(PS2-37) Factor analysis and measurement invariance of the Five Facet Mindfulness Questionnaire (FFMQ): Higher-order factor structure and invariance across gender identity and meditation experience

Abstract Text: The past decade has seen a steep increase investigating mindfulness and mindfulness-based interventions, necessitating a psychometrically sound assessment of mindfulness. The Five Facet Mindfulness Questionnaire (FFMQ) is a widely-used tool for assessing mindfulness skills, despite questions surrounding its factor structure - including the use of item parcels, the stability of the five-factor model, and the appropriateness of a total score. In addition, the impact of how the FFMQ items are keyed has not received adequate attention in the extant literature. Finally, the stability of the FFMQ’s factor structure across different groups remains an open question, with past findings suggesting that items function differently across gender identities and meditation experience. The current study addresses these issues by examining the factor structure of the FFMQ in two university student samples (Sample 1: n = 938; Sample 2: n = 344). Confirmatory factor analysis (CFA) revealed that the original five-factor model, tested at the item level, did not adequately fit the data. Exploratory structural equation modeling (ESEM) using Sample 1 identified a four-factor model using a shortened 25-item pool (after redundant items were removed). This model was confirmed using CFA in Sample 2 and model fit was found to be largely invariant across participant gender identity and meditation experience. Bifactor analysis found good fit for a model with two higher-order factors added to the aforementioned four-factor model which represented negatively- (NEG) and positively-keyed items (POS). POS correlated significantly with extra-test measures of mindfulness and emotion regulation, while the four first-order factors did not. A similar pattern emerged for NEG, which correlated significantly with measures of depression, worry, stress, and anxiety. The current study represents a step forward in validating an important process measure in mindfulness-based interventions as a function of gender identity and meditation experience. The effort to ensure that we have empirically-validated inclusive and affirming therapies requires that we have psychotherapy process measures whose psychometrics are understood across dimensions of identity.