Language use in the Adult Attachment Interview: Evidence for attachment‐specific emotion regulation

Adult attachment classification is traditionally based on qualitative coding of participants' discourse about their attachment history. Word count‐based analyses have proven useful for assessing emotional states from narrative. To expand the understanding of how language is used in emotion regulation processes related to attachment, the authors assess 102 college‐aged adults' language on the Adult Attachment Interview (AAI). Autonomous adults use more emotion words and, in particular, negative emotion words. Preoccupied adults use more anger words. Disorganized adults use more experientially connected language and more death/dying words, but also use more second‐person pronouns when discussing loss. Language use during the AAI explains variability in self‐reported emotional distress above and beyond attachment classifications. Results are discussed in terms of their relevance to emotion and attachment.     

Stability of attachment representations: the transition to marriage

This study examined the stability of adult attachment representations across the transition to marriage. One hundred fifty-seven couples were assessed using the Adult Attachment Interview (AAI; C. George, N. Kaplan, & M. Main, 1985), the Current Relationship Interview (J. A. Crowell & G. Owens, 1996), and measures describing relationship functioning and life events 3 months prior to their weddings and 18 months into their marriages. The authors tested the hypotheses that attachment classifications are stable and that change is related to experiences in the relationship and/or life events; 78% of the sample received the same primary AAI classification (secure, preoccupied, and dismissing) at both times. Change was toward increased security and was associated with feelings and cognitions about the relationship. Only 46% of participants initially classified as unresolved retained the classification. Stability of the unresolved classification was associated with stressful life events and relationship aggression.     

Individual differences in adult attachment: Disentangling two assessment traditions

The goal of this paper was to disentangle the contributions of type of measurement and relational domain to the lack of concordance between adult attachment measures. Eighty-six college students completed the Adult Attachment Interview (AAI) and self-reported measures of attachment security adapted to their relationships with four different relational partners. Results showed that whereas the AAI main dimensions were unrelated to self-reports relative to any partner, reports of attachment security with different partners were interrelated. Further, very few of the AAI State of Mind scales related to self-reported attachment security, but the AAI Experience scales did relate to self-reports in a theoretically consistent manner. Finally, the AAI scales of Idealization and Lack of Recall correlated positively with self-reported attachment when controlling for experiences reported on the AAI. The results suggest that a fundamental distinction between adult attachment measures could be the degree of automaticity of the constructs tapped, and that attachment self-reports may be biased by defensive information-processing strategies among dismissing individuals.     

The Adolescent Unresolved Attachment Questionnaire: The Assessment of Perceptions of Parental Abdication of Caregiving Behavior

This article reports on the Adolescent Unresolved Attachment Questionnaire (AUAQ), a brief questionnaire that assesses the caregiving experiences of unresolved adolescents (as recipients of caregiving). The AUAQ was developed and validated in a large normative sample (n = 691) and a sample of 133 adolescents in psychiatric treatment. It is a self-report questionnaire consisting of 3 scales with Likert-type responses ranging from strongly disagree to strongly agree. The Aloneness/Failed Protection Scale assesses the adolescent's perception of the care provided by the attachment figure. The Fear Scale taps the fear generated by the adolescent's appraisal of failed attachment figure care. The Anger/Dysregulation Scale assesses negative affective responses to the perceived lack of care from the attachment figure. All scales demonstrated satisfactory internal reliability and agreement between scores for adolescents (n = 91) from the normative sample who completed the AUAQ twice. Adolescents in the clinical sample also completed the Adult Attachment Interview (AAI; C. George, N. Kaplan, & M. Main, 1984/1985/1996); the AUAQ demonstrated high convergent validity with the AAI.     

The emotional integration of childhood experience: physiological, facial expressive, and self-reported emotional response during the adult attachment interview

Attachment researchers claim that individual differences in how adults talk about their early memories reflect qualitatively distinct organizations of emotion regarding childhood experiences with caregivers. Testing this assumption, the present study examined the relationship between attachment dimensions and physiological, facial expressive, as well as self-reported emotional responses during the Adult Attachment Interview (AAI). Consistent with theoretical predictions, more prototypically secure adults behaviorally expressed and reported experiencing emotion consistent with the valence of the childhood events they described. Insecure adults also showed distinctive and theoretically anticipated forms of emotional response: Dismissing participants evidenced increased electrodermal activity during the interview, a sign of emotional suppression, whereas preoccupied adults showed reliable discrepancies between the valence of their inferred childhood experiences and their facial expressive as well as reported emotion during the AAI. Results substantiate a case that the AAI reflects individual differences in emotion regulation that conceptually parallel observations of attachment relationships in infancy.     

