Figure 2.

Perceptual decision model schematic. A logistic regression was used to model location discrimination choices based on four predictors (and their respective fitted beta coefficients): (1) current stimulus location (β_stim), (2) stimulus location history (β_{prior_stim}), (3) choice history (β_{prior_choice}) and (4) baseline bias (β₀). Black stems on the stimulus and choice axes mark stimulus (S_i) and choice (C_i) at time step i for an example trial, with test stimulus at t (black dashed box), and prior stimuli and choices q steps back (for example, priors biased to the right, green dashed boxes). The trial began with a central fixation-point (F.P.). The sum of the product of the predictors with their respective coefficients (z) is passed through a logistic function to yield the probability of making a rightward choice in response to the test stimulus.