Data-driven studies in face identity processing rely on the quality of the tests and data sets

Area of Science:

• Cognitive Psychology
• Neuroscience
• Human Perception

Background:

• Data-driven approaches are increasingly used to study individual differences in face identity processing (FIP).
• Existing FIP tests are often used interchangeably, raising questions about their validity, reliability, and consistency.
• The variability in FIP test performance can influence the outcomes of data-driven analyses.

Purpose of the Study:

• To investigate the underlying factors of face identity processing (FIP) using multiple common tests.
• To assess the reliability, inter-test correlations, and individual consistency across various FIP tests.
• To evaluate the suitability of current FIP tests for data-driven research.

Main Methods:

• 211 participants completed eight frequently used face identity processing (FIP) tests.
• Principal Component Analysis and Agglomerative Clustering were employed to analyze performance factors.
• Reliability, test inter-correlations, and participant consistency were quantified.

Main Results:

• Participant performance in FIP tests could be explained by two factors: confirmation and elimination of an identity match.
• Participants were categorized into clusters based on their performance profiles across these factors.
• The reliability of the FIP tests was found to be moderate at best, with weak correlations between tests and low individual consistency.

Conclusions:

• Current face identity processing (FIP) tests exhibit limitations in reliability and consistency.
• The heterogeneity of FIP tests poses challenges for data-driven research aiming to understand individual differences.
• Developing and rigorously evaluating FIP measures is crucial for advancing data-driven insights into face perception.