Prediction Accuracy in Multivariate Repeated-Measures Bayesian Forecasting Models with Examples Drawn from Research on Sleep and Circadian Rhythms

Figure 5

MMSE predictions of the expected primary response from the bivariate random intercept model specified in (23), and corresponding 95% confidence intervals (shaded areas) for individual #19. The MMSE estimator was iteratively determined assuming only primary responses were observed, as well as assuming both primary and secondary responses were observed. For the former case, the MMSE was iteratively determined by incorporating observations ; for the latter case, the MMSE was iteratively determined by incorporating pairs of observations . Confidence intervals were obtained from quantiles of the posterior distribution, which is defined by the posterior mean (the MMSE estimator) (31) and the posterior variance (32).