GOODNESS-OF-FIT TESTS FOR GEE MODEL
METHODS AND APPLICATIONS
Versandkostenfrei!
Versandfertig in 6-10 Tagen
32,99 €
inkl. MwSt.
PAYBACK Punkte
16 °P sammeln!
During the last decades, a vertical exploration of statistical methodology in the ground of longitudinal data has been observed. One of the most exciting developments in this area is the Generalized Estimating Equation (GEE). This is a widely used approach that does not require the complete specification of the joint distribution of repeated measurements. In addition, GEE takes into account which results in attaining more efficiency in estimating parameters of marginal models. The model used in this study was developed by Barnhart & Williamson (1998). To test the goodness-of-fit we have applie...
During the last decades, a vertical exploration of statistical methodology in the ground of longitudinal data has been observed. One of the most exciting developments in this area is the Generalized Estimating Equation (GEE). This is a widely used approach that does not require the complete specification of the joint distribution of repeated measurements. In addition, GEE takes into account which results in attaining more efficiency in estimating parameters of marginal models. The model used in this study was developed by Barnhart & Williamson (1998). To test the goodness-of-fit we have applied model based and empirically corrected tests according to their implication. In their suggested GEE approach the correlation between two responses was not considered. Here an alternative procedure is proposed based on GEE where the correlation between two responses was considered. This study has also indicated that the model with only main effects did not fit the data well. There is a significant region, time and interaction effect. The identity correlation structure provides most efficient estimates to represent the relationship among the covariates and the responses.
Andere Kunden interessierten sich für
