A multidimensional normative framework for fine-grained social cognition assessment with the BCS-Greco

Current social cognition assessments often yield global summary scores that obscure clinically meaningful variability across subcomponents. This study proposes a multilevel interpretive framework for the Social Cognition Battery (Batterie de Cognition Sociale, BCS-Greco) that enables fine-grained analysis of facial emotion recognition, mental state inference, and theory of mind reasoning.

Analyses were based on partially overlapping normative samples of healthy French-speaking adults (age range: 20-69 years) who completed facial emotion recognition (FER/IDEMO; n = 155), Reading the Mind in the Eyes Test (RMET; n = 175), and theory of mind (ToM) tasks assessing first-order and second-order belief attribution as well as faux pas recognition (n = 129-145; adapted from Baron-Cohen, 2000). A data-driven stratification approach identified optimal normative groupings based on empirically observed age and sex effects. Scoring dimensions included emotion-specific accuracy, intensity-dependent performance profiles, systematic misattribution patterns (confusion matrices), response latencies, and ToM subcomponents by inferential order and cognitive dimension.

Age-related decline emerged for anger and fear recognition (p = .001) but not for joy or disgust, which showed ceiling effects. Response times revealed nonlinear age trajectories, with adults aged 50-59 demonstrating longer latencies than those over 60 (p < .001). Older adults (60+) exhibited significantly elevated misattribution rates (p = .008), whereas a descriptive tendency toward higher anger-disgust confusion in young men did not reach significance after correction for within-participant clustering. Theory of mind performance remained stable across age groups, with significant differentiation across inferential orders (p < .001) and cognitive dimensions (p < .001).

This multidimensional framework provides normative benchmarks that may enhance clinical utility by enabling identification of emotion-specific performance profiles, detection of processing efficiency variations through response latencies, and characterization of systematic error patterns; thereby extending the interpretive precision of well-validated instruments whose clinical sensitivity has been established across neurological and psychiatric populations.