Associations among relational maintenance behaviors, attachment-style categories, and attachment dimensions.

Communication is critical for maintaining satisfying relationships. As Dindia (2003) put it, "to maintain the quality of a relationship, one must maintain the quality of the communication" (p. 1). Research has identified a set of prosocial maintenance behaviors that associate with positive relational qualities, such as commitment, trust, love, liking, and satisfaction (e.g., Dainton, Stafford, & Canary, 1994; Dindia & Baxter, 1987; Stafford & Canary, 1991; Stafford, Dainton, & Haas, 2000; Weigel & Ballard-Reisch, 1999a, 1999b). Furthermore, couples who use prosocial maintenance behaviors are more likely to stay together or escalate their relationships (Guerrero, Eloy, & Wabnik, 1993). According to Canary and Zelley (2000), three antecedent factors predict how much prosocial maintenance behavior people use in their relationships--equity, relational type and history, and individual differences.

The present study focuses on one individual difference variable that has been shown to associate with maintenance behavior--attachment style. Attachment styles are "relatively coherent and stable patterns of emotion and behavior [that] are exhibited in close relationships" (Shaver, Collins, & Clark, 1996, p. 25). Thus, it is reasonable to expect that people with various attachment styles would maintain their relationships in line with differentiated patterns of emotion and behavior (Dainton, Zelley, & Langan, 2003). Indeed, Simon and Baxter (1993) demonstrated that people with secure attachment styles report using more assurances and romance in their romantic relationships than do people with dismissive attachment styles. Similarly, Bippus and Rollin (2003) found people to perceive secure friends as using more prosocial maintenance behavior than insecure friends.

The present study extends this line of work in three primary ways. First, we examine a wider variety of prosocial maintenance behaviors--assurances, romantic affection, positivity, openness, social networking, task sharing, and comfort/support--than have past studies on attachment and maintenance. Second, in addition to investigating attachment style categories (secure, dismissive, fearful, and preoccupied), we examine the attachment dimensions of anxiety and avoidance. These two dimensions represent the underlying structure of attachment styles (e.g., low anxiety and high avoidance is characteristic of the dismissive style). Finally, we collected data from both members of romantic couples, allowing us to determine how one's own versus one's partner's attachment orientations associate with maintenance behavior.

Attachment Theory

According to attachment theorists, the way people think and act in their intimate relationships is guided by cognitive models about themselves and significant others (Brennan & Shaver, 1995; Collins & Read, 1990; Hazan & Shaver, 1987; Kirkpatrick & Davis, 1994). These working models first develop in infancy as a result of interaction with one's caregiver (Ainsworth, Blehar, Waters, & Wall, 1978; Bowlby, 1969, 1973, 1980). Children who develop positive models of self perceive themselves as confident and likable, whereas those with negative models of self have low self-esteem. Children with positive versus negative models of others learn to trust or to distrust others, respectively. These early working models provide a foundation for personality development. However, attachment theory is more than a description of personality types; the theory provides an account of how people develop cognitive schema that guide perceptions and social behavior. According to Bartholomew (1993), social behavior that is consistent with one's attachment orientation reinforces one's working models of self and others, creating stability in attachment styles (i.e., a preoccupied person wants to escalate intimacy too quickly, which pushes the partner away and reinforces the preoccupied person's low self-esteem). However, attachment styles can also be modified based on new interactions with significant others (Feeney & Noller, 1996). Thus, the theory predicts that interpersonal communication influences and is influenced by working models of self and others.

Bartholomew's (1990; Bartholomew & Horowitz, 1991) work recognizes the critical role that models of self and others play within attachment theory. She proposed a two-dimensional model of attachment styles that is based on crossing these two working models. The model of self has been conceptualized as an anxiety dimension, with secure and dismissive individuals perceiving themselves as relatively low in anxiety (or high in self-confidence) whereas preoccupied and fearful individuals perceive themselves as relatively high in terms of anxiety, especially as related to low self-esteem and fear of abandonment or rejection (see also, Brennan, Clark, & Shaver, 1998; Feeney, Noller, & Hanrahan, 1994). The model of others, in contrast, has been conceptualized as an avoidance dimension, with secure and preoccupied individuals theorized to hold approach orientations toward relationships (and to be comfortable with intimacy) whereas dismissive and fearful individuals are theorized to have avoidant orientations toward relationships (and to be uncomfortable with intimacy).

When these two dimensions are crossed, four distinct attachment styles emerge. Secures have positive views of self and others; they tend to have high self-esteem, trust others and are comfortable with intimacy and independence. Preoccupieds have a negative view of self and a positive view of others; they tend to have low self-esteem, worry that others will abandon them and are preoccupied with relationships. Dismissives have a positive view of themselves and negative views of others; they tend to have high self-esteem, avoid intimacy and are overly independent. Fearful Avoidants have a negative perception of themselves and others; they tend to have low self-esteem, tend to have low trust in others and fear commitment and intimacy (Bartholomew & Horowitz, 1991).