Attachment representations and characteristics of friendship relations during adolescence

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

When "new" meets "old": configurations of adult attachment representations and their implications for marital functioning

Two studies addressed the implications of concordance versus discrepancy of attachment representations in individuals at 2 stages in their marital relationships. Engaged (n = 157) and dating (n = 101) couples participated in a multimethod 6-year longitudinal study of adult attachment. Individuals completed the Adult Attachment Interview (AAI), the Current Relationship Interview (CRI), and various questionnaires and were observed in interactions with partners. On the basis of AAI and CRI classifications, participants were placed in one of four groups: Secure-sub(AAI)/Secure-sub(CRI), Secure-sub(AAI)/Insecure-sub(CRI), Insecure-sub(AAI)/Secure-sub(CRI), or Insecure-sub(AAI)/Insecure-sub(CRI). Each of the configurations showed a particular pattern of behavior, feelings about relationships and the self, and likelihood of relationship breakup. The findings of the studies address important points about the protective effects of attachment security and have interesting implications for the extension of attachment theory into adulthood.     

The Adolescent Unresolved Attachment Questionnaire: the assessment of perceptions of parental abdication of caregiving behavior

This article reports on the Adolescent Unresolved Attachment Questionnaire (AUAQ), a brief questionnaire that assesses the caregiving experiences of unresolved adolescents (as recipients of caregiving). The AUAQ was developed and validated in a large normative sample (n = 691) and a sample of 133 adolescents in psychiatric treatment. It is a self-report questionnaire consisting of 3 scales with Likert-type responses ranging from strongly disagree to strongly agree. The Aloneness/Failed Protection Scale assesses the adolescent's perception of the care provided by the attachment figure. The Fear Scale taps the fear generated by the adolescent's appraisal of failed attachment figure care. The Anger/Dysregulation Scale assesses negative affective responses to the perceived lack of care from the attachment figure. All scales demonstrated satisfactory internal reliability and agreement between scores for adolescents (n = 91) from the normative sample who completed the AUAQ twice. Adolescents in the clinical sample also completed the Adult Attachment Interview (AAI; C. George, N. Kaplan, & M. Main, 1984/1985/1996); the AUAQ demonstrated high convergent validity with the AAI.     

The adult attachment interview and the relationship questionnaire: Relations to reports of mothers and partners

The study asked the following questions: How do interview and self-report assessments of attachment status correspond? What is the relation between attachment status assessed by these methods to self-reports of relationships with mothers and partners? Participants were 53 married white women assessed with the Adult Attachment Interview (AAI), and the Relationship Questionnaire (RQ), and a variety of questionnaires assessing love, trust, and closeness in relationships with mothers and partners. Results indicated a trend toward a relation between AAI and RQ classifications: 81% of women classified as Secure with the AAI identified themselves as Secure with the RQ, but only 42% of AAI-Insecure women identified themselves as Insecure with the RQ. Secure and Insecure AAI groups did not differ in their reports of mothers or partners, whereas Secure and Insecure RQ groups did. Results indicated classifications derived from the measures are not equivalent.     

The Adult Attachment Interview and self-reports of attachment style: an empirical rapprochement

Although 10 studies have been published on the empirical overlap of the Adult Attachment Interview (AAI) and measures of self-reported attachment style, results in this literature have been inconsistently interpreted in narrative reviews. This report was designed as a rapprochement of the AAI and attachment style literatures and includes 3 studies. Study 1 (combined N = 961) is a meta-analytic review showing that by J. Cohen's (1992) criteria (mean r = .09), the association between AAI security and attachment style dimensions is trivial to small. Study 2 (N = 160) confirms meta-analytic results with state-of-the-art assessments of attachment security and also examines attachment dimensions in relation to the Big 5 personality traits. Finally, Study 3 is an investigation of 50 engaged couples that shows that developmental and social psychological measures of attachment security predict somewhat distinct--though theoretically anticipated--aspects of functioning in adult relationships.     

