Poster #142 - Measuring attachment insecurity in middle-childhood: a comparison across instruments

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Parenting & Parent-Child Relationships

Integrative Statement

Introduction. In the last 20 years, the study of attachment in middle childhood has received much attention from scholars, leading to the development of several assessment measures for this specific period. Two self-report questionnaires have been proposed to measure the two main insecurity dimensions of attachment, namely the Experiences in Close Relationships-Revised Child version (ECR–RC, Brenning, Van Petegem, & Soenens, 2014) and the Preoccupied and Avoidant Coping Questionnaire (PACQ; Finnegan, Hodges, & Perry, 1996). The former is designed to assess avoidance and anxiety, while the latter evaluates avoidant and preoccupied/ambivalent attachment. Although the PACQ preoccupied subscale has been defined as “similar to anxious attachment” (Brenning et al., 2012, p. 2), its association with the ECR-RC anxiety dimension has not been formally tested.
The current study aimed to a) evaluate the mutual association between the PACQ and the ECR-RC, and b) test the concurrent, convergent and predictive validity of the two questionnaires.
Method. Participants were 196 native-born Italian children (48,2% boys, age 8.1-10.4 years) attending public schools in Italy who completed the mother-forms of the PACS and the ECR-RC, together with a questionnaire of felt security. The main teacher also reported on children’s internalizing, externalizing, and prosocial behaviors at school. After two years, 165 children participated in a second wave and were assessed for their perceived self-worth and use of emotion regulation strategies, i.e. cognitive reappraisal and expressive suppression.
Analytic approach. As a preliminary step, the factorial structure of the two questionnaires was evaluated by Confirmatory Factor Analysis (CFA). Then, a series of Structural Equation Models (SEM) was performed to evaluate the association between latent attachment dimensions (i.e., PACQ: preoccupied and avoidant; ECR-RC: anxious and avoidant) and their links with felt security (concurrent validity), with internalizing, externalizing, and prosocial behaviors (convergent validity), and with global self-worth and the two emotion regulation strategies after two years (predictive validity).
Results. CFAs indicated a good factor structure of both questionnaires (see Table 1). Strong and meaningful associations were found between the PACQ and ECR-RC avoidance dimensions (.77), but not between the PACQ-preoccupied and ECR-RC-anxious subscales (-.09). The avoidant dimensions as assessed via the PACQ and ERC-RC correlated similarly with the considered criterion validity measures, whereas the preoccupied and anxiety dimensions showed different, and sometimes opposite, patterns of associations with almost all external measures (see Table 2). More preoccupied coping was linked to higher levels of felt security and predicted increased use of cognitive reappraisal after two years. On the contrary, the ECR-RC anxiety dimension was negatively related to felt security and not associated with cognitive reappraisal.
Conclusion. Our findings provide only partial support for the mutual association between the two questionnaires. While the avoidance dimensions of the PACQ and ECR-RC showed similar correlations with the external measures, a less clear pattern was found for the PACQ-preoccupied subscale. Thus, the two instruments may be considered as complementary rather than overlapping. Further studies are needed to ascertain whether in middle childhood, the PACQ-preoccupied subscale reflects normative aspects of child-mother attachment more than preoccupation per se.

Authors

• Tatiana Marci, University od Padova
  Presenting Author

• Ughetta Moscardino, University of Padova
  Non-Presenting Author

• Xiaoyu Lan, University of Padova
  Non-Presenting Author

• Gianmarco Altoè, University of Padova
  Non-Presenting Author