Introduction: Although contemporary approaches conceptualize mentalizing as a multidimensional construct, many available instruments assess only a limited range of its facets. The Multidimensional Mentalizing Questionnaire (MMQ) was developed to simultaneously capture both adaptive and maladaptive dimensions of mentalizing. However, critical reviews have highlighted that the conceptual distinction between the capacity/process of mentalizing and the outcomes of impaired mentalizing is not clearly articulated in the original MMQ structure. The present study aimed to evaluate the psychometric properties of the Persian version of the MMQ in a sample of Iranian university students and to revise its factor structure in order to achieve a clearer differentiation between mentalizing dimensions.
Method: In this descriptive–correlational study, 248 students from universities in Isfahan were selected using convenience sampling. The Persian version of the MMQ was prepared through a forward–backward translation procedure. Face and content validity were examined using quantitative indices and qualitative feedback from expert reviewers. Three competing models—including the original six factor model, a five factor model, and a second order two factor model—were tested via confirmatory factor analysis using the diagonaly weighted least squares (DWLS) . Convergent validity was assessed through correlations with the Reflective Functioning Questionnaire, the Toronto Alexithymia Scale, and the Satisfaction With Life Scale, and discriminant validity was evaluated using the heterotrait–monotrait ratio of correlations (HTMT). Internal consistency of the subscales was estimated using appropriate reliability indices.

Results: Based on expert feedback, several items were revised and some were identified as potential candidates for removal. Nonetheless, in the first step, all items were tested within a six factor confirmatory model, which showed poor fit and suboptimal performance of the “Distrust” factor and several items. In the next step, integrating the evidence from the previous phases, nine items were removed and a five factor, 24 item model was tested, followed by a second order two factor structure comprising “Mentalizing Propensity” and “Impaired Mentalizing.” Although both models demonstrated acceptable fit, the evidence specifically supported the the second order two factor model for the Persian version of the MMQ. Indices of validity and reliability were also found to be satisfactory.

Discussion and conclusion: The suboptimal statistical performance of the “Mistrust” factor, which effectively removed it from the nomological network of mentalizing, may indicate a misalignment between the original MMQ formulation and patterns of trust in the predominantly collectivistic Iranian cultural context. In contrast, reverse-scoring the “Ego Strength” factor and reallocating it to the “Mentalizing Impairment” cluster yielded a clearer differentiation between “mentalizing capacity/processes” and the outcomes of its failure.