Fathers' and mothers' attachment representations as predictors of preadolescents' attachment security: A ten‐year follow‐up of Finnish families

The question of how mothers' and fathers' representations of attachment correlate ten years later with children's perceptions of attachment relationships was examined in a longitudinal study on Finnish families (N = 42). The parents completed the Adult Attachment Interview (AAI) during the child's first year of life. At 11 years, the children filled out three scales on how secure they perceive the relationship with each parent. Parents' AAI classifications and AAI dimensions based on continuous scales were used as predictors of the preadolescents' attachment security. Regression analyses demonstrated that fathers' but not mothers' State‐of‐Mind and Experience dimensions predicted preadolescents' security of attachment to father. The discussion focuses on the predictive validity of the classical categorical versus the recently proposed continuous approach and the different roles of parents in transmitting security from one generation to another.     

The impact of adult attachment style on organizational commitment and adult attachment in the workplace

Adult attachment style has only recently been considered as having a role in explaining work behavior. The present research aimed to explore the impact of adult attachment style, assessed by the Adult Attachment Interview (AAI), on organizational commitment (OC) and on adult attachment in the workplace (AAW). We hypothesized that a secure attachment style would be positively related to affective and normative commitment, while preoccupied and avoidant styles would be negatively related to affective commitment; we also hypothesized that there would be a correspondence between the AAI categories and the AAW dimensions. Using the AAI categories as group variable, analysis of average OC and AAW scores confirmed the hypotheses. Secure workers had a higher mean score for affective commitment than avoidant and preoccupied workers; normative commitment was higher in avoidant than in secure and preoccupied workers; continuance commitment was higher in preoccupied than in secure and avoidant workers. Moreover, AAI categories converged with AAW dimensions: secure workers had higher secure AAW scores than avoidant and preoccupied workers; avoidant workers had higher avoidant AAW scores than secure and preoccupied workers; preoccupied workers had higher preoccupied AAW scores than secure and avoidant workers.     

Observing attachment behavior in couples: the Adult Attachment Behavior Q-Set (AABQ)

Knowledge as to how attachment behavior is observable in couple interaction can be very useful to clinicians who use attachment theory or related theories to guide their work with couples. The development of the Adult Attachment Behavior Q-Set (AABQ), a 100-item Q-sort designed to be consistent with Main's Adult Attachment Interview (AAI) coding system, is described. Videotaped discussions were coded for a sample of 28 couples. The majority of the couples were married and had sought therapy for couple problems. Reliability of coding was adequate. Evidence for validity was promising, with a match of 84% between the AAI and AABQ for secure versus insecure, and 70% for the three categories of secure, dismissing, and preoccupied. Even though additional reliability and validity information are needed for the AABQ, it holds promise for providing insights into couple dynamics that could be useful in guiding interventions and in evaluating the outcome of couple therapy.     

Psychometric Properties of Two Implicit Associations Tests measuring Adult Attachment

Adult attachment style has consequences for mental health, interpersonal functioning and emotion regulation. This occurs partly deliberately, also referred to as explicit, and partly on an automatic level outside of conscious awareness, also referred to as implicit. Whereas explicit adult attachment can be assessed with self-report instruments, measurement of implicit adult attachment requires indirect methods. This paper describes the psychometric properties of two Implicit Association Tests measuring general adult attachment in a population sample. The study evaluated the reliability and the validity of the Avoidant Attachment IAT (ANX-IAT) and the Anxious Attachment IAT (AVOID-IAT). Validity was evaluated against self-report measures of adult attachment style (RQ), psychopathology (SQ-48), and well-being (MHC-SF). The split-half reliabilities of both IATs were good; the test-retest reliability of the ANX-IAT was adequate; however the AVOID-IAT had low test-retest reliability. Both IATs did not explain variance in psychopathology additional to explicit measures. The AVOID-IAT showed added value over explicit measurement of avoidant attachment in explaining variance in well-being, particularly regarding emotional and psychological well-being. The ANX-IAT did not explain variance in any measure of well-being additional to the explicit measure of anxious attachment. Our findings provide a basis from which more valid IATs measuring general adult attachment can be developed. Furthermore, they suggest that implicit avoidant attachment might be related to well-being, particularly emotional and psychological well-being. However, further research is needed to investigate the role of implicit general adult attachment in mental health and to optimize the two IATs in terms of validity before clinical use is recommended.     

