On Being Insecure about the Assessment of Attachment Styles

The psychometric properties of subscores from Collins and Read's (1990) Adult Attachment Scale (AAS, Ns = 61 gay, 42 lesbian, and 155 heterosexual couples) and Griffin and Bartholomew's (1994b) Relationship Scales Questionnaire (RSQ, N = 33 gay, 52 lesbian, and 79 heterosexual couples) were examined. Type of couple did not moderate any findings. Dependency and Closeness emerged as reliable factors from the AAS, but, with controls for the Big Five personality factors, neither attachment style accounted for unique variability in relationship satisfaction, and only Closeness accounted for unique variability in relationship commitment. Closeness exerted an indirect effect on commitment through both expressiveness and dysfunctional beliefs regarding relationship standards. Avoidance and Anxiety emerged as reliable factors from the RSQ, but, with controls for the six facets of Neuroticism, only Anxiety accounted for variability in satisfaction and commitment. Anxiety exerted an indirect effect on satisfaction and commitment through positive models of the self and positive models of the other in the current relationship. It is concluded that, although closeness and anxiety may be distinct individual differences variables of relevance to close relationships that exert their indirect effects on relationship outcomes through relationship schemas, additional work is needed to develop a single measure that reliably assesses them.