relationships bullying adolescence teacher-student-interaction

Attachment represents a special relationship which involves an affective bond between the infant and the caregiver that can be characterized in terms of emotion regulation (Jacobite & Hazen, 1999). Early attachment relationships with the caregiver provide an internal working model from which the child filters aspects of the social world, and thus form the basis for all future close relationships through childhood and adolescence into later life, in the form of schemes of multiple relationships (Armsden & Greenberg, 1987; Bowlbly, 1969). Research has proposed three major schemes regarding the way the initial infant-caregiver relationship affects future close relationships: The independent, the hierarchical, and the integrative model (Overall, Fletcher, & Friesen, 2003; Raikes & Thomson, 2005). However, whereas adolescence signals a dramatic shift in the social needs of children as they now seek autonomy from their caregivers (Hong & Espelage, 2012; Nickerson & Nagle, 2005), empirical evidence regarding the scheme under which parental attachment influences other attachment relationships (such as those with their peers and their teacher) in adolescence are scarce.

Bullying defined as an intentional and repetitive form of aggression, characterized by an imbalance of power between the victim and the perpetrator (Olweus, 1993), poses a major threat to adolescents as it is associated with detrimental emotional and behavioral outcomes (Haltigan & Vaillancourt, 2014; Menesini, Modena, & Tani, 2009; O’Brennan, Bradshaw, & Sawyer 2009). Several attempts have been made to better understand bullying involvement, including the investigation of the association between bullying and the attachment relationship framework (Marini, Dane, Bosacki, & Cura, 2006; Murphy, Laible, & Augustine, 2017), in line with the conceptualization of bullying behavior as a dysfunctional way of interpersonal interaction, resonated in attachment theory (e.g., Eliot & Cornell, 2009). However, to the present no study has comprehensively examined the distinct and interactive effects of multiple attachment relationships on bullying involvement. Delineating the scheme of multiple attachment relationships in adolescents and clarifying how multiple close relationships relate to students’ engagement in bullying would provide significant information in directing bullying prevention programs (e.g., Nation, Vieno, Perkins, & Santinello, 2008; Walden & Beran, 2010).

The present study sought to examine the independent, hierarchical and integrative models of multiple attachment relationships (parental and peer attachment, and teachers’ interpersonal behavior) in a sample of Greek-Cypriots middle adolescents and to test the distinct and interactive effects of these relationships on bullying involvement.

Attachment Theory

The attachment relationship signals a special relationship involving an affective bond between the infant and the caregiver. This relationship can be characterised in terms of emotion regulation in infants and can be conceptualised in the form of a continuum of emotional regulation for managing affect, events, and relationships (Jacobite & Hazen, 1999). At the one end of the continuum lies the anxious–avoidant attachment relationship, for which over-emphasis is placed on controlling and minimising affect. At the opposite end lies the anxious–resistant style, characterised by relatively uncontrolled, poorly-managed affect. Secure attachment represents the equilibrium between the two extremes of emotional regulation.

Early attachment relationships with the caregiver form the basis for all future close relationships through childhood and adolescence into later life, and they provide an internal working model from which the child filters aspects of the social world (Armsden & Greenberg, 1987; Bowlby, 1969). As the child develops, additional social experiences interfere with this working model, with research supporting a relative stability of the child–caregiver relationship over time (Bretherton & Munholland, 2008; Fraley, 2002). It should be noted that attachment is usually assessed in infancy with the Ainsworth Strange Situation procedure (Ainsworth, 1969), whereas questionnaires and interviews are used with adolescent populations (Crowell, Fraley, & Shaver, 1999).

Three Proposed Schemes for the Representations of Multiple Attachment Relationships

Three major schemes have been proposed regarding the conceptualization of the parental attachment as an internal working model which defines future close relationships (Overall et al., 2003; Raikes & Thomson, 2005): 1. The hierarchical scheme suggests that the initial attachment relationship between the child and the primary caregiver is the predominant relationship that determines the quality and nature of the future attachment relationships; 2. the independent scheme postulates that future close relationships are independent of the initial attachment relationship; and, 3. the integrative scheme suggests that children incorporate aspects of their relationships with the primary caregiver and adopt similar roles in their future attachment relationships.

Previous studies have supported the integrative scheme for multiple attachment relationships with respect to preadolescents (Charalampous et al., 2016) and adults (Overall et al., 2003). Yet, no study has specifically examined these schemes in middle and late adolescence. Given that adolescence is a developmental period of considerable developmental changes including a shift in the social needs of children, which begin to seek autonomy from their caregivers (Hong & Espelage, 2012; Nickerson & Nagle, 2005) and that adolescents report more positive perceptions of their relationships with peers and less positive perceptions of their relationships with adults than children (Lynch, & Cicchetti, 1997), it is important to understand the role of initial caregiver attachment relationships in the formation of future close relationships (e.g. with peers or teachers) at this developmental stage.

Parental-Peer Attachment, Parental-Teacher Relationship, Peer-Teacher Relationship

Previous studies have highlighted the significant effect of early parent-child relationships on the quality of subsequent relationships with peers (Schneider et al., 2001). In a recent meta-analysis, a significant yet small to moderate effect was found between parental attachment security and close peer relationships (r = 0.19) (Pallini, Baiocco, Schneider, Madigan, & Atkinson, 2014). The authors analyzed data from 44 studies and a total sample of 8.505 participants, underscoring a number of possible moderators of the effect. However, the authors used a mean effect size in case the study reported multiple outcomes (e.g., number of friendships, quality of peer relations), which might have an impact on the mean correlation between parental security and peer relationships.

A number of studies report on the similarities between the parent-child and the teacher-child relationship, suggesting the existence of an ‘attachment component’ in the teacher-child relationships (Thompson, 2008). Vershueren and Koomen (2012) also highlight the importance of teachers as persons whose sensitivity and responsiveness can determine whether children would perceive them as a secure base. Rydell, Bohlin, and Thorell (2005) examined the interactions between multiple attachment relationships, by investigating the effect of parent-child attachment representation of preschool children on their relationships with teachers. Their findings showed that secure parental attachment relationships was related to a more positive relationship to teachers. In addition, children with avoidant parental attachment showed more conflict, lower closeness, and lower prosocial behaviour in their relationships with teachers compared to their peers with secure parental attachment, whereas children with ambivalent attachment patterns showed low levels of security with their teachers. Verschueren, Doumen, and Buyse (2012) also reported a positive association between attachment security in the mother-child relationship at preschool and the quality of the teacher-child relationship at the first grade.

Few studies have addressed the relationship between student-teacher and student-peer relationships. Security in the relationship between children and their teachers at age two could predict competent peer interactions at age four, as characterized by complex play, prosocial behaviour and socialization, low hostile aggression, and low withdrawn behaviour (Howes, Matheson, & Hamilton, 1994).

