Figure 1
Histogram of samples from our default prior distribution for affirmative voting probability.
Figure 2
Histogram of samples from our alternative prior distribution for affirmative voting probability.
Figure 3
Posterior median rank of the ideal points of legislators in the 116th House of Representatives under IDEAL (x axis) vs. those under PUM (y axis).
Figure 4
Maximum likelihood estimates of the ideal points of legislators in the 116th House of Representatives under WNOMINATE (x axis) vs. the posterior median rank of legislators’ ideal points under PUM (y axis).
Figure 5
Posterior mean (solid lines) and 95% pointwise posterior credible intervals (gray regions) for the response functions associated with selected votes in the 116th House of Representatives.
Figure 6
Posterior mean (solid lines) and 95% credible intervals (shaded regions) of the trajectories of the ideal points of selected U.S. Supreme Court Justices.
Figure 7
Posterior distributions for Justice Hugo Black’s preferences.
Figure 8
Prior and posterior distributions of the autocorrelation parameter ρ.
Figure 9
Differences in WAIC scores between a dynamic IRT model and a dynamic PUM. In this graph, ΔWAIC = WAIC(IRT)– WAIC(PUM), so that positive values of ΔWAIC indicate that PUM provides better complexity-adjusted fit to the data.