Facial first impressions form two clusters representing approach-avoidance.

Area of Science:

• Cognitive Psychology
• Social Psychology
• Computational Neuroscience

Background:

• Existing models propose 2-4 factors underlying social trait judgments from facial first impressions.
• Previous research focused on grouping traits rather than faces.

Purpose of the Study:

• To investigate the underlying structure of facial first impressions using unsupervised learning.
• To determine if faces can be clustered into distinct groups based on social trait judgments.
• To link these clusters to behavioral responses and facial cues.

Main Methods:

• Applied unsupervised learning algorithms to large facial impression rating databases.
• Utilized Bayesian modeling to extract impression profiles and separating vectors.
• Conducted behavioral experiments to assess observer responses and facial cue influence.

Main Results:

• Identified two stable clusters of faces across datasets and dimensional representations.
• The separating vector between clusters strongly correlated with valence and approachability.
• Facial appearance cues, such as smiling, were sufficient to differentiate these clusters.
• Observer responses showed a clear approach-avoidance behavior bias towards the two clusters.

Conclusions:

• Facial first impressions are organized into two fundamental clusters.
• These clusters are driven by basic approach-avoidance mechanisms, influenced by facial appearance.
• Findings provide a computational and behavioral basis for understanding social perception.