However, most studies that have been conducted so far related to student-teacher relationships investigated preschool children and thus, there is a lack of studies using samples from other age groups (e.g. adolescence). At the same time, as the literature supports the mutual interactions between multiple attachment relationships, the scarcity of studies concurrently testing these interactions and their effects on adolescent behaviour, confines our effort to understand maladaptive behaviours in this context. One such maladaptive behaviour, attracting much attention in recent adolescent literature, is school bullying.

Bullying Behaviour, Parental Attachment, and Peer Attachment

Bullying is a form of aggression which is intentional, repetitive over time, and it is characterized by power imbalance between the perpetrator/s and the victim/-s (Olweus, 1993). A broad list of individual and contextual factors have been linked and hypothesized to explain the childrens’ or adolescents’ involvement in bullying behaviour. However, limited research has focused on the relationship between parental attachment and bullying involvement and even less has investigated the role of peer attachment and the interplay between parental attachment and bullying involvement (Murphy et al., 2017; Nikiforou, Georgiou, & Stavrinides, 2013). As attachment relationships form the main interaction prototypes for children and adolescents, insecure forms of attachment have an impact on social information processing and emotion regulation systems (Jacobite & Hazen, 1999). Emotion regulation systems are central in the development and/or maintenance of bullying behaviours, because of their impact in the social interaction of the individual. Specific characteristics of parental attachment have been related to specific bullying or bullying-related behaviors. Specifically, high attachment avoidance and anxiety in the relationship with the father has been related to higher relational aggression in adolescents (Williams, 2011). At the same time, alienation from mother (a characteristic of insecure attachment relationships) and trust were detected to be higher and lower respectively, between adolescents involved in bullying compared to their uninvolved peers (Marini et al., 2006). A recent meta-analysis on studies investigating the parental factors which are related to bullying involvement, found that high parental communication, warmth and affection in the attachment relationship were among the major factors that were likely to protect children and adolescents from victimization (Lereya, Samara, & Wolke, 2013).

Peer attachments may function comparably to parental attachments and seem to have similar effects on adolescents’ outcomes, including their involvement in bullying behaviour (Murphy et al., 2017; Nikiforou et al., 2013). Laible, Carlo, and Raffaelli (2000) have proposed a link between secure peer attachments and the development of empathy, social information processing, and perspective taking, suggesting that secure attachment with peers should enhance social development and should prevent them from bullying involvement. Indeed, higher security in peer attachments have been reported by uninvolved middle adolescents from grades six to eight compared to adolescents involved in bullying (Burton, Florell, & Wygant, 2013). Demanet and van Houtte (2012) found that 9th and 11th graders tended to perceive higher security in their peer attachment relationships if they reported higher school belonging and higher perceived teacher support. Even though peer attachment was associated with more school misconduct in that study, suggesting the generalization of deviance between peers with high attachment to each other, this effect stopped being significant after controlling for teacher support and school belonging, which had a preventive effect on school misconduct (Demanet & van Houtte, 2012). It should be noted, however, that the measurement tools used in this study were not well established (e.g., parental attachment was measured with a 7-items questionnaire and school misconduct with a 17-items scale that were developed for the scope of the study).

According to Murphy and colleagues (2017) the interplay between parental and peer attachments is important, as previous research has demonstrated that when adolescents have a shortcoming in one of the two relationships, then the other one can act as a protective factor. For example, it has been found that high peer acceptance could protect adolescents with high parental rejection from the development of internalizing and externalizing problems (Sentse, Lindenberg, Omvlee, Ormel, & Veenstra, 2010). This may indicate that adolescents use peer attachment relationships as a way to compensate if they have insecure attachment relationships to their parents. Murphy et al. (2017) investigated the relationship between parental attachment, peer attachment and bullying in a sample of middle adolescents. Their findings supported the protective role of parental and peer attachment security in bullying involvement. In addition they found that for males, greater peer attachment security predicted less bullying involvement for those with lower parent attachment security, but not for those with higher parent attachment security. The latter finding suggests that having a secure attachment to peers may be a potential protective factor against bullying involvement for males with insecure attachments to parents.

Teacher’s Interpersonal Behaviour and Bullying Involvement

Based on the communicative systems approach, teachers’ interpersonal behavior affects their pupils’ social behaviors in two ways (Wubbels, Opdenakker, & Den Brok, 2012): 1. Students might respond to teacher behavior in a symmetrical way and thus model teacher behaviors. Hence, if a teacher displays proximity behaviors to their students, students might also show caring and respectful patterns of interaction with their teachers and peers (Gest & Rodkin, 2011); 2. Students can behave complementarily towards their teachers’ interpersonal behaviors by showing behaviors that are opposite on the agency dimension and identical on the communion dimension (Wubbels et al., 2012). For example, students who have teachers with an interpersonal behavior characterized by high control and directivity are likely to behave in a conforming manner (Thijs, Koomen, Roorda, & ten Hagen, 2011). Research supports that the student-teacher relationship is not only affected by maternal attachment relationship, but that it also mediates the effect of maternal attachment on internalizing and externalizing behaviors (O’Connor, Collins, & Supplee, 2012).

Previous research has shown that pupils of primary schools who perceive their teacher as supportive and sensitive to their needs show more prosocial behavior, less school misbehavior, are less likely to reject aggressive and withdrawn peers, and have more reciprocal friendships (Gest & Rodkin, 2011). In this respect research has shown that higher teacher agency and communion relate to lower bullying and victimization levels (Richard, Schneider, & Mallet, 2011; van der Zanden, Denessen, & Scholte, 2015). In line with the principles of complementarity and symmetry, researchers assumed that teachers who are in control and demonstrate care towards pupils may inhibit bullying by creating a sense of connectedness in the classroom in which pupils care about each other (Allen, 2010; Orpinas & Horne, 2006).

Previous studies have underscored the distinct importance of attachment relationships with main figures (i.e., parents, peers, and teachers) in the prediction of bullying involvement. However, research concerning the nature of the interactions between multiple attachment relationships and bullying, whether hierarchical or integrative, is still inconclusive. As children and adolescents function simultaneously in more than one social environment and their attachment style may constantly change based on their interactions, research needs to provide answers on the additive or independent effects of multiple attachment relationships on adolescent dysfunctional behavior. Under this framework, prevention and intervention approaches would emphasize on specific developmental phases that would allow the in-depth processing of attachment relationships and related mechanisms developed (e.g., social information, emotion regulation).

The Present Study

The present study sought to examine the independent, hierarchical, and integrative models of multiple attachment relationships (parental and peer attachment, and teachers’ interpersonal behavior) in a sample of Greek-Cypriot middle to late adolescents and to test the distinct and interactive effects of these relationships on bullying involvement.

