Ignoring Non-ignorable Missingness

Sophia Rabe-Hesketh; Anders Skrondal

doi:10.1007/s11336-022-09895-1

Ignoring Non-ignorable Missingness

Psychometrika. 2023 Mar;88(1):31-50. doi: 10.1007/s11336-022-09895-1. Epub 2022 Dec 20.

Authors

Sophia Rabe-Hesketh¹, Anders Skrondal^{2

3

4}

Affiliations

¹ University of California, Berkeley, 2121 Berkeley Way, Berkeley, CA, 94720, USA. sophiarh@berkeley.edu.
² Norwegian Institute of Public Health, 2121 Berkeley Way, Berkeley, CA, 94720, USA.
³ University of Oslo, Oslo, Norway.
⁴ University of California, Berkeley, 2121 Berkeley Way, Berkeley, CA, 94720, USA.

Abstract

The classical missing at random (MAR) assumption, as defined by Rubin (Biometrika 63:581-592, 1976), is often not required for valid inference ignoring the missingness process. Neither are other assumptions sometimes believed to be necessary that result from misunderstandings of MAR. We discuss three strategies that allow us to use standard estimators (i.e., ignore missingness) in cases where missingness is usually considered to be non-ignorable: (1) conditioning on variables, (2) discarding more data, and (3) being protective of parameters.

Keywords: MAR; data deletion; m-graph; make-MAR; missing data; ordered factorization; protective estimation.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Models, Statistical*
Psychometrics