Two-stage reinforcement learning task predicts psychological traits.

Area of Science:

• Cognitive Psychology
• Behavioral Neuroscience
• Psychometrics

Background:

• External information influences actions, but internal psychological traits (PTs) are crucial for decision-making.
• PTs are stable individual characteristics shaping behavioral repertoires across contexts.

Purpose of the Study:

• To explore the relationship between adaptive behavior metrics under uncertainty and various PTs.
• To develop a multimetric behavioral profile for predicting PTs.

Main Methods:

• Participants completed a reversal-learning task with volatile contingencies.
• Behavioral profiles were derived from response sequences and linked to self-report PT questionnaires (Hierarchical Taxonomy Of Psychopathology - HiTOP).
• Multiple linear regression models (MLRMs) analyzed the relationships, with significance confirmed via random permutations.

Main Results:

• Learning curves from the task predicted significant differences in PTs and response times.
• Behavioral profile configurations served as a reliable "fingerprint" for participant identification.
• The MLRM findings demonstrated a strong predictive link between behavior and psychological traits.

Conclusions:

• Behavioral profiling offers a novel approach to understanding and quantifying psychological traits.
• This method has potential applications in refining nosological classifications within psychopathology.
• Characterizing behavior under uncertainty provides insights into internal psychological variables.