QIC.gee: QIC and QICu for GEE

In wgeesel: Weighted Generalized Estimating Equations and Model Selection

Description

Calculate quasi-likelihood under the independence model criterion (QIC) and QICu based on GEE.

Usage

QIC.gee(object)

Arguments

object

a fitted model object of class "wgee".

Value

Return a vector of QIC, QICu and Quasi-likelihood.

Note

QIC can be used to select the best correlation structure and the best fitting model in GEE analyses. The GEE is fitted by geeglm (geepack). QICu is a simplified version of QIC, which can not be applied to select the optimal working correlation structure. geeglm (geepack) only works for complete data. Thus if there are NA’s in data, the missing values are automatically removed by na.omit.

Author(s)

Cong Xu, Zheng Li and Ming Wang

References

Liang, K.Y. and Zeger, S.L., 1986. Longitudinal data analysis using generalized linear models. Biometrika, pp.13-22.

Pan, W., 2001. Akaike's information criterion in generalized estimating equations. Biometrics, 57(1), pp.120-125.

Prentice, R.L. and Zhao, L.P., 1991. Estimating equations for parameters in means and covariances of multivariate discrete and continuous responses. Biometrics, pp.825-839.

See Also

geeglm (geepack). MuMIn also provides QIC value.

Examples

data(imps)
fit <- wgee(Y ~ Drug+Sex+Time, data=imps, id=imps$ID, family="binomial",
            corstr="exchangeable")

QIC.gee(fit)

data(seizure)
###reshapre the seizure data to "long" format
seiz.long <- reshape(seizure, varying=list(c("base","y1", "y2", "y3", "y4")),
                      v.names="y", times=0:4, direction="long")
seiz.long <- seiz.long[order(seiz.long$id, seiz.long$time),]

fit <- wgee(y ~ age + trt + time, data=seiz.long, id=seiz.long$id, 
            family="poisson", corstr="independence")

QIC.gee(fit)

Example output

     QIC   QICu Quasi_lik
1 1386.4 1380.5    -686.2
     QIC     QICu Quasi_lik
1 -12933 -12945.9    6476.9

wgeesel documentation built on May 2, 2019, 3:41 a.m.