Testing the factor structure underlying behavior using joint cognitive models: Impulsivity in delay discounting and...

Area of Science:

• Cognitive Neuroscience
• Behavioral Economics
• Psychometrics

Background:

• Cognitive models quantify decision-making processes but comparing models across tasks is challenging.
• Existing methods struggle to determine if different tasks measure the same underlying cognitive traits.
• Understanding shared versus distinct cognitive processes is crucial for accurate assessment.

Purpose of the Study:

• Develop a method to model behavior across multiple tasks by linking cognitive model parameters to common latent constructs.
• Evaluate whether two distinct tasks measure the same cognitive dimensions.
• Improve cognitive model estimation by accounting for overlapping processes.

Main Methods:

• Developed a novel approach for cross-task behavioral modeling.
• Applied the method to data from the delay discounting task and the Cambridge gambling task.
• Connected cognitive model parameters to latent constructs to assess shared dimensions of impulsivity.

Main Results:

• Discounting rate parameters from the two tasks were not closely related.
• Substance users exhibited more impulsive behavior on both tasks.
• Delay discounting was linked to externalizing psychopathology (aggression), while Cambridge gambling task performance related to response inhibition failures.

Conclusions:

• The developed method effectively connects behavior across tasks, offering new insights into cognitive processes.
• Impulsivity is a multi-dimensional construct, with different tasks capturing distinct facets.
• Findings highlight the differential relationships between specific decision-making deficits, impulsivity dimensions, and substance use.