Modeling 2-alternative forced-choice tasks: Accounting for both magnitude and difference effects

Area of Science:

• Cognitive psychology
• Computational neuroscience
• Decision science

Background:

• Two-alternative forced-choice (2AFC) tasks are common for studying decision-making.
• Stimuli in 2AFC tasks can vary in difference and magnitude, influencing performance.
• Standard decision models struggle to explain effects of stimulus magnitude on accuracy and reaction time (RT).

Purpose of the Study:

• To develop and test a model-based analysis of 2AFC tasks with two side-by-side stimuli.
• To investigate how stimulus difference and magnitude affect decision-making processes.
• To evaluate the performance of the diffusion decision model and leaky competing accumulator model.

Main Methods:

• Model-based analysis of 2AFC tasks where stimuli vary in difference and magnitude.
• Fitting the standard diffusion decision model with modifications to variability parameters.
• Comparing diffusion model fits with the leaky competing accumulator model.

Main Results:

• The standard diffusion model with only drift rate variation failed to fit the data.
• Models allowing stimulus magnitude to influence drift variability or response threshold variability successfully fit the data.
• Both modified diffusion models explained accuracy and RT distributions, with one showing theoretical preference.

Conclusions:

• Decision-making in 2AFC tasks is influenced by both stimulus difference and magnitude.
• Modified diffusion models, particularly one with across-trial drift variability, can accurately capture these effects.
• The diffusion model provides a better fit than the leaky competing accumulator model for this task type.