QICW.gee: QICWr and QICWp for WGEE
In wgeesel: Weighted Generalized Estimating Equations and Model Selection

Description Usage Arguments Value Note Author(s) References See Also Examples

Calculate the QICWr and QICWp (an information criterion based on the weighted quasi-likelihood function) for selection of mean model and correlation structure based on the WGEE.

1	QICW.gee(object)

object

a fitted model object of class "wgee".

Return a data frame of QICWr, QICWp and Wquasi_lik.

QICWr can be used for variable selection and for selecting the correlation structure in WGEE analyses. QICWp is a simplified version of QICWr, which can not be applied to select the optimal working correlation structure in WGEE.

Cong Xu, Zheng Li and Ming Wang

Gosho, M., 2015. Model selection in the weighted generalized estimating equations for longitudinal data with dropout. Biometrical Journal.

Platt, R.W., Brookhart, M.A., Cole, S.R., Westreich, D. and Schisterman, E.F., 2013. An information criterion for marginal structural models. Statistics in medicine, 32(8), pp.1383-1393.

Robins, J.M., Rotnitzky, A. and Zhao, L.P., 1995. Analysis of semiparametric regression models for repeated outcomes in the presence of missing data. Journal of the American Statistical Association, 90(429), pp.106-121.

wgee

data(imps)

### variable selection by QICWr, not rum###
#fit <- wgee(IMPS79 ~ Drug+Sex+Time, mismodel= R ~ Drug+Time, data=imps,
##           id=imps$ID, family="gaussian", corstr="exchangeable")
##QICW.gee(fit)

#fit <- wgee(IMPS79 ~ Drug+Sex+Time+Time:Sex+Time:Drug, mismodel= R ~ Drug+Time,
#        data=imps, id=imps$ID, family="gaussian", corstr="exchangeable")
##QICW.gee(fit)