Abstract
Social media has intensified parasocial relationships, one-sided bonds between individuals and media figures. While extensively researched in Western populations, parasocial engagement remains unexplored in culturally diverse contexts, particularly Latin America. The Multidimensional Measure of Parasocial Relationships (MMPR) is a validated tool that captures four dimensions of parasocial relationship engagement: Affective, Cognitive, Behavioral, and Decisional. Cross-cultural validation of parasocial relationship measures is essential for understanding how cultural context shape media psychology phenomena. We aimed to validate the Brazilian Portuguese adaptation of the MMPR, with four primary objectives: (1) replicate the correlated bifactor structure from previous research, (2) evaluate internal consistency and dimensional performance in the Brazilian cultural context, (3) analyze the intercorrelations among those dimensions, and (4) examine measurement invariance across gender. Brazilian participants (N = 398; M(age) = 24.66, SD = 9.24) completed the 18-item Brazilian MMPR. We conducted a Confirmatory Factor Analysis (CFA) using the Weighted Least Square Mean and Variance method to test the correlated bifactor model. Internal consistency was evaluated using Cronbach's Alpha (α) and McDonald's Omega (ω), with Spearman correlations examining inter-dimensional relationships. Additionally, multi-group confirmatory factor analysis was conducted to assess measurement invariance across gender. CFA supported the correlated bifactor structure with acceptable fit indices (χ² = 190.89, RMSEA = 0.08, CFI = 0.91). Standardized loadings ranged from - 0.40 to 0.79. The total scale demonstrated high internal consistency (α/ω = 0.88/0.87), while coefficients varied across dimensions (Affective: α/ω = 0.64/0.66; Cognitive: α/ω = 0.75/0.75; Behavioral: α/ω = 0.58/0.58; Decisional: α/ω = 0.82/0.81). The Decisional dimension demonstrated the strongest correlation with overall parasocial engagement (ρ = 0.86), whereas the Behavioral dimension showed the weakest (ρ = 0.65). Measurement invariance testing showed acceptable model fit for women, consistent with the full sample; however, convergence issues for the male subsample prevented further assessment of configural invariance. The Brazilian adaptation of the MMPR replicated the correlated bifactor structure and demonstrated strong overall reliability, supporting its uses as a valid measure of parasocial engagement in Brazilian populations. However, variability in the reliability of individual dimensions, particularly the strong contribution of the Decisional dimension and the weaker performance of the Behavioral dimension, suggest culturally specific patterns in parasocial engagement. Additionally, partial evidence for gender invariance suggests the scale performs more consistently among women than men, warranting further investigation. These findings highlight the need for culturally sensitive psychometric adaptations and provide a foundation for future cross-cultural research in consumer behavior, social media psychology, and therapeutic applications.