A diffusion model decomposition of the practice effect.

Area of Science:

• Cognitive psychology
• Computational neuroscience
• Psycholinguistics

Background:

• Repeatedly performing cognitive tasks leads to decreased mean response times (RTs).
• The precise mathematical function describing this practice effect remains debated.
• Understanding practice effects requires examining accuracy and RT distributions for both correct and error responses.

Purpose of the Study:

• To investigate the practice effect by analyzing changes in accuracy and RT distributions.
• To decompose the practice effect into its underlying psychological processes using a diffusion model.
• To determine how practice influences information processing, response caution, bias, and peripheral timing.

Main Methods:

• Utilized the Ratcliff diffusion model, a computational tool for analyzing two-choice RT data.
• Analyzed data from a large-scale (10,000-trial) lexical decision task.
• Simultaneously modeled changes in accuracy and RT distributions across practice trials.

Main Results:

• Practice effects encompass more than just increased processing speed.
• Significant changes were observed in response caution, response bias, and peripheral processing time.
• The diffusion model successfully captured these multiple subcomponents of practice.

Conclusions:

• The practice effect is a complex phenomenon with multiple interacting subcomponents.
• Relying solely on mean RT for correct responses can obscure the full impact of practice.
• A comprehensive understanding necessitates analyzing RT distributions and accuracy alongside mean RTs.