BackHome
Refresh: OffView Personal Schedule
Individual Submission Summary
Share...

Direct link:

The Dynamics of Attachment Stability and Change: An Empirical Test of Alternative Developmental Processes

Fri, April 9, 10:15 to 11:15am EDT (10:15 to 11:15am EDT), Virtual

Abstract

Attachment theory and research has long emphasized the importance of stability in attachment security, or an individual’s confidence in the availability and responsiveness of their primary caregiver/secure base (e.g. Grossmann et al., 2005). This largely stems from the prediction that early experiences with primary caregivers are organized to form a mental representation or prototype of attachment relationships which is carried forward and informs relationship functioning across development (e.g. Bowlby, 1969/1982). However, evaluating the assumed prototype-like stability of attachment has been complicated by reliance on two-wave designs and a focus on infancy to the exclusion of later developmental periods (Fraley et al., 2011)

Fraley (2002) proposed a modeling approach for evaluating the prototype-like dynamics of attachment stability which leverages multi-wave data. Under a prototype model, early attachment experiences are represented as a single latent factor which contributes equally and consistently to attachment across time. Although this model allows for new experiences to update attachment expectations, the impact of the latent variable on attachment across time is invariant. In contrast to the prototype model, they also proposed a revisionist-contextual model of attachment stability. This model predicts that later experience progressively overwrites prior attachment experiences thus making prior attachment history an increasingly poor predictor of later attachment.

In this study we examined the prototype v. revisionist dynamics of attachment stability data from a 5-wave, 6-year, longitudinal study of childhood attachment security from middle childhood and across the transition into adolescence (N = 157; Wave 1 Mage=10.91, SD=0.87; 51% female). The retention rates for Wave 2-5 follow-ups were 93%, 84.7%, 67.5%, and 83.4% and attrition analyses suggest no consistent significant differences based on demographic data. Attachment security was assessed at each wave using both the Experiences in Close Relationships questionnaire (ECR-RC; Brenning et al., 2014) and the Attachment Script Assessment (ASA; Waters & Waters, 2006). Based on prior research (Fraley et al., 2011), we predicted that both measures would show prototype-like stability.

Results indicated that the ASA and ECR-RC data best fit the prototype model (Table 1). The slopes of the ASA stability correlations (Figure 1) further supported this interpretation as the data demonstrated a clear non-zero asymptote at a moderate effect size. The self-report attachment styles data were somewhat less conclusive. Although the stability data best fit a prototype model (Table 1), the stability correlations plots (Figure 1) did not as clearly indicate a non-zero asymptote as would have been expected under a prototype model. This was especially true of attachment avoidance. Moderator analyses indicated that biological sex and parental divorce significantly undermined the prototype-like stability of attachment style. These findings are consistent attachment theory’s prediction that separations from a primary caregiver due to divorce can undermine security and findings that indicate significant sex differences in attachment security at least by adolescence. We also found that parent-child conflict actually increased prototype-like stability for attachment anxiety. This was not in line with previous work but suggests that parent-child conflict may reinforce/stabilize attachment anxiety more so than it disrupts security (Kobak & Bosmans, 2019).

Authors