Automatic Change Detection of Human Attractiveness: Comparing Visual and Auditory Perception

TL;DR

This study explores how people automatically detect changes in facial and vocal attractiveness, finding that unattractive cues are detected more strongly than attractive ones.

Contribution

The study introduces a novel ERP-based approach to compare automatic processing of facial and vocal attractiveness.

Findings

Mismatch negativities (MMNs) were elicited by both high- and low-attractive faces and voices.

Low-attractive voices induced larger MMNs than high-attractive ones, while low-attractive faces induced larger P3 amplitudes.

Results suggest a negativity bias in attractiveness processing, with stronger responses to unattractive stimuli.

Abstract

Background/Objectives: Change detection of social cues across individuals plays an important role in human interaction. Methods: Here we investigated the automatic change detection of facial and vocal attractiveness in 19 female participants by recording event-related potentials (ERPs). We adopted a ‘deviant-standard-reverse’ oddball paradigm where high- or low-attractive items were embedded as deviants in a sequence of opposite attractive standard stimuli. Results: Both high- and low-attractive faces and voices elicited mismatch negativities (MMNs). Furthermore, low-attractive versus high-attractive items induced larger mismatch negativities in the voice condition but larger P3 amplitudes in the face condition. Conclusions: These data indicate that attractiveness can be automatically detected but that differences exist between facial and vocal attractiveness processing. Generally, change detection seems to work better for unattractive than attractive information, possibly in line with a negativity bias.

Figures