Figure 3.

Test-retest properties of the Bayesian state-space model. A, B, Regression of $B\left[s\right]$ onto ${\sigma }_{\eta }\left[s\right]$ (A) and ${\sigma }_{ϵ}\left[s\right]$ (B) for the simulated optimal dataset. C, D, Regression of $B\left[s\right]$ onto ${\sigma }_{\eta }\left[s\right]$ (C) and ${\sigma }_{ϵ}\left[s\right]$ (D) for the simulated permuted dataset. Parameter estimates with 68% HDIs are shown for every simulated learner as a dot with error bars. The black solid line shows the regression on the model parameters estimated with the Bayesian state-space model, the green dashed line the regression on the original model parameters.