The integrative scheme regarding multiple attachment relationships has been previously supported with respect to preadolescents (Charalampous et al., 2016) and adults (Overall et al., 2003). Yet, given the developmental changes in adolescence, including parental autonomy seeking (Hong & Espelage, 2002) and changes in the perceptions of the relationships with adults and peers (Lynch & Cicchetti, 1997), it is difficult to hypothesize which scheme of multiple attachment relationship would apply to these age group, even though previous findings favor the interactive model (Murphy et al., 2017; Verschueren et al., 2012).

Previous studies have supported that poor attachment relationship quality with parents and peers significantly predicts the involvement in bullying and victimization (Murphy et al., 2017; Nikiforou et al., 2013). At the same time research supports that a positive teacher relationship, characterized by high agency and communion, may serve as a protective factor against bullying involvement (e.g., van der Zanden et al., 2015). The present study is one of the first to examine the distinct and interactive contribution of these variables in bullying perpetration and victimization. It is expected that all three parental, peer and teacher relationships will relate to bullying involvement, and that at least some partial mediation effects will be present.

In order to test the proposed schemes of multiple attachment relationships, several conditions were set (see Fig. 1): Given that the independent scheme posits that future close relationships are independent from the maternal attachment relationship, support to this scheme would be provided if maternal attachment had no effect on peer attachment or on teacher interaction, and if all close relationships with attachment figures have simultaneous unmediated effects on bullying involvement. In the hierarchical scheme the maternal relationship is expected to have a significant effect on future close relationships, and the potential effect of maternal attachment on behavioral outcomes is expected to be channeled through this future close relationships. In addition, the presence of maternal attachment relationship is expected to overshadow any associations between future close relationships. Thus, the hierarchical scheme would be supported if the effect of parental attachment on bullying is mediated by peer attachment relationships or by the teacher’s interaction and if no effects are in place between peer attachment and teacher’s interaction. Finally, in the integrative scheme the maternal relationship is also expected to have a significant effect on future close relationships (albeit of lower magnitude compared to the hierarchical scheme), and the potential effect of maternal attachment on behavioral outcomes is also expected to be channeled through this future close relationships. However, in this case, interactions between future close relationships, even in the presence of maternal attachment, are expected. Thus, the integrative model would be supported if the effect of parental attachment on bullying is mediated by the peer attachment relationship or by teacher’s interaction, and if effects are in place between peer attachment and teacher’s interaction.

Figure 1

The statistical models representing the inter-relations of the variables of the study supporting each of the hypothesized schemes.

Methods

Participants and Procedure

Participants were 406 Greek-Cypriot students aged from 15 to 18 years (M = 16.01, SD = 0.82) attending secondary schools in Nicosia, Cyprus. Participants were randomly selected to participate in the study. A percentage of 55.4% of the sample were females and the rest were males. Among the participants 58.6% were attending the fourth grade and 41.4% the fifth grade.

The present study was part of a large-scale project investigating risk and protective factors for school bullying in middle adolescence in Cyprus. The participants completed a questionnaire package (average time for completion was 40 minutes), which included the measures used in the present study and others, and were not compensated for their participation. The study was approved by the Ministry of Education in Cyprus, and both parental and participant consents were obtained. The participants were invited to participate from six randomly selected schools in urban and rural areas in Nicosia, Cyprus. One of the randomly selected schools was a Technical and Occupational School and rest were High Schools. Therefore, the number of students participating in the study who came from High and Technical Schools respectively, was representative of the general population ratio of students attending high or technical schools in Cyprus. Trained members of the research team administered self-report questionnaires to the participants in school during school hours.

Instruments

Teacher-student interaction

The Questionnaire on Teacher Interaction (QTI) is a self-report measure consisting of 40 items, using a 5-point Likert scale, assessing the relationship interactions between a student and his/her teacher. The QTI has been adapted to Greek by Kokkinos, Charalambous, and Davazoglou (2009; Charalampous & Kokkinos, 2013, 2018) and was based on the version for elementary students that has been developed by Goh and Fraser (1996). Previous research using the QTI has showed adequate psychometric properties (e.g., Charalampous et al., 2016). The QTI is theoretically based on the Model of Interpersonal Teacher Behavior (MITB), which conceptualizes the interpersonal teacher behavior in two dimensions, agency and communion, and eight sectors, labeled: directing (sample item: “The teacher is a good leader.”), supporting (sample item: “The teacher is someone we can depend on.”), understanding (sample item: “If we have something to say this teacher will listen.”), acquiescing (sample item: “The teacher gives us a lot of free time in class.”, hesitating (sample item: “The teacher seems uncertain.”), objecting (sample item: “The teacher is suspicious.”, confronting (sample item: “The teacher gets ungry.”, and imposing (sample item: “The teacher is strict.”(Wubbels et al., 2012). These sectors represent eight types of teacher behavior placed in a circle according to their combination of interpersonal valence on the two dimensions of agency and communion. The internal consistency of the subscales ranged from α= 0.78 to 0.91.

Parental and peer attachment

The Inventory of Parent and Peer Attachment (IPPA; Armsden & Greenberg, 1987) is a self-report questionnaire with 75 questions (25 for each of the attachment figures: mother, father, and peers) assessing adolescents’ perceptions of their relationships to close others who serve as sources of security. Participants respond on a 5-point Likert scale (1 = full disagreement; 5 = full agreement). Each of the three scales has three subscales: trust (sample item: “When we discuss things, my mother cares about my point of view”), communication (sample item: “I tell my mother about my problems and troubles”), and alienation (sample item: “I don’t get much attention from my mother”). The IPPA is theoretically based on attachment theory (Bowlby, 1969, 1988) and has shown good psychometric properties in previous research done in Cyprus (e.g., Nikiforou et al., 2013). The internal consistency for the subscale of maternal attachment was: α= 0.76 (for trust α= 0.84, for communication α= 0.75, and for alienation α= 0.82); and for peer attachment α= 0.83 (for trust α= 0.87, for communication α= 0.79, and for alienation α= 0.87).

Bullying and victimization

The Bully/Victim Questionnaire-Revised (BVQ-R; Olweus, 1993, 1996, 1997) consists of 20 items to assess children’s bullying and victimization experiences (10 statements for each subscale). Answers are given on a 5-point Likert-type scale ranging from 1 (not true) to 5 (absolutely true). Sample items from the bullying subscale include, “Other children are afraid of me” and “Other children complain that I hit them.” Sample items from the victimization subscale include, “Other children have hit me or tried to hit me” and “Other children constantly tease me and call me names.” In earlier studies conducted with Cypriot populations (e.g., Georgiou, 2008; Georgiou, Fousiani, Michaelides, & Stavrinides, 2013; Kyriakides, Kaloyirou, & Lindsay, 2006; Stavrinides, Georgiou, Nikiforou, & Kiteri, 2011; Stavrinides, Georgiou, & Theofanous, 2010), the measurement tool yielded two factors representing “bullying” (perpetrator perspective) and “bullying victimization” and showed good psychometric properties (α values ranging from 0.70 to 0.80). Construct validity has been previously supported with a large sample of Greek-Cypriot pupils (Kyriakides et al., 2006). Cronbach’s alpha had a value of 0.76 and 0.83 for the bullying and the victimization subscales respectively.

