Research Article

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

Figure 7

Example of data collection cost minimization that shows the number of observations on primary and secondary responses needed to obtain a fixed BMSE in the MMSE prediction of the expected primary response , for different values of the between-subjects correlation between primary and secondary response parameters. The solid curve represents the number of measurements to collect from the primary response, , and the dashed curve represents the number of measurements to collect from the secondary response, . In this example, for , no data is to be collected from the secondary response, and the number of data points to collect from the primary response is obtained from (50). For , the number of secondary observations increases and the number of primary observations decreases with increasing correlation, as specified by and in (48). For , observations are to be collected only from the secondary response, the number of which is given by (51). The equation numbers are indicated near the relevant pieces of the curves.