Ignorability for general longitudinal data

Farewell, D. M.; Huang, C.; Didelez, V.

Abstract

Likelihood factors that can be disregarded for inference are termed ignorable. We demonstrate that close ties exist between ignorability and identification of causal effects by covariate adjustment. A graphical condition, stability, plays a role analogous to that of missingness at random, but is applicable to general longitudinal data. Our formulation of ignorability does not depend on any notion of missing data, so is appealing in situations where missing data may not actually exist. Several examples illustrate how stability may be assessed.

Citation

Farewell, D. M., Huang, C., & Didelez, V. (2017). Ignorability for general longitudinal data. Biometrika, 104(2), 317-326. https://doi.org/10.1093/biomet/asx020

• Journal Article Type: Article
• Acceptance Date: Feb 12, 2017
• Online Publication Date: May 8, 2017
• Publication Date: 2017-06
• Deposit Date: Apr 9, 2019
• Publicly Available Date: Apr 10, 2019
• Journal: Biometrika
• Print ISSN: 0006-3444
• Electronic ISSN: 1464-3510
• Publisher: Oxford University Press (OUP)
• Peer Reviewed: Peer Reviewed
• Volume: 104
• Issue: 2
• Pages: 317-326
• DOI: https://doi.org/10.1093/biomet/asx020
• Public URL: https://hull-repository.worktribe.com/output/1567943
• Publisher URL: https://academic.oup.com/biomet/article/104/2/317/3804413