The factor structure of the CA-MIR as evaluated using confirmatory factor analysis

Introduction: The aim of the current study was to evaluate the factor structure of the CArtes-Modeles Individuels de Relations (CA-MIR), a self-report questionnaire designed to tap into the relational strategies of adults that was developed by a French-speaking research group coordinated by Blaise Pierrehumbert. The CA-MIR's particular merit lies in the richness and complexity of the theoretical model underpinning it. However, to date, this model has only been partially reproduced in studies using exploratory analysis and has never been tested via confirmatory factor analysis (CFA). Objective and Method: We thus conducted CFA on data collected from a sample of 979 subjects, recruited using a snowball sampling method during the spring and fall of 2005. To assess if some item multidimensionality was present, we estimated both the independent clusters model (ICM-CFA) and a model in which some zero loading restrictions were removed. Results: The results supported the originally proposed structure of the CA-MIR; the large majority of items were good indicators of the expected latent dimensions and only few items showed relevant secondary loadings or loaded in an unexpected factor. The instrument adequately differentiates the three attachment styles, taking into account both past and present experiences of attachment relationships, and providing a rich and complex assessment of multiple features of attachment. In terms of internal consistency, alpha values were satisfactory and comparable to those found in the original Swiss validation study. Conclusions: Our results are of key importance for both research and clinical work, given the lack of valid and easy-to-administer tools for evaluating adult attachment.