Joint Cognitive Models Reveal Sources of Robust Individual Differences in Conflict Processing

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Psychometrics

Background:

• Conflict tasks (e.g., Stroop, Flanker, Simon) show poorer performance in incongruent conditions.
• Debate exists on whether individual differences in conflict tasks reflect reliable, trait-like mechanisms.
• Previous methods like difference scores have poor psychometric properties, and separate modeling fails to account for measurement error.

Purpose of the Study:

• To simultaneously model distinct mechanisms of conflict task performance and their covariance.
• To improve individual estimation and account for measurement error using hierarchical Bayesian joint modeling.
• To investigate the reliability and generalizability of conflict-specific versus task-general mechanisms.

Main Methods:

• Applied the conflict linear ballistic accumulator model (LBA) to large datasets with multiple conflict tasks and test-retest sessions.
• Utilized hierarchical Bayesian joint modeling to estimate cognitive process parameters and their covariation simultaneously.
• Included a dataset with a conflict task and a simple perceptual decision-making task to assess task-general mechanisms.

Main Results:

• Found moderate test-retest reliability for both conflict-specific and task-general mechanisms within conflict tasks.
• Task-general mechanisms, but not conflict-specific ones, showed correlations across different conflict tasks.
• Task-general mechanisms were also correlated between conflict tasks and a simple decision-making task.

Conclusions:

• Robust individual differences exist in computational mechanisms underlying general decision-making.
• Evidence does not support reliable individual differences in mechanisms specifically for conflict processing.
• Findings suggest that general cognitive abilities, rather than conflict-specific ones, drive individual differences in performance on conflict tasks.