Can different adult attachment profiles be distinguished when categorical and dimensional approaches are combined? Evidence based on two questionnaires

Abstract The study of adult attachment styles has been approached either from a categorical or from a dimensional perspective, although some authors advocate a combination of both. In this last line, the present study assessed which attachment styles can be identified in two different questionnaires (CAA-r and ECR) using Factor Mixture Analysis (FMA), a technique that combines the categorical and dimensional approaches by allowing the hypothesis of different numbers of classes within a well-defined factorial structure to be rigorously tested. Furthermore, the appropriateness of different class-solutions can be assessed instead of imposing a fixed number a priori. These advantages are particularly relevant in attachment research overcoming the theoretical and methodological limitations of the previous approaches and contributing to both theoretical and practical advances in this domain. The sample consisted of 619 adults (60% female) and the FMA and validity analyses were performed with MPlus . In both measures, the FMA results suggest that there is a majority profile of individuals with neither high nor low means in any attachment factor, and a minority profile of insecure attachment, as had already been found previously with the CAA questionnaire. Furthermore, a different pattern of relationships between the attachment factor levels and other variables (perceived maternal care, perceived maternal overprotection and neuroticism) was found for each profile. Both the FMA and validity-analyses results point to the interest in differentiating between profiles, especially in the CAA-r. However, the results seem to be halfway between what is proposed from the dimensional and from the categorical perspectives.