The purpose of this study was to examine the coherence of attachment organization during late adolescence. In a sample of 53 first-year college students, 3 kinds of working models of attachment were assessed with the Adult Attachment Interview: Dismissing of Attachment, Secure, and Preoccupied with Attachment. Affect regulation was evaluated with peer Q-sort ratings of Ego-Resiliency, Ego-Undercontrol, Hostility, and Anxiety, and representations of self and others were assessed with self-report measures of distress, perceived competence, and social support. The Secure group was rated as more ego-resilient, less anxious, and less hostile by peers and reported little distress and high levels of social support. The Dismissing group was rated low on ego-resilience and higher on hostility by peers and reported more distant relationships in terms of more loneliness and low levels of social support from family. The Preoccupied group was viewed as less ego-resilient and more anxious by peers and reported high levels of personal distress, while viewing their family as more supportive than the Dismissing group. These findings are interpreted in terms of different styles of affect regulation and representational bias associated with particular working models of attachment.

We present a control theory analysis of adolescents' attachment strategies in the Adult Attachment Interview (AAI). In Study 1, Q-sort prototypes for secure/anxious and deactivating/hyperactivating strategies were used to differentiate between Main and Goldwyn's AAI classifications. In Study 2, we examined how AAI strategies were associated with emotion regulation during mother-teen problem solving. 4 aspects of mother-teen problem solving (dysfunctional anger, support/validation, avoidance of problem solving, and maternal dominance) were used to predict teens' AAI strategies. Teens with secure strategies engaged in problem-solving discussions characterized by less dysfunctional anger and less avoidance of problem solving. In addition, attachment security showed a curvilinear relation with maternal dominance, indicating that secure teens maintained balanced assertiveness with their mothers. Teens with deactivating strategies engaged in problem-solving interactions characterized by higher levels of maternal dominance and dysfunctional anger. The contribution of attachment strategies to teens' autonomy and to transformations in mother-teen relationships is discussed.

The purpose of this study was to examine the coherence of attachment organization during late adolescence. In a sample of 53 first-year college students, 3 kinds of working models of attachment were assessed with the Adult Attachment Interview: Dismissing of Attachment, Secure, and Preoccupied with Attachment. Affect regulation was evaluated with peer Q-sort ratings of Ego-Resiliency, Ego-Undercontrol, Hostility, and Anxiety, and representations of self and others were assessed with self-report measures of distress, perceived competence, and social support. The Secure group was rated as more ego-resilient, less anxious, and less hostile by peers and reported little distress and high levels of social support. The Dismissing group was rated low on ego-resilience and higher on hostility by peers and reported more distant relationships in terms of more loneliness and low levels of social support from family. The Preoccupied group was viewed as less ego-resilient and more anxious by peers and reported high levels of personal distress, while viewing their family as more supportive than the Dismissing group. These findings are interpreted in terms of different styles of affect regulation and representational bias associated with particular working models of attachment.

Working models of attachment in marital functioning were examined. The security and accuracy of working models were measured with a new Q-sort method. Spouses with secure working models (self as relying on partner and partner as psychologically available) showed more constructive modulation of emotion and reported better marital adjustment. The accuracy of internal working models, measured with an objective index of spouses' agreement about models, was associated with independent reports of marital adjustment and observers' ratings of communication in problem-solving and confiding tasks. Behavior in communication tasks showed predictable associations with attachment security. Husbands' attachment security covaried with wives' rejection during problem solving, and wives' security covaried with quality of husbands' listening during a confiding task. A reciprocal interaction view of working models and marital functioning is supported.

By asking the subject to consider a host of potentially threatening attachment-related issues, the Adult Attachment Interview (AAI) allows an assessment of different strategies for regulating the attachment system. These strategies can be assessed along the 2 dimensions of security/anxiety and deactivation/hyperactivation. The greatest inferential leaps may be in characterizing strategies as deactivating. For example, individuals using deactivating strategies often report extremely positive relationships with parents, display restricted recall of attachment memories, and play down the significance of early attachment experiences. If these descriptive features are guided by a strategy that requires diverting attention from attachment information, subjects employing this strategy should experience conflict or inhibition during the Attachment Interview. In the present study, skin conductance levels were monitored for 50 college students during a baseline period and throughout the Attachment Interview. Subjects employing deactivating strategies showed marked increases in skin conductance levels from baseline to questions asking them to recall experiences of separation, rejection, and threat from parents. This finding supports the notion that individuals employing deactivating attachment strategies experience conflict or inhibition during the Attachment Interview.