Statistical analyses

The psychometric properties of the questionnaires were examined through the use of exploratory and confirmatory factor analyses and structural equation models were applied for testing the research questions. The maternal attachment variables served as the independent variables (exogenous) for the examined models and the peer and teacher relationships as mediators of the relationship between maternal attachment and bullying/victimization (dependent variables, endogenous). It was decided to keep attachment parameters separately, firstly due to the distinctiveness of the constructs as confirmed by the exploratory factor analyses on the maternal and peer subscales of the IPPA and secondly due to a lack of such analyses in previous studies. Even though the conceptualization and measurement of attachment as a single construct might provide more specific findings, it may also blur the distinct effects of attachment dimensions on the examined constructs.

In order to examine mediation effects, the direct effects between the variables were tested first. Then, the mediational variables were inserted to the models and the effects of both the mediating and the exogenous factors on the endogenous factors were estimated. Finally, the indirect effects were tested by constraining the direct effects to zero. Full, partial, or no mediation conclusions are established based on the improvement in model fit and on the remaining significant effects at each step.

The maximum likelihood estimation was used and analyses were performed using AMOS (Arbuckle, 2006). Several indices were examined to assess model fit including the Chi-squared goodness of fit statistic (χ²), the Comparative Fit Index (CFI), the Tucker Lewis index (TLI), and the Root Mean Squared Error of Approximation (RMSEA) (Browne & Cudeck, 1989; Byrne, 2009; Hu & Bentler, 1999; Marsh, Wen, & Hau, 2004; Smith & McMillan, 2001). Adequate fit is indicated by non-significance for the χ². For the CFI and TLI values over 0.90 indicate adequate and over 0.95 excellent fit. For the RMSEA values less than 0.08 indicate adequate fit and less than 0.05 excellent fit.

Results

The psychometric properties and factorial structure of all questionnaires were confirmed, using Principal Axis Factoring in EFAs run in SPSS 24.0. The correlations of the mean scores between the studied constructs are presented in (Table 1). Associations were in the expected direction. Positive relationship indices with the mother, peers, and the teacher were positively inter-related and correlated negatively with bullying involvement. On the other hand, teacher agency had no significant correlations with any of the variables of the study, and thus, was not expected to play any significant role in the upcoming models.

Table 1

Bivariate Correlations

Variables	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.
1. Peer communication
2. Peer trust	0.83^***
3. Peer alienation	–0.13 *	–0.31^***
4. Maternal alienation	–0.17^***	–0.18^***	0.33^***
5. Maternal trust	0.26^***	0.27^***	–0.14 **	–0.54^***
6. Maternal communication	0.41^***	0.35^***	–0.14 **	–0.55^***	0.81^***
7. Teacher agency	0.06	0.09	–0.06	0.01	0.09	0.05
8. Teacher communion	0.19^***	0.19^***	–0.11 *	–0.24^***	0.31^***	0.26^***	–0.09
9. Victimization	–0.35^***	–0.44^***	0.28^***	0.16 **	–0.27^***	–0.29^***	0.02	–0.20^***
10. Bullying	–0.26^***	–0.27^***	0.12 *	0.18* **	–0.18 **	–0.16 **	–0.01	–0.21^***	0.37^***

^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

Next, the latent constructs were created and confirmatory factor analyses were tested in AMOS 24.0. All of the CFAs showed adequate model fit indices. However, due to the complexity of the hypothesized models, only the bullying and victimization constructs were used as latent variables, whereas the rest were introduced in the models as observed variables representing the factor scores of each subscale (QTI and IPPA). Therefore, factor scores were used as observed variables in the path models for maternal communication, trust and alienation, for peer communication, trust and alienation, and for teacher agency and communion. Directing, Imposing, Confronting and Objecting interpersonal behavior with the teacher formed the teacher agency factor, and Supporting, Understanding, Acquiescing, Directing and Objecting behavior formed the teacher communion factor, based on previous research using the QTI and the evidence for the circumplex positioning of the eight subscales (e.g., Charalampous et al., 2016; Den Brok et al., 2003).

We proceeded with the examination of the integrative and hierarchical models for multiple attachment relationships. In the first model (Model 1a) maternal trust, communication, and alienation were allowed to independently predict peer trust, communication and alienation, and teacher agency and communion. In Model 1b, in addition to the relationships in Model 1a, teacher communion and agency were predicting peer attachment dimensions, and in Model 1c, the effect of teacher communion and agency on peer attachment dimensions was reversed. The model fit indices are presented in (Table 2).

Table 2

Model Fit Indices and Model Comparisons

	χ ²	df	p	CFI	TLI	RMSEA (90% CI)	SQMC (b/v)	Δχ² (Δdf), p
Model 1a	17.337	7	0.015	0.99	0.97	0.060 (0.025–0.097)
Model 1b	5.969	1	0.015	1.00	0.89	0.111 (0.039–0.203)		11.368 (6), p < 0.10
Model 1c	7.138	1	0.008	1.00	0.87	0.123 (0.051–0.214)		1.169 (0)
Model 2a	471.301	193	0.000	0.91	0.90	0.060 (0.053–0.067)	0.057/0.085
Model 2b	663.232	285	0.000	0.91	0.88	0.057 (0.052–0.063)	0.141/0.249
Model 2c	651.864	279	0.000	0.91	0.88	0.057 (0.052–0.063)	0.150/0.255	11.368 (6), p < 0.10
Model 2d	653.032	279	0.000	0.91	0.88	0.058 (0.052–0.063)	0.149/0.254	1.168 (0)

SQMC; squared multiple correlations (variance explained by the model for bullying/victimization).

Results from Model 1a indicated that maternal trust had a significant positive effect on teacher communion (b = 0.24, p < 0.001) and a negative effect on peer communication (b = –0.18, p = 0.02). Maternal communication positively predicted peer communication (b = 0.57, p < 0.001) and peer trust (b = 0.39, p < 0.001). At the same time, maternal alienation was positively related to peer alienation (b = 0.37, p < 0.001). Model 1b showed that the addition of effects from teacher’s interaction dimensions resulted in an increase in model fit (at the p < 0.01 level). Significant positive effects were found from teacher communion on peer trust (b = 0.12, p = 0.01) and on peer communication (b = 0.13, p = 0.008). On the other hand, the examination of the reversed effects from peer attachment quality on interpersonal relationships with teachers (Model 1c) showed that these effects were not significant. Importantly, the effects from the maternal attachment quality parameters (maternal trust, communication, and alienation) on peer attachment dimensions and on teacher interaction dimensions did not change in the three examined models. Overall, these results show that the independent scheme of multiple attachment relationships is not a good explanation to the data.

