A Confirmatory Factor Analysis and Validation of the Vulnerable Attachment Style Questionnaire

Abstract

The Vulnerable Attachment Style Questionnaire (VASQ; Bifulco et al., Psychological Medicine, 33, 1099–1110, 2003) was developed to assess adult attachment as a vulnerability factor for developing depression and identified two subscales, insecure attachment and proximityseeking. The present study sought to confirm and further validate the factor structure of the VASQ in a large community convenience sample. The VASQ was completed by a large sample of men and women (N=1236) as part of an online survey. The data were randomly split to allow both independent exploratory (EFA) and confirmatory factor analyses (CFA) to be conducted. A four-factor model consisting of two types of proximity-seeking (lack of autonomy and anxious-dependent) and insecurity (ambivalent and avoidantdismissive) attachment patterns proved to be the best-fitting measurement model in this sample (X2=186.7, df=71, p<.001; CFI=.945, TLI=.929, RMSEA=.05). Although similar to the original questionnaire, the new factor structure resulted in the elimination of several items. Validity was confirmed with the shortened VASQ as similar associations with mood, stress, eating pathology and sex were observed for both the new shortened VASQ and original version of the VASQ. The structure of the VASQ was broadly consistent with the original solution although some items were removed and both subscales were further split into two sub-factors. Future research should use this tool in clinical and nonclinical groups to provide further support for its factor structure and to determine the clinical and theoretical usefulness of the different subscales.