Figure 7
From: Drift-diffusion explains response variability and capacity for tracking objects
Two alternative drift-diffusion models can be used for TTC of two balls. (a) Schematic of the TTC model, when the participant considers a single decision variable for both balls. Here, the participant estimates the TTC of the first ball when the decision variable hits the first decision threshold and TTC of the second ball when the same decision variable now hits the second decision threshold. (b) Schematic of the drift-diffusion model for two objects, when the participant considers a separate decision variable for each ball. Here, the participant estimates TTC of the first ball when the first decision variable hits the first decision threshold and TTC of the second ball when the second decision variable hits the second decision threshold. (c) TTC estimation errors for the second ball versus estimation errors for the first ball. (d) σ values for the second ball versus σ values for the first ball in the two-ball experiment. Each dot belongs to one participant in a given condition. Number of dots = conditions × participants = 15 × 18. (e) The variance of difference in TTC estimate for ball one and ball two in two ball experiment is plotted along with the sum of variances and subtraction of variances of two TTC estimates. (f) Fitted λ values for each subject in the two-ball experiment. Each blue dot shows the average of λs for one participants in all fifteen conditions (n = 18). The blue histograms at the right part of the panel show the distribution of fitted λ values for a 100,000 simulated samples of two completely dependent DDMs (upper histogram), and for a 100,000 simulated samples of two completely independent DDMs (lower histogram).
