saws-package: Small-Sample Adjustments for Wald tests Using Sandwich...

In saws: Small-Sample Adjustments for Wald Tests Using Sandwich Estimators

Small-Sample Adjustments for Wald tests Using Sandwich Estimators

Description

Tests coefficients with sandwich estimator of variance and with small samples. Regression types supported are gee, linear regression, and conditional logistic regression.

Details

Package:	saws
Type:	Package
Version:	0.9-7.0
Date:	2022-06-23
License:	GPL-2 or greater

The main function of this package is saws, which takes output from some regression models (gee, linear regression, conditional logistic regression) and gives inferences (confidence intervals, p-values) using small sample adjusted sandwich estimators of variance. Using methods described in Fay and Graubard (2001).

The output from the regression model must be a list including the following three elements: The 'coefficients' is a vector with p parameter estimates, and is a standard output from the regression. The matrix 'u' is K by p with u[i,] the ith estimating equation, where there are K approximately independent estimating equations. The array 'omega' is K by p by p where omega[i,,] is a p by p matrix estimating - du/dbeta (here beta=coefficients). See Fay and Graubard (2001) for details.

Since the 'u' matrix and 'omega' array are not normally part of standard output, there are three specialized functions for creating regression output for use in the saws function: mgee (for gee), lmfitSaws (for linear models) and clogistCalc (for conditional logistic regression). For example, the function mgee does a gee analysis using the gee function of the gee package, takes the output and runs it through the geeUOmega function to create the 'u' matrix and the 'omega' array, and adds those onto the output from the gee (in the process other output from the gee may be corrected, see geeUOmega).

The cox regression function is not included in this version of the package. Unless there is demand (and I have time) it will not be included in a future version. There is a demo recreating the example in Fay and Graubard (2001).

Author(s)

M.P. Fay

Maintainer: Michael Fay <[e-mail]>

References

Fay and Graubard (2001). Small-Sample Adjustments for Wald-Type Tests Using Sandwich Estimators. Biometrics 57: 1198-1206. (for copy see /inst/doc/ directory

Examples

library(gee)
data(warpbreaks)
g<-gee(breaks~tension,id=wool, data=warpbreaks, corstr="exchangeable")
guo<-geeUOmega(g)
saws(guo)

saws documentation built on June 24, 2022, 1:07 a.m.