Rethinking the PERMA-Profiler: A multidimensional Item Response Theory (MIRT) evaluation

The PERMA-Profiler is a self-report instrument designed to assess personal psychological well-being, including “PERMA” (an acronym for Positive Emotions, Engagement, Relationships, Meaning, and Accomplishment). Despite consistent validation of its descriptive utility and latent five-factor measurement structure, debates persist regarding these dimensions' distinctiveness and incremental validity. This study applied MIRT to evaluate the instrument’s between-item multidimensionality and the functioning of its 11-point rating scale using Category Boundary Discriminations (CBDs). Data were drawn from a culturally diverse sample (N = 2,337), including several previously published datasets. Results indicated that a Rasch-compliant model did not fit, raising questions about mean or sum scores. The intended five-factor PERMA structure was not replicated due to unmodeled latent factor associations that inflated model residuals. These unmodeled associations likely reflect convergence difficulties, potentially driven by high inter-item and inter-factor correlations. Stochastic dependencies within Meaning emerged (particularly regarding item M1), which could have inflated previously reported reliability and validity indices. Participants predominantly used the upper end of the scale, with categories 5 through 10 accounting for 84% of responses, suggesting redundancy. Categories 0 to 3 therefore showed minimal incremental discrimination and unordered thresholds. Merging lower categories or using time-framed or percentage labels could improve measurement precision and discrimination.