Figure 9.

Utility of including boundary adaptation compared with drift-only model. A, Relative error of simulated compared with observed RT, accuracy, and posterror slowing on probe trial measures based on 20 simulated datasets for the drift-only and drift and bound adaptive models. Posterror slowing in each context condition as predicted by the (B) drift-only and (C) drift and bound models. For comparison with patterns observed in empirical data, see Figure 3C. Error bars indicate the 95% CI around the mean.