This study tests for the first time the validity of universality and normativity assumptions related to the attachment theory in a non-Western culture, using a novel design including psychiatric and non-psychiatric samples as part of a comprehensive exploratory and advanced confirmatory framework. Three attachment assessments were distributed to 212 psychiatric outpatients and 300 non-psychiatric samples in Yemen. The results of the fourteen approaches of exploratory factor analysis (EFA) produce a similar result and assertion that the psychiatric outpatients tend to explore attachment outcomes based on multi-methods, while the non-psychiatric samples suggest an attachment orientation based on multi-traits (self-other). The multiple group-confirmatory factor analysis (MG-CFA) demonstrates that the multi-method model fits the psychiatric samples better than the non-psychiatric samples. Equally, the MG-CFA suggests that the multi-traits model also fits the psychiatric samples better than the non-psychiatric samples. Implications of the results are discussed.
Keywords: attachment orientations; attachment psychosis; multiple group CFA.