dynConfiR: An R package for sequential sampling models of decision confidence

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Psychometrics

Background:

• Sequential sampling models are foundational for understanding decision-making.
• Perceptual decisions involve interconnected choices, confidence judgments, and reaction times.
• Simultaneously modeling these variables offers a more comprehensive understanding of decision processes.

Purpose of the Study:

• To introduce dynConfiR, an R package implementing various sequential sampling models.
• To provide tools for fitting parameters, predicting outcomes, and simulating data for decision-making research.
• To facilitate the analysis of choice, response time, and decision confidence.

Main Methods:

• Implementation of multiple sequential sampling models in R.
• Development of probability density functions and high-level fitting functions.
• Step-by-step workflow for data preprocessing, model fitting, prediction, comparison, and assessment.

Main Results:

• The dynConfiR package offers robust computations for probability density calculations.
• Parameter and model recovery analyses demonstrate the reliability of the implemented models.
• The package provides intuitive usability and high flexibility for researchers.

Conclusions:

• dynConfiR enhances the modeling of decision-making by integrating choice, response time, and confidence.
• The package supports researchers in analyzing empirical data and advancing the understanding of decision processes.
• It lowers the technical barrier for applying complex computational models in cognitive science research.