Fast and slow errors: Logistic regression to identify patterns in accuracy-response time relationships.

Area of Science:

• Cognitive Psychology
• Psychometrics
• Behavioral Neuroscience

Background:

• Understanding cognitive performance relies on analyzing error and response time (RT) patterns.
• Distributional analyses, like conditional accuracy functions (CAFs), offer deeper insights than mean statistics.
• Existing CAF methods face issues with averaging accuracy within RT bins, potentially obscuring true patterns.

Purpose of the Study:

• To present a novel method for analyzing accuracy-RT relationships using nonlinear logistic regression.
• To address the limitations of traditional RT binning in CAF computation.
• To demonstrate the utility of this new method in describing common behavioral patterns.

Main Methods:

• Developed a nonlinear logistic regression model for CAF analysis.
• Evaluated the parametric robustness of the logistic regression CAF via parameter recovery.
• Applied the novel CAF method to three existing datasets.

Main Results:

• The logistic regression CAF effectively handles issues related to RT binning.
• Specific parametric changes in the logistic regression CAF consistently describe known behavioral patterns.
• Demonstrated the method's ability to identify response capture, time pressure effects, and speed-accuracy trade-offs.

Conclusions:

• The proposed logistic regression approach offers a more robust and nuanced analysis of accuracy-RT relationships.
• This method provides a valuable tool for cognitive psychology research, enhancing the understanding of cognitive performance.
• Future research can explore modifications to further refine this analytical technique.