Probabilistic learning of emotion categories.

Area of Science:

• Cognitive Psychology
• Developmental Psychology
• Social Neuroscience

Background:

• Human emotion perception often categorizes continuous facial muscle configurations.
• Categorical perception, seen in speech and color, suggests attunement to environmental thresholds.
• Learning processes for emotion cue representation remain largely unknown.

Purpose of the Study:

• Investigate the role of statistical distributions of facial cues in emotion categorization.
• Examine how learning these distributions affects emotion category boundaries in children and adults.
• Determine if learners can track distributions for multiple actors and how this influences social judgments.

Main Methods:

• Experiment 1: Supervised learning of neutral-angry boundary with varying facial cue distributions in children and adults.
• Experiments 2 & 3: Replication and investigation of actor-specific category formation and influence on trait judgments.

Main Results:

• Participants rapidly shifted emotion category boundaries when exposed to different statistical distributions of facial cues.
• Learners formed actor-specific emotion categories.
• Facial cue distributions influenced participants' trait judgments about actors.

Conclusions:

• Human emotion construal is flexible and shaped by learning the statistical distributions of social cues.
• Environmental statistical regularities in facial expressions are critical for adaptive emotion perception.
• This learning extends to forming actor-specific social impressions.