The restricted maximum likelihood was used to compare mixed models, given the same mean value structure.
Of the methods shown, the mixed model gave almost unbiased treatment estimates, even when the dropout rates were different, suggesting that myth 2 may also be false.
Fitting the linear mixed model gives the following results presented in Table 3.
Fitting the linear mixed model gives the following results in Table 2.
A special case is the linear mixed model given by, Yi |Xi,Zi =Xiβ + Zibi + ei, where ei~N 0,Ri) and independent of bi.
To estimate the among-individual and residual variance for each capture, an interaction between ID and capture number (1 4) was included in the among-individual and residual covariance structure of the mixed model, giving separate estimates for among-individual and residual variance for each of the captures.
The mixed effects models gave only one significant effect of time, on the IBS-QOL (B = 4.5, t20.1 = 2.2, p =.04).
Within the linear models, the results are mixed, and all models gave similar predictions.
In order to adhere to the intention-to-treat principle, the continuous outcome variables were analysed using mixed effects models, given their ability to handle missing data.
In order to adhere to the intention-to-treat principle, the outcome variables will be analysed using mixed effects models, given their ability to handle missing data.