Working models of attachment and reactions to different forms of caregiving from romantic partners

Inspired by attachment theory, the authors tested a series of theoretically derived predictions about connections between attachment working models (attachment to one's parents assessed by the Adult Attachment Interview; M. Main & R. Goldwyn, 1994) and the effectiveness of specific types of caregiving spontaneously displayed by dating partners during a stressful conflict-resolution discussion. Each partner first completed the Adult Attachment Interview. One week later, each couple was videotaped while they tried to resolve a current problem in their relationship. Trained observers then rated each interaction for the degree to which (a) emotional, instrumental, and physical caregiving behaviors were displayed; (b) care recipients appeared calmed by their partners' caregiving attempts; and (c) each partner appeared distressed during the discussion. Individuals who had more secure representations of their parents were rated as being more calmed if/when their partners provided greater emotional care, especially if they were rated as more distressed. Conversely, individuals who had more insecure (dismissive) representations of their parents reacted more favorably to instrumental caregiving behaviors from their partners, especially if they were more distressed. The broader theoretical implications of these findings are discussed.     

Calibrating Use of Emotion Regulation Strategies to the Relationship Context: An Attachment Perspective

This research tested whether adult attachment orientations predict use of emotion regulation strategies in theoretically consistent ways, and whether associations among attachment orientations and emotion regulatory strategies are moderated by critical features of the relationship context. Ninety‐six couples (192 individuals) reported on their attachment orientations, habitual use of emotion regulation strategies (cognitive reappraisal, expressive suppression, negative emotion expressivity), and perceptions of relationship closeness and negative partner behaviors. Highly secure individuals reported greater use of cognitive reappraisal, especially when they felt closer to their partners, and engaged in less suppression when their partners behaved more negatively toward them. Highly avoidant individuals reported greater use of suppression, especially when they perceived more negative partner behaviors, and when their partners were more avoidant. Highly anxious individuals also used more suppression when their partners were more avoidant, but they expressed more negative emotions when they were paired with less avoidant partners. Fearful‐avoidant individuals' emotion regulation patterns resembled those of both highly secure and dismissive‐avoidant individuals. This study illustrates how attending to moderating effects within specific relationships and testing joint effects of both partners' personality characteristics can help identify contextual boundaries of emotion regulation strategies and clarify emotional response patterns in couples.     

Le rôle joué par les événements stressants sur la transmission intergénérationnelle de l’attachement

Associations between maternal internal working models (IWM) (adult attachment projective: AAP) [George, C., West, M., Pettem, O., 1997. Adult attachment projective: protocol and classification scoring system. Manuscrit non publié. Mills College, Californie], preschool child attachment (separation–reunion procedure) [Cassidy, J., Marvin, R. S., 1992. Attachment organization in preschool children: procedures and coding manual. Manuscrit non publié. MacArthur Working Group on Attachment; Dev Psychol 24 (1988) 415–526] and stressful life-events were examined in 119 dyads. A convergence for transmission similar to what has been reported in the literature [Psychol Bull 117(3) (1995) 387–403] was found (76% for security, kappa = 0.53; p < 0.001). Moreover, unresolved mothers with disorganized–controlling children reported more stressful life-events (F [3, 106] = 4.70; p < 0.01). The present study therefore indicates the importance of including contextual factors [van IJzendoorn, M.H., Bakermans-Kranenburg, M.J., 1997. Intergenerational transmission of attachment: a move to the contextual level. In: Atkinson et, L., Zucker, K.J. (Eds.), Attachment and Psychopathology. The Gilford Press, New York, pp. 135–170] to better understand attachment transmission and offer support for the validity of the AAP.     