Next, we proceeded with the addition of bullying involvement. First, we examined a model with independent effects from each of the attachment figures on bullying. Results were consistent in their depiction of an ill-fitting model [χ² (299) = 840.90, p < 0.001, CFI = 0.87, TLI = 0.84, RMSEA = 0.067 (90% CI 0.062–0.072)].

We then tested the direct effects from mother attachment parameters to bullying involvement (Model 2a – Fig. 2) and subsequently examined Models 2b (Fig. 3), 2c (Fig. 4) and 2d, which were similar to Models 1a–1c, respectively, with the addition of bullying and victimization as endogenous variables, in order to examine the multiple attachment relationships’ effects on bullying and victimization. Model 2a showed that direct effects existed from maternal trust on bullying (b = –0.18, p < 0.001) and from maternal communication on victimization (b = –0.21, p < 0.001). Maternal trust and communication explained 6% and 9% of the total bullying and victimization variance, respectively. The addition of the teacher interaction dimensions and the peer attachment indicators and their independent effects on bullying and victimization (Model 2b) led to non-significant effects from maternal trust and communication on bullying involvement. Specifically, less bullying involvement was now predicted by higher peer trust (b = –0.23, p = 0.021) and by higher teacher communion (b = –0.16, p = 0.003). Victimization was predicted by low peer trust (b = –0.37, p < 0.001), high peer alienation (b = 0.18, p = 0.002), and by low teacher communion (b = –0.14, p = 0.009). These findings were retained even when we tested only the indirect effects by fixing the direct effects from maternal attachment to bullying involvement to zero.

Figure 2

Model 2a based on the independent scheme of multiple attachment relationships. Values reported are standardized and significant (p < 0.05). Dashed arrows represent relationships that were tested but found statistically non-significant. Arrows above the endogenous variables represent squared multiple correlations.

Figure 3

Model 2b based on the hierarchical scheme of multiple attachment relationships. Values reported are standardized and significant (p < 0.05). Dashed arrows represent relationships that were tested but found statistically non-significant. Arrows above the endogenous variables represent squared multiple correlations.

Figure 4

Model 2c based on the integrative scheme of multiple attachment relationships. Values reported are standardized and significant (p < 0.05). Dashed arrows represent relationships that were tested but found statistically non-significant. Arrows above the endogenous variables represent squared multiple correlations.

The examination of the association of the teacher interaction and the peer attachment (Models 2c and 2d) showed that significant effects existed only from the teacher interaction to peer attachment and not the opposite, replicating the findings from Models 1b and 1c. Importantly, the direct effects from teacher communion on bullying and victimization involvement were retained, even when peer attachment quality was considered a potential mediator of this relationship (Model 2c) and apart from the significant effects of teacher communion on peer trust and communication. Overall, Model 2c was found to be a significantly better fit to the data than Model 2b at the p < 0.10 level, and explained a higher percentage of the variance of the latent endogenous variables (bullying and victimization). That is, Model 2c explained 26% and 15% of the victimization and the bullying behaviors, respectively.

Discussion

The present study sought to examine the independent, hierarchical, and integrative models of multiple attachment relationships in a sample of Greek-Cypriot middle adolescents and to test the distinct and interactive effects of these relationships on bullying involvement. The findings provide support for the integrative model of multiple attachment relationships in middle adolescence and some first indications in how multiple attachment relationships in adolescence might relate to bullying perpetration and victimization.

The integrative model was supported by the fact that maternal attachment had a significant effect on peer attachment and teacher interaction (rejection of the independent scheme) and at the same time an effect from teacher-student interaction onto peer attachments was also found (rejection of the hierarchical scheme). The effect from teacher communion on peer trust and peer communication indicates that the relationship with the teacher, which is partially influenced by the maternal attachment, can additively impact the peer attachment relationship. This finding is in line with the integrative model of multiple attachment relationships, which implies that the attachment relationship with the primary caregiver (in the case of the present study the mother) influences future close relationships (e.g. peer relationships), yet other close relationships (e.g. with the teacher) may also have an additive effect on future relationships. The fact that the student-teacher interaction was a significant predictor of peer attachment quality parameters and not the opposite seems to be in line with previous findings (Charalampous et al., 2016) and adds to the literature in that higher security in peer relationships is perceived, if higher perceived support by teachers is preceded (Demanet & van Houtte, 2011). These findings are in accordance with previous studies examining the multiple attachment relationships in early adolescents (Charalampous et al., 2016).

In addition, the effect of maternal attachment on bullying involvement was mediated by peer attachment relationships and the student-teacher interaction. It thus seems that the effect of the initial attachment figure on future behavior seems to be channeled through future close relationships. This finding, along with the association between subsequent close relationships (peer and teacher relationships), again provides clear support for the integrative model of multiple attachment relationships and highlights the importance of accounting for multiple close relationships when examining bullying involvement (Nation et al., 2008; Walden & Beran, 2010).

The findings of the present study also showed that maternal trust had a negative effect on peer communication. Even though this effect was of small magnitude, it might reflect the fact that adolescents who perceive having highly secure and trust-worthy maternal relationships may feel less need for increased communication with their peers (as they may share their personal concerns with mothers). However, one should be cautious on the interpretation of this finding, since its replicability should be first established in future studies.

The findings regarding the effects of maternal attachment quality on bullying involvement are generally consistent with the literature, though a direct significant effect from alienation with mother to increased bullying behaviour was not replicated in the present study (e.g., Marini et al., 2006). In addition, an important finding of the present study was the full mediation of the effect of maternal attachment quality on bullying involvement, when the distinct effects of peer attachment and teacher interaction were concurrently examined. Previous concurrent investigation of parental and peer attachment as predictors of bullying roles found that attachment with parents was the only significant predictor of the two (Murphy et al., 2017). However, in the present study, peer trust and peer alienation were shown to be the most significant predictors of involvement in bullying and victimization, in accordance with other researchers (e.g., Charalampous et al., 2018; Laible et al., 2000) indicating the importance of secure peer attachments for developing empathy and high emotion regulation skills that could serve as protective factors for bullying involvement.

