ابعاد دلبستگی و صفات شخصیتی پنج عامل بزرگ:ارتباطات و توانایی مقایسه ای برای پیش بینی کیفیت رابطه ای

Several studies have explored associations between measures of adult attachment style and the Big Five personality traits or factors, but the studies have not included current dimensional measures of attachment style (Brennan, Clark, & Shaver, 1998) or the most complete (NEO-PI-R; Costa & McCrae, 1992) and frequently used (BFI; John, Donahue, & Kentle, 1991) measures of the Big Five. Moreover, most studies after Shaver and Brennan’s (1992) have not compared attachment style and Big Five measures as predictors of relationship quality. Here, we summarize past research and report two studies comparing Brennan et al.’s two-dimensional measure of attachment style with the BFI and NEO-PI-R measures of the Big Five. There are consistent and theoretically meaningful associations between the attachment-style and personality trait measures, but attachment-style dimensions still predict relationship quality better than measures of the Big Five. Implications are discussed.

Adult attachment theory (Fraley and Shaver, 2000, Hazan & Shaver, 1987 and Mikulincer and Shaver, 2003) is an extension of Bowlby and Ainsworth’s attachment theory (Ainsworth and Bowlby, 1991 and Bowlby, 1969), designed to explain individual differences in cognitions, feelings, and behaviors that occur in the context of adolescent and adult close relationships. According to the theory, individual differences in “attachment style” emerge from experiences in previous close relationships, beginning with the attachment relationships between children and their primary caregivers. Since 1987, when the theory was first proposed, scores of studies (reviewed by Mikulincer & Shaver, 2003) have shown that measures of attachment style are associated in theoretically predictable ways with mental processes related to close relationships, behaviors observed in such relationships, and outcomes of such relationships, both subjective (e.g., satisfaction) and objective (e.g., breakup or divorce). In recent years, many studies have included both individual-difference measures and experimental manipulations, and have illuminated some of the mental processes, many of them implicit, that underlie variations in attachment style (Mikulincer & Shaver, 2003). In their early research, Hazan & Shaver, 1987 and Hazan and Shaver, 1990 used a simple three-category self-report measure of attachment style based on hypothesized parallels between Ainsworth’s (Ainsworth, Blehar, Waters, & Wall, 1978) three-category typology of infants’ patterns of attachment to their parents. The three patterns were called secure, anxious (or anxious/ambivalent), and avoidant. This measure, which produced both self-ratings of the three category descriptions and selection of the most self-descriptive category, was used by Shaver and Brennan (1992) in an early longitudinal study of predictors of relationship quality and outcomes. In that study, the three category ratings were systematically associated with the then-current measure of the “Big Five” personality traits,1 the NEO-PI (Costa & McCrae, 1985); but the attachment ratings proved to be better predictors of relationship outcomes over time. The study was important in the history of adult attachment research, because it was interpreted as a license to pursue attachment theory as a conceptual framework that was not easily or completely assimilated to the Big Five framework. As is well known, the Big Five personality traits—Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness—have emerged as an overarching, empirically based framework capturing major between-person differences in personality (John & Srivastava, 1999). It is therefore considered parsimonious and sensible to make sure, when any new individual-difference variables are introduced, that they are not simply clones of the Big Five variables bearing new names (the so-called “jangle fallacy”; Block, 2000). Shaver and Brennan’s (1992) study accomplished this task for the early measure of adult attachment style. Over the years, many improvements in the measurement of attachment style have been proposed (e.g., Bartholomew & Horowitz, 1991, Carver, 1997, Collins & Read, 1990 and Simpson, 1990). Some of the improvement efforts are based on the assumption that dimensional measures are more accurate and valid than categorical measures; some are based on dimensional theoretical conceptions of the attachment-style domain, which supersede a simple categorical conception. The most influential of the dimensional schemes is Bartholomew’s (1990), which posits two essentially orthogonal dimensions, model of self (or attachment anxiety) and model of partner (or attachment avoidance) as the factors defining four adult attachment styles. In 1998, Brennan, Clark, and Shaver reported a large factor-analytic study involving virtually all of the self-report attachment style measures proposed up to that time. They found that a two-dimensional, continuous measure of attachment style (the Experiences in Close Relationships scale, or ECR), compatible with the conceptual scheme proposed by Bartholomew, 1990 and Bartholomew & Horowitz, 1991, could represent all of the existing measures while adding considerably to measurement precision. Brenan, Clark, and Shaver (1998) called the two dimensions “attachment-related anxiety” and “attachment-related avoidance,” the first referring to anxiety about rejection, abandonment, and unlovability, and the second to avoidance of intimacy and dependency. Recent research has supported this two-dimensional representation of adult attachment (e.g., Fraley & Shaver, 2000), and also of infant attachment to parents (Fraley & Spieker, 2003). In the personality arena, John, Donahue, and Kentle (1991), coming from the lexical tradition of personality research (rather than the questionnaire approach of Costa and McCrae), introduced their Big Five Inventory (BFI), which has become one of the most commonly used measures of the Big Five traits. During the same period, Costa and McCrae (1992) also improved their NEO-PI, creating the NEO-PI-R (for “revised”), which included six “facet” subscales for each of the Big Five traits. In the 1985 version of the NEO-PI, there were no facet scales for two of the traits, agreeableness and conscientiousness; now there are. For each major trait, the corresponding six facet scales correlate substantially with each other and, together, provide a microanalytic interpretation of the overarching trait. The studies reported here had two main purposes. First, building on a brief review of studies that have examined the relation between attachment style and the Big Five traits (a review summarized in Table 1), we wished to examine how the ECR measure of attachment style relates to the BFI and the NEO-PI-R, two measures that have not been examined previously in association with attachment style. Second, we wanted to update Shaver and Brennan’s (1992) conclusions by seeing whether the ECR, a dimensional measure of attachment style, provides unique predictive power with respect to a measure of relationship quality when the Big Five trait scales or the 30 facet scales of the NEO-PI-R are taken into account. Few studies since Shaver and Brennan (1992) have addressed this question, but since most studies find that attachment style measures and scales assessing the Big Five traits are only modestly or moderately related, it seems likely that the ECR attachment scales will still account for unique variance in relationship quality even after the Big Five traits are statistically controlled (but see Kurdek, 2002).