DstarM: an R package for analyzing two-choice reaction time data with the D∗M method.

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Psychometrics

Background:

• Choice reaction time (CRT) data analysis traditionally employs diffusion models to describe decision processes.
• Standard diffusion models assume non-decision time is uniformly distributed, which can bias parameter estimates.
• Accurate modeling of decision and non-decision processes is crucial for understanding cognitive tasks.

Purpose of the Study:

• To present an R package implementing the D*M method for diffusion model parameter estimation.
• To address the limitations of assuming a uniform non-decision time distribution.
• To provide a tool for unbiased estimation of decision and non-decision time parameters.

Main Methods:

• Utilized the D*M method to estimate diffusion model parameters without prior distributional assumptions for non-decision time.
• Developed an R package for practical implementation of the D*M method.
• Conducted extensive simulation studies to validate the method's performance.

Main Results:

• The D*M method successfully estimated decision model parameters.
• Non-parametric estimates of the non-decision time distribution were accurately retrieved.
• Simulation studies confirmed correct retrieval of both decision and non-decision time parameters.

Conclusions:

• The developed R package offers a robust tool for analyzing choice reaction time data using diffusion models.
• The D*M method effectively mitigates bias introduced by misspecified non-decision time distributions.
• This approach enhances the accuracy and reliability of cognitive process modeling.