Abstract

The "happy-face advantage" (HFA)—faster categorization of happy vs. fearful expressions—is typically established using mean reaction times (RTs), which obscure the structure of the RT distribution. The present research examined how facial expression and social-category cues influence emotion categorization using ex-Gaussian modelling, decomposing RTs into mean (μ), variability (σ), and slow-tail responding (τ). Ex-Gaussian modelling revealed that expression effects were distributed across distinct components of the RT distribution. Whereas mean RT (μ) replicated the happy-face advantage, σ indexed systematic differences in response consistency and τ captured selective increases in unusually slow responses. These findings demonstrate that social-category influences on emotion categorization are distributed across multiple components of the RT distribution that remain hidden in conventional analyses of mean RT.