A Different Approach to the Problem of Missing Data
Methodology
2015-09-17 v1
Abstract
There is a long history of devleopment of methodology dealing with missing data in statistical analysis. Today, the most popular methods fall into two classes, Complete Cases (CC) and Multiple Imputation (MI). Another approach, Available Cases (AC), has occasionally been mentioned in the research literature, in the context of linear regression analysis, but has generally been ignored. In this paper, we revisit the AC method, showing that it can perform better than CC and MI, and we extend its breadth of application.
Cite
@article{arxiv.1509.04992,
title = {A Different Approach to the Problem of Missing Data},
author = {Xiao Gu and Norman Matloff},
journal= {arXiv preprint arXiv:1509.04992},
year = {2015}
}
Comments
Software at https://github.com/matloff/regtools