2019/03/21 by John C. Galati, Galati, John C.
Computer Science · Mathematics · #Bayesian Modeling and Causal Inference #FOS: Computer and information sciences #Other Statistics (stat.OT) #Statistical Methods and Bayesian Inference #Statistical Methods and Inference
paper · pdf · doi:10.48550/arxiv.1903.08880
openalex publication_date 2019/03/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
We identify three issues permeating the literature on statistical methodology\nfor incomplete data written for non-specialist statisticians and other\ninvestigators. The first is a mathematical defect in the notation Yobs, Ymis\nused to partition the data into observed and missing components. The second are\nissues concerning the notation `P(R|Yobs, Ymis)=P(R|Yobs)' used for\ncommunicating the definition of missing at random (MAR). And the third is the\nframing of ignorability by emulating complete-data methods exactly, rather than\ntreating the question of ignorability on its own merits. These issues have been\npresent in the literature for a long time, and have simple remedies. The\npurpose of this paper is to raise awareness of these issues, and to explain how\nthey can be remedied.\n