The findings of the present study are also in partial accordance with the findings by van der Zanden and colleagues (2015) and by Richard and colleagues (2011), in terms of the positive impact of high communion in the relationship with teachers on bullying and victimization. However, the effect of teacher agency on bullying involvement was not replicated. This may be a result of cultural or developmental differences. For example, some authors (Charalampous & Kokkinos, 2013, 2018; Flanagan & Stout, 2010 Lempers & Clark-Lempers, 1992) have shown differences in the way early, middle, and late adolescents perceived their relationships with significant others, including teachers, and differences in terms of social trust. In addition, other authors have supported that students with differing cultural backgrounds may interpret their teachers’ agency in a different manner (Evans & Fisher, 2000; Koul & Fisher, 2005)

Limitations

The cross-sectional research design is a limitation of the present study. If longitudinal data had been gathered, inferences on causal relationships between the variables under study would be more easily drawn. A recent meta-analysis (Lereya et al., 2013) concluded that the cross-sectional design is the main problem for studies (met for 62 out of a total of 70 studies included in the review) investigating risk and protective factors for bullying. Even though the sample was randomly selected (in terms of randomized selection of schools) and representative of the ratio between rural and urban areas and of the ratio between student populations studying in high schools and in technical and occupational schools, a shortcoming of the present study might also be considered the rather small number of schools from which participants were approached. Furthermore, the present study included attachment to parents through the measurement of attachment to mother. Although, previous studies report that the relationships between attachment security and bullying involvement seem to be consistent for both parents (e.g., Murphy et al., 2017), the study would benefit from the measurement of paternal attachment security. At the same time, we recognize the imbalance that may stem from the use of attachment scales for the mother and peers and the use of an agency/communion scale for teachers. By its own, this might have created a specific conceptualization for teacher-student relationships, which may prompted adolescents to respond differently in that scale. Also, we used mostly measured rather than latent variables in our examined models, due to the complexity of the interactions needed to be examined and the large number of items in each questionnaire. Path models are statistically easier to come to solutions, as they come with a lower number of parameters to be estimated. Finally, the questionnaires used are self-reports and include the pitfalls usually associated with such instruments, such as unaccounted common method variance (Richardson, Simmering, & Sturman, 2009).

Strengths of the Study and Implications

This study is probably one of the first to investigate the schemes of multiple attachment relationships in middle adolescents, and to our knowledge the first that attempted to concurrently examine the multiple attachment relationships’ effects on bullying and victimization. Previous studies have examined only one or two close relationships, and for student-teacher relationships most studies investigated this effect in younger-aged children and not in adolescents. The recognition of the distinct and additive importance of relationships to multiple attachment figures during adolescence is a major finding of the present study.

Another major implication is the need to theoretically improve our understanding of the interrelation between the different close relationships in adolescence and to delineate the pathways, through which these relationships affect adolescents’ involvement in bullying. The present study provided preliminary indications regarding the parameters which are important through each attachment relationship. Maternal attachment seems to affect bullying and victimization only through its effects on peer attachment relationships’ quality and interpersonal communion felt in the relationship with teachers, and these future relationships developed during the school years (i.e., with peers and teachers) have considerable effect in the involvement in dysfunctional behaviors. Especially, when peer attachment relationships and teacher interactions replicate the negative attachment working models developed in the relationships with parents (e.g., low security reflected by high alienation), adolescents have difficulties to protect themselves from experiencing victimization as the patterns of alienation seem to be replicated in their peer interactions. Theoretically this is important, as the systemic investigation of multiple relationships and bullying involvement increases the support for socio-ecological and multi-systemic models for bullying.

At the same time, the findings of our study support the need to apply multi-level prevention and intervention programs at schools, which consider both the interactions with peers and teachers. Skills training programs for teachers, which could enhance their skills to show friendliness and understanding towards their students, without a judgmental stance. High school teachers spend less hours with their students compared to elementary school teachers due to the school schedule, but they seem to remain important figures for adolescents’ development. The present study reveals the importance of applying anti-bullying programs which will also include interpersonal/social skills training with peers, in order to increase the communication and trust in peer attachment relationships.

Future studies should investigate emotion regulation and empathy dimensions as potential mediating mechanisms through which a secure mother-child relationship impacts future relationships. These constructs are strongly related to attachment and bullying (e.g., Nickerson, Mele, & Princiotta, 2008) and a simultaneous exploration of the multiple attachment relationships and the mechanisms through which they affect behavior could guide prevention and intervention efforts. Moreover, the present study did not examine demographic factors as moderators of the models (e.g., gender, SES, racial status), therefore future studies should examine the generalizability of the models for different sub-populations. Finally, research including qualitative assessments with middle adolescents involved in bullying incidents would help underpin the specific relational mechanisms, through which multiple attachment relationships interact and impact school behaviors and would help with the replication of this study’s findings.

Author’s note

This research was supported by internal grants from the University of Cyprus.

References

Arbuckle, J. L. (2006). Amos 7.0 User’s guide. Spring House, PA: Amos Development.

Ainsworth, M. D. (1969). Attachment and exploratory behavior of one-year-olds in strange situation. Determinants of Infant Behavior, 4, 111–136.

Allen, K. P. (2010). Classroom management, bullying, and teacher practices. The Professional Educator, 34, 1–15.

Armsden, G. C., & Greenberg, M. T. (1987). The inventory of parent and peer attachment: Individual differences and their relationship to psychological well-being in adolescence. Journal of Youth and Adolescence, 16,427–454. doi: 10.1007/BF02202939

Bowlby, J. (1969). Attachment and loss. New York: Basic Books.

Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge.

Bretherton, I., & Munholland, K. A. (2008). Internal working models in attachment relationships: Elaborating a central construct in attachment theory. In Cassidy

& Shaver

P. R.

(Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 102–127). New York, NY, US: Guilford Press.

Browne, B. W., & Cudeck, R. (1989). Single sample croos-validation indices for covariance structures. Multivariate Behavioral Research, 24, 136–162. doi: 10.1207/s15327906mbr2404_4

Burton, K. A., Florell, D., & Wygant, D. B. (2013). The role of peer attachment and normative beliefs about aggression on traditional bullying and cyberbullying. Psychology in the Schools, 50, 103–115. doi: 10.1002/pits.21663

10.

Byrne, B. M. (2009). Structural equation modelling with AMOS: Basic concepts, applications and programming (2nd ed.). New York: Routledge Taylor and Francis Group.

11.

Charalampous, K., Demetriou, C., Tricha, L., Ioannou, M., Georgiou, S., Nikiforou, M., & Stavrinides, P. (2018). The effect of parental style on bullying and cyber bullying behaviors and the mediating role of peer attachment relationships: A longitudinal study. Journal of Adolescence, 64, 109–123. doi: 10.1016/j.adolescence.2018.02.003

12.

Charalampous, K., & Kokkinos, C. M. (2013). The Model of Interpersonal Teacher Behaviour: A qualitative cross-cultural validation within the Greek elementary education context. British Educational Research Journal, 39, 182–205. doi: 10.1080/01411926.2011.617437

13.

Charalampous, K., & Kokkinos, C. M. (2018). The structure of pre-adolescents' perceptions of their teacher's interpersonal behaviours and their relation to pre-adolescents' learning outcomes. Educational Studies, 44, 167–189. doi: 10.1080/03055698.2017.1347492

14.

