Face evaluation: Findings, methods, and challenges.

Area of Science:

• Psychology
• Cognitive Science
• Computer Science

Background:

• Facial appearance judgments are rapid, consequential, and consistent across time and cultures.
• Trustworthiness judgments correlate with general facial valence and resemblance to emotional expressions, signaling approach/avoidance behaviors.

Purpose of the Study:

• To review key findings and methods in face evaluation research over the past two decades.
• To highlight the limitations of aggregated models and advocate for individual-level judgment modeling.

Main Methods:

• Review of existing literature on face evaluation.
• Discussion of data-driven computational models for feature configuration discovery.
• Introduction of a novel method combining established techniques with machine learning for individual participant modeling.

Main Results:

• Evaluative judgments, particularly trustworthiness, emerge early and are remarkably consistent.
• Computational models have identified facial features driving judgments, often linked to emotional expressions.
• Most variance in complex judgments is idiosyncratic, stemming from individual differences and participant-stimulus interactions.

Conclusions:

• Understanding complex facial judgments requires modeling individual differences.
• Novel methods integrating machine learning are crucial for capturing idiosyncratic judgment variations.
• Future research should focus on developing and applying individual-level computational models.