Swgee: An R Package for Analyzing Longitudinal Data with Response Missingness and Covariate Measurement Error

Juan Xiong,Grace,Y Yi

doi:10.32614/rj-2019-031

Abstract

swgee: An R Package for Analyzing Longitudinal Data with Response Missingness and Covariate Measurement Error

Highlights

Longitudinal studies are commonly conducted in the health sciences, biochemical, and epidemiology fields; these studies typically collect repeated measurements on the same subject over time
It has been well documented that ignoring missing responses and covariate measurement error may lead to severely biased results, leading to invalid inferences (Fuller, 1987; Carroll et al, 2006)
Yi (2008) proposed an estimation method based on the marginal model for the response process, which does not require the full specification of the distribution of the response variable but models only the mean and variance structures

Summary

Introduction

Longitudinal studies are commonly conducted in the health sciences, biochemical, and epidemiology fields; these studies typically collect repeated measurements on the same subject over time. Missing observations and covariate measurement error frequently arise in longitudinal studies and they present considerable challenges in statistical inference about such data (Carroll et al, 2006; Yi, 2008). It has been well documented that ignoring missing responses and covariate measurement error may lead to severely biased results, leading to invalid inferences (Fuller, 1987; Carroll et al, 2006). A functional method is applied to relax the need of modeling the covariate process. These features make the method of Yi (2008) flexible for many applications

Objectives

Methods

Findings

Discussion

Conclusion