Attachment and Autism: Parental Attachment Representations and Relational Behaviors in the Parent-Child Dyad

While attachment research has demonstrated that parents’ internal working models of attachment relationships tend to be transmitted to their children, affecting children’s developmental trajectories, this study specifically examines associations between adult attachment status and observable parent, child, and dyadic behaviors among children with autism and associated neurodevelopmental disorders of relating and communicating. The Adult Attachment Interview (AAI) was employed to derive parental working models of attachment relationships. The Functional Emotional Assessment Scale (FEAS) was used to determine the quality of relational and functional behaviors in parents and their children. The sample included parents and their 4- to 16-year-old children with autism and associated neurodevelopmental disorders. Hypothesized relationships between AAI classifications and FEAS scores were supported. Significant correlations were found between AAI classification and FEAS scores, indicating that children with autism spectrum disorders whose parents demonstrated secure attachment representations were better able to initiate and respond in two-way pre-symbolic gestural communication; organize two-way social problem-solving communication; and engage in imaginative thinking, symbolic play, and verbal communication. These findings lend support to the relevance of the parent’s state of mind pertaining to attachment status to child and parent relational behavior in cases wherein the child has been diagnosed with autism or an associated neurodevelopmental disorder of relating and communicating. A model emerges from these findings of conceptualizing relationships between parental internal models of attachment relationships and parent-child relational and functional levels that may aid in differentiating interventions.     

Mothers' attachment status as determined by the Adult Attachment Interview predicts their 6-year-olds' reunion responses: a study conducted in Japan

Following a 1986 study reporting a predominance of ambivalent attachment among insecure Sapporo infants, the generalizability of attachment theory and methodologies to Japanese samples has been questioned. In this 2nd study of Sapporo mother-child dyads (N=43), the authors examined attachment distributions for both (a) child, based on M. Main and J. Cassidy's (1988) 6th-year reunion, and (b) adult, via the Adult Attachment Interview (AAI). In contrast to the previous Sapporo study, children's 3-way or "organized" distribution did not differ from the global distribution. However, when the disorganized-controlling (D) and cannot classify (CC) categories were applied to the analyses, a high proportion of D/CC children was found. Comparable analyses for Japanese mothers, including the unresolved/disorganized (U) and CC categories, were found to deviate slightly from the global norm. However, turning from global distributions to mothers' AAI classification as related to their child's reunion classification, all matches were surprisingly close to those established worldwide. When, as is customary, mothers' U and CC classifications were combined (U/CC) and compared with the child's D and CC classifications (also customarily combined as D/CC), mothers' U/CC status strongly predicted child D/CC status (r=.60, d=1.50). Additionally, mothers' AAI subclassifications predicted child subclassifications.     

POSTPARTUM BONDING DIFFICULTIES AND ADULT ATTACHMENT STYLES: THE MEDIATING ROLE OF POSTPARTUM DEPRESSION AND CHILDBIRTH‐RELATED PTSD

Despite decades of research demonstrating the role of adult attachment styles and early mother–infant bonding in parenting behaviors and maternal mental health, these constructs have seldom been studied together. The present study aimed to investigate the relationship between attachment styles and specific bonding difficulties of mothers. In addition, as postpartum depression and childbirth‐related posttraumatic stress symptoms have been associated with both constructs, we explored their possible mediation effect. One hundred fourteen mothers, 4 to 12 weeks’ postpartum, completed a demographic questionnaire, the Adult Attachment Style Questionnaire (M. Mikulincer, V. Florian, & A. Tolmacz, 1990), the Postpartum Bonding Questionnaire (L.F. Brockington, C. Fraser, & D. Wilson, 2006), the Modified Perinatal Posttraumatic Stress Disorder Questionnaire (J.L. Callahan, S.E. Borja, & M.T. Hynan, 2006), and the Edinburgh Postnatal Depression Scale (J.L. Cox, G. Chapman, D. Murray, & P. Jones, 1996), using an online survey system. As predicted, insecure attachment styles were associated with bonding difficulties wherein anxious/ambivalent attachment was associated with greater infant‐focused anxiety, mediated by postpartum depression but not childbirth‐related PTSD symptoms. In contrast, greater avoidant attachment style was associated with greater rejection and anger, mediated by childbirth‐related posttraumatic stress disorder (PTSD), but not depression symptoms. The current study confirmed the association of different attachment styles with bonding as well as the mediating roles of childbirth‐related PTSD and postpartum depression symptoms. Future psychological interventions may utilize such evidence to target interventions for bonding disorders in accordance with individual differences.