Abstract
The "happy-face advantage" (HFA)-faster categorization of happy versus 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.