Charalampous, K., Kokkinos, C. M., Apota, E., Iliadou, A., Iosifidou, M., Moysidou, S., & Vriza, E. (2016). Pre-adolescents' representations of multiple attachment relationships: The role of perceived teacher interpersonal behaviour. Learning Environments Research, 19, 63–86. doi: 10.1007/s10984-015-9196-z

15.

Crowell, J. A., Fraley, R. C., & Shaver, P. R. (1999). Measurement of individual differences in adolescent and adult attachment. In Cassidy

& Shaver

P. R.

(Eds.), Handbook ofattachment: Theory, research, and clinical applications (pp. 434–465). New York, NY, US: Guilford Press.

16.

Demanet, J., & van Houtte, M. (2011). School belonging and school misconduct: The differing role of teacher and peer attachment. Journal of Youth Adolescence, 41, 499–514. doi: 10.1007/s10964-011-9674-2

17.

Den Brok, P., Fisher, D., Brekelmans, M., Rickards, T., Wubbels, T., Levy, J., & Waldrip, B. (2003, March). Students' perceptions of secondary science teachers' interpersonal style in six countries: A study on the cross national validity of the Questionnaire on Teacher Interaction. Paper presented at the annual meeting of the National Association for Research in Science Teaching (NARST), Philadelphia, PA, USA.

18.

Eliot, M., & Cornell, D. G. (2009). Bullying in middle school as a function of insecure attachment and aggressive attitudes. School Psychology International, 30, 201–214. doi: 10.1177/0143034309104148

19.

Evans, H., & Fisher, D. (2000). Cultural differences in students' perceptions of science teachers' interpersonal behaviour. Australian Science Teachers Journal, 46, 9–17.

20.

Flanagan, C. A., & Stout, M. (2010). Developmental patterns of social trust between early and late adolescence: Age and school climate effects. Journal of Research on Adolescence, 20,748–773. doi: 10.1111/j.1532-7795.2010.00658x.

21.

Fraley, C. R. (2002). Attachment stability from infancy to adulthood: Meta-analysis and dynamic modeling of developmental mechanisms. Personality and Social Psychology Review, 6, 123–151. doi: 10.1207/S15327957PSPR0602_03

22.

Georgiou, S. N. (2008). Bullying and victimization at school: The role of mothers. British Journal of Educational Psychology, 78, 109–125. doi: 10.1348/000709907X204363

23.

Georgiou, S. N., Fousiani, K., Michaelides, M., & Stavrinides, P. (2013). Cultural value orientation and authoritarian parenting as parameters of bullying and victimization at school. International Journal of Psychology, 48, 69–78. doi: 10.1080/00207594.2012.754104

24.

Gest, S. D., & Rodkin, P. C. (2011). Teaching practices and elementary classroom peer ecologies. Journal of Applied Developmental Psychology, 32, 288–296. doi: 10.1016/j.appdev.2011.02.004

25.

Goh, S. C., & Fraser, B. J. (1996). Validation of an elementary school version of the Questionnaire on Teacher Interaction. Psychological Reports, 79, 515–522. doi: 10.2466/pr0.1996.79.2.515

26.

Haltigan, J. D., & Vaillancourt, T. (2014). Joint trajectories of bullying and peer victimization across elementary and middle school and associations with symptoms of psy-chopathology. Developmental Psychology, 50, 2426–2436. doi: 10.1037/a0038030

27.

Hong, J., & Espelage, D. L. (2012). A review of research on bullying and peer victimization in school: An ecological system analysis. Aggression and Violent Behavior, 17, 311–322. doi: 10.1016/j.avb.2012.03.003

28.

Howes, C., Matheson, C. C., & Hamilton, C. E. (1994). Maternal, teacher, and child care history correlates of children's relationships with peers. Child Development, 65, 264–273. doi: 10.1111/j.1467-8624.1994.tb00749.x

29.

Jacobite, D., & Hazen, N. (1999). Developmental pathways from infant disorganization to childhood peer relationships. In J. Cassidy & P.R. Shaver (Eds.), Handbook of attachment (pp. 671–687). New York: Milford Press

30.

Kokkinos, C. M., Charalambous, K., & Davazoglou, A. (2009). Interpersonal teacher behaviour in primary school classrooms: A cross-cultural validation of a Greek translation of the Questionnaire on Teacher Interaction. Learning Environments Research, 12, 101–114. doi: 10.1007/s10984-009-9056-9

31.

Koul, R. B., & Fisher, D. L. (2005). Cultural background and students' perceptions of science classroom learning environment and teacher interpersonal behaviour in Jammu, India. Learning Environments Research, 8, 195–211. doi: 10.1007/s10984-005-7252-9

32.

Kyriakides, L., Kaloyirou, C., & Lindsay, G. (2006). An analysis of the Revised Olweus Bully/Victim Questionnaire using the Rasch measurement model. British Journal of Educational Psychology, 76, 781–801. doi: 10.1348/000709905X53499

33.

Laible, D. J., Carlo, G., & Raffaelli, M. (2000). The differential relations of parent and peer attachment to adolescent adjustment. Journal of Youth and Adolescence, 29, 45–59. doi: 10.1023/A:1005169004882

34.

Lempers, J. D., & Clark-Lempers, D. S. (1992). Young, middle, and late adolescents' comparisons of the functional importance of five significant relationships. Journal of Youth and Adolescence, 21,53–96. doi: 10.1007/BF01536983

35.

Lereya, S. T., Samara, M., & Wolke, D. (2013). Parenting behavior and the risk of becoming a victim and a bully/victim: A meta-analysis study. Child Abuse and Neglect, 37, 1091–1108. doi: 10.1016/j.chiabu.2013.03.001

36.

Lynch, M., & Cicchetti, D. (1997). Children's relationships with adults and peers: An examination of elementary and junior high school students. Journal of School Psychology, 35, 81–99. doi: 10.1016/S0022-4405(96)00031-3.

37.

Marini, Z. A., Dane, A. V., Bosacki, S. L., & Cura, Y. L. C. (2006). Direct and indirect bully-victims: Differential psychosocial risk factors associated with adolescents involved in bullying and victimization. Aggressive Behavior, 32, 551–569. doi: 10.1002/ab.20155

38.

Marsh, H. W., Wen, Z., & Hau, K.-T. (2004). Structural equation models of latent interactions: Evaluation of alternative estimation strategies and indicator construction. Psychological Methods, 9, 275–300. doi: 10.1037/1082-989x.9.3.275

39.

Menesini, E., Modena, M., & Tani, F. (2009). Bullying and victimization in adolescence: Concurrent and stable roles and psychological health symptoms. The Journal of Genetic Psychology, 170, 115–133. doi: 10.3200/GNTP.170.2.115-134

40.

