Measures of discriminatory ability, correlation, bias and accuracy of Bayesian hierarchical regression models.

ROC, receiver operating characteristic, a plot of sensitivity vs. one minus specificity; values of 0.9 outstanding discriminatory performance. Correlation coefficient of 1 indicates perfect correlation between observed and expected prevalence. Mean error is the mean difference between observed and expected prevalence (assesses bias of estimates). Mean absolute error is the mean absolute difference between observed and expected prevalence (assesses accuracy of estimates).