Yinghao-Pan/ODS: Statistical Methods for Outcome-Dependent Sampling Designs

Outcome-dependent sampling (ODS) schemes are cost-effective ways to enhance study efficiency. In ODS designs, one observes the exposure/covariates with a probability that depends on the outcome variable. Popular ODS designs include case-control for binary outcome, case-cohort for time-to-event outcome, and continuous outcome ODS design (Zhou et al. 2002). Because ODS data has biased sampling nature, standard statistical analysis such as linear regression will lead to biases estimates of the population parameters. This package implements four statistical methods related to ODS designs: (1) An empirical likelihood method analyzing the primary continuous outcome with respect to exposure variables in continuous ODS design (Zhou et al., 2002). (2) A partial linear model analyzing the primary outcome in continuous ODS design (Zhou, Qin and Longnecker, 2011). (3) Analyze a secondary outcome in continuous ODS design (Pan et al. 2018). (4) An estimated likelihood method analyzing a secondary outcome in case-cohort data (Pan et al. 2017).

Getting started

Package details

AuthorYinghao Pan [aut, cre], Haibo Zhou [aut], Mark Weaver [aut], Guoyou Qin [aut], Jianwen Cai [aut]
MaintainerYinghao Pan <[email protected]>
LicenseGPL (>= 2)
Version0.2.0
URL https://github.com/Yinghao-Pan/ODS
Package repositoryView on GitHub
Installation Install the latest version of this package by entering the following in R:
install.packages("devtools")
library(devtools)
install_github("Yinghao-Pan/ODS")
Yinghao-Pan/ODS documentation built on Nov. 11, 2018, 4:23 p.m.