Breadcrumbs Section. Click here to navigate to respective pages.
Chapter
Chapter
If we had fitted the mixed model to the full set of data the confidence intervals for µ would have been (−0.0678,0.2482) for log(AUC) and (−0.0907,0.1843) for log(Cmax). No analysis is complete until the assumptions that have been made in the modelling have been checked. As in Chapter 2, we can check the normality of the studentized residuals by plotting histograms and nor-mal probability plots. The histograms of the studentized residuals and normal probability plots for log(AUC) and log(Cmax) are given in Fig-ure 7.4. We note that in a two-period cross-over trial the two residuals for each subject are equal in magnitude and of opposite signs. Hence the plots in Figure 7.4 are only for the residuals from the first period (we could, of course, have chosen the second period). There is clear evi-dence that the residuals for log(AUC) are not normally distributed and some evidence that the residuals from log(Cmax) are not normally dis-
DOI link for If we had fitted the mixed model to the full set of data the confidence intervals for µ would have been (−0.0678,0.2482) for log(AUC) and (−0.0907,0.1843) for log(Cmax). No analysis is complete until the assumptions that have been made in the modelling have been checked. As in Chapter 2, we can check the normality of the studentized residuals by plotting histograms and nor-mal probability plots. The histograms of the studentized residuals and normal probability plots for log(AUC) and log(Cmax) are given in Fig-ure 7.4. We note that in a two-period cross-over trial the two residuals for each subject are equal in magnitude and of opposite signs. Hence the plots in Figure 7.4 are only for the residuals from the first period (we could, of course, have chosen the second period). There is clear evi-dence that the residuals for log(AUC) are not normally distributed and some evidence that the residuals from log(Cmax) are not normally dis-
If we had fitted the mixed model to the full set of data the confidence intervals for µ would have been (−0.0678,0.2482) for log(AUC) and (−0.0907,0.1843) for log(Cmax). No analysis is complete until the assumptions that have been made in the modelling have been checked. As in Chapter 2, we can check the normality of the studentized residuals by plotting histograms and nor-mal probability plots. The histograms of the studentized residuals and normal probability plots for log(AUC) and log(Cmax) are given in Fig-ure 7.4. We note that in a two-period cross-over trial the two residuals for each subject are equal in magnitude and of opposite signs. Hence the plots in Figure 7.4 are only for the residuals from the first period (we could, of course, have chosen the second period). There is clear evi-dence that the residuals for log(AUC) are not normally distributed and some evidence that the residuals from log(Cmax) are not normally dis-
ABSTRACT