GEE.var.fg: A Modified GEE Variance Estimator Proposed by Fay and...
In geesmv: Modified Variance Estimators for Generalized Estimating Equations

Description Usage Arguments Details Value Author(s) References Examples

Calculate the modified GEE variance estimator proposed by Fay and Graubard (2001).

1	GEE.var.fg(formula,id,family=gaussian,data,corstr="independence",b=0.75)

`formula`	an object of class "formula" (or one that can be coerced to that class): a formula expression as for other regression models to be fitted, of the form response ~ predictors. The details of formula specification can be seen in glm() and gee().
`id`	a vector which identifies the clusters. The length of id should be the same as the total number of observations. Data is assumed to be sorted so that observations on a cluster are contiguous rows for all entities in the formula.
`family`	a family object: a list of functions and expressions for defining link and variance functions to be used in the model. This can be a character string naming a family function including "gaussian", "binomial", and "poisson". See family for details of family functions in glm() and gee().
`data`	an optional data frame, list or environment (or object coercible by as.data.frame to a data frame) containing the variables in the model. If not found in data, the variables are taken from environment(formula), typically the environment from which glm is called.
`corstr`	a character string specifying working correlation structure: "independence", "AR-M","exchangeable", "unstructured" are possible.
`b`	adjustment factor in the proposed variance estimator, which is prespecfied subjectively to avoid extreme adjustments when the cluster size is close to 1. The default value is 0.75.

Fit the model by GEE and calculate the variance estimator proposed by Fay and Graubard (2001).

`cov.beta`	estimate of the variance for beta
`cov.var`	estimate of the variance-covariance matrix for the variance estimator

Ming Wang <mwang@phs.psu.edu>

Maintainer: Zheng Li <zheng.li@outlook.com>

Fay MP and Graubard BI. Small-sample adjustments for Wald-type tests using sandwich estimators. Biometrics 2001;57: 1198-1206.

data(dental)
data_alt <- reshape(dental, direction="long", timevar="Time", 
          varying=names(dental)[3:6], v.names="response", times=c(8,10,12,14))
data_alt <- data_alt[order(data_alt$subject),]
data_alt$gender <- as.numeric(data_alt$gender)
data_alt$Time <- sqrt(data_alt$Time)
formula <- response~Time+gender
fg.ind <- GEE.var.fg(formula,id="subject",family=gaussian,
       data_alt,corstr="independence") ##Independence correlation structure;
fg.exch <- GEE.var.fg(formula,id="subject",family=gaussian,
        data_alt,corstr="exchangeable") ##Exchangeable correlation structure;
fg.ar1 <- GEE.var.fg(formula,id="subject",family=gaussian,
        data_alt,corstr="AR-M") ##AR-1 correlation structure;
fg.unstr <- GEE.var.fg(formula,id="subject",family=gaussian,
       data_alt,corstr="unstructured") ##Unstructured correlation structure;