Figure 5

Estimated marginal posterior probabilities for Scenario 2 datasets. The top estimated posterior probabilities were summarized across the Scenario 2 datasets. The posterior estimates were very similar across algorithms, indicating convergence. Marginally, X₁₀ had low posterior estimates, meaning it was infrequently included in the model without the interaction with X₇. Note that the whiskers and outliers were not drawn on the box plots.