Murphy, T. P., Laible, D., & Augustine, M. (2017). The influences of parent and peer attachment on bullying. Journal of Child and Family Studies, 26, 1388–1397. doi: 10.1007/s10826-017-0663-2

41.

Nation, M., Vieno, A., Perkins, D. D., & Santinello, M. (2008). Bullying in school and adolescent sense of empowerment: An analysis of relationships with parents, friends, and teachers. Journal of Community & Applied Social Psychology, 18, 211–232. doi: 10.1002/casp.921

42.

Nickerson, A. B., Mele, D., & Princiotta, D. (2008). Attachment and empathy as predictors of roles as defenders or outsiders in bullying interactions. Journal of School Psychology, 46, 687–703. doi: 10.1016/j.jsp.2008.06.002

43.

Nickerson, A. B., & Nagle, R. J. (2005). Parent and peer attachment in late childhood and early adolescence. The Journal of Early Adolescence, 25, 223–249. doi: 10.1177/0272431604274174

44.

Nikiforou, M., Georgiou, S. N., & Stavrinides, P. (2013). Attachment to parents and peers as a parameter of bullying and victimization. Journal of Criminology. Retrieved on 14/7/2018 fromhttp://downloads.hindawi.com/journals/jcri/2013/484871.pdf. doi: 10.1155/2013/484871

45.

O'Brennan, L. M., Bradshaw, C. P., & Sawyer, A. L. (2009). Examining developmental differences in the social-emotional problems among frequent bullies, victims, and bully/victims. Psychology in the Schools, 46, 100–115. doi: 10.1002/pits.20357

46.

O'Connor, E. E., Collins, B. A., & Supplee, L. (2012). Behavior problems in late childhood: The roles of early maternal attachment and teacher-child relationship trajectories. Attachment & Human Development, 14, 265–288. doi: 10.1080/14616734.2012.672280

47.

Olweus, D. (1993). Bullying at school. What we know and whatwe can do?Oxford: Blackwell.

48.

Olweus, D. (1996). The revised Olweus Bully/Victim Questionnairefor Students. Bergen, Norway: University of Bergen.

49.

Olweus, D. (1997). Bully/victim problems in school: Facts and intervention. European Journal of Psychology of Education, 12, 495–510. doi: 10.1007/BF03172807

50.

Orpinas, P., & Horne, A. M. (2006). Creating a positive school climate and developing social competence. Washington, DC: American Psychological Association.

51.

Overall, N. C., Fletcher, G. J., & Friesen, M. D. (2003). Mapping the intimate relationship mind: Comparisons between three models of attachment representations. Personality and Social Psychology Bulletin, 29, 1479–1493. doi: 10.1177/0146167203251519

52.

Pallini, S., Baiocco, R., Schneider, B. H., Madigan, S., & Atkinson, L. (2014). Early child-parent attachment and peer relations: A meta-analysis of recent research. Journal of Family Psychology, 28,118–123. doi: 10.1037/a0035736

53.

Richard, J. F., Schneider, B. H., & Mallet, P. (2011). Revisiting the whole-school approach to bullying: Really looking at the whole school. School Psychology International, 33, 263–284. doi: 10.1177/0143034311415906

54.

Richardson, H. A., Simmering, M. J., & Sturman, M. C.A tale of three perspectives: Examining post hoc statistical techniques for detection and correction of common method variance. Organizational Research Methods, 12, 762–800.

55.

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. 255–282). New York, NY: Guilford Press. doi: 10.1177/1094428109332834

56.

Rydell, A. M., Bohlin, G., & Thorell, L. B. (2005). Representations of attachment to parents and shyness as predictors of children's relationships with teachers and peer competence in preschool. Attachment & Human Development, 7, 187–204. doi: 10.1080/14616730500134282

57.

Schneider, B. H., Atkinson, L., & Tardif, C. (2001). Child-parent attachment and children's peer relations: A quantitative review. Developmental psychology, 37, 86–100. doi: 10.1037/00121649.37.1.86

58.

Sentse, M., Lindenberg, S., Omvlee, A., Ormel, J., & Veenstra, R.Rejection and acceptance across contexts: Parents and peers as risks and buffers for early adolescent psychopathology. The TRAILS study. Journal of Abnormal Child Psychology, 38, 119–130. doi: 10.1007/s10802-009-9351-z

59.

Smith, T. D., & McMillan, B. F. (2001). A primer of model fit indices in structural equation modelling. Paper presented at the annual meeting of the Southwest educational research association, (New Orleans, LA).

60.

Stavrinides, P., Georgiou, S., Nikiforou, M., & Kiteri, E. (2011). Longitudinal investigation of the relationship between bullying and psychosocial adjustment. European Journal of Developmental Psychology, 8, 730–743. doi: 10.1080/17405629.2011.628545

61.

Stavrinides, P., Georgiou, S., & Theofanous, V. (2010). Bullying and empathy: A short-term longitudinal investigation. Educational Psychology, 30, 793–802. doi: 10.1080/01443410.2010.506004

62.

Thijs, J., Koomen, H., Roorda, D., & ten Hagen, J. (2011). Explaining teacher-pupil interactions in early childhood: An interpersonal theoretical approach. Journal of Applied Developmental Psychology, 32, 34–43. doi: 10.1016/j.appdev.2010.10.002

63.

Thompson, R. A. (2008). Measure twice, cut once: Attachment theory and the NICHD study of early childcare and youth development. Attachment & Human Development, 10, 287–297. doi: 10.1080/14616730802113604

64.

van der Zanden, P. J., Denessen, E. J., & Scholte, R. H. (2015). The effects of general interpersonal and bullying-specific teacher behaviors on pupils' bullying behaviors at school. School Psychology International, 36, 467–481. doi: 10.1177/0143034315592754

65.

Verschueren, K., Doumen, S., & Buyse, E. (2012). Relationships with mother, teacher, and peers: Unique and joint effects on young children's self-concept. Attachment and Human Development, 14, 233–248. doi: 10.1080/14616734.2012.672263

66.

Verschueren, K., & Koomen, H. M. (2012). Teacher-child relationships from an attachment perspective. Attachment & Human Development, 14, 205–211. doi: 10.1080/14616734.2012.672260

67.

Walden, L. M., & Beran, T. N. (2010). Attachment quality and bullying behavior in school-aged youth. Canadian Journal of School Psychology, 25, 5–18. doi: 10.1177/0829573509357046

68.

Williams, K. (2011). Bullying behaviors and attachment styles. Electronic Theses & Dissertations. 444. Retrieved on 14/7/2018 fromhttps://digitalcommons.georgiasouthern.edu/etd/444

69.

Wubbels, T., Opdenakker, M. C., & Den Brok, P. (2012). Let's make things better. In Wubbels,

den

Brok, P.

van

Tartwijk, J.

& Levy

(Eds.), Interpersonal relationships in education (pp. 225–249). Rotterdam: Sense Publishers.