Nonstandard conditionally specified models for nonignorable missing data
Abstract
We consider data-analysis settings where data are missing not at random. In these cases, the two basic modeling approaches are 1) pattern-mixture models, with separate distributions for missing data and observed data, and 2) selection models, with a distribution for the data preobservation and a missing-data mechanism that selects which data are observed. These two modeling approaches lead to distinct factorizations of the joint distribution of the observed-data and missing-data indicators. In this paper, we explore a third approach, apparently originally proposed by J. W. Tukey as a remark in a discussion between Rubin and Hartigan, and reported by Holland in a two-page note, which has been so far neglected.
Document Details
- Document Type
- Pub Defense Publication
- Publication Date
- Jul 28, 2020
- Source ID
- 10.1073/pnas.1815563117
Entities
People
- Alexander Franks
- Donald B. Rubin
- Edoardo Airoldi
Organizations
- National Science Foundation
- Office of Naval Research
- Temple University
- Tsinghua University
- University of California, Santa Barbara