Figure 2.

Relationship between the effect size in diffusion model parameters manipulated in each scenario (x-axis) and the effect size observed in the behavioral measures derived from the simulated data (y-axis). Positive effect sizes on the y-axis indicate larger costs in the older adult group. See Table 1 and Figure 1 for parameters manipulated in each scenario. The effect sizes are nonzero for all raw and transformed costs in Scenarios B and C, and for proportional response time (RT) costs in Scenario D. This indicates that they do not control for group differences in these confounding parameters.