Saved in:
Bibliographic Details
Main Author: von Hippel, Paul T.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.22981
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916814975401984
author von Hippel, Paul T.
author_facet von Hippel, Paul T.
contents Predictive mean matching (PMM) is a popular imputation strategy that imputes missing values by borrowing observed values from other cases with similar expectations. We show that, unlike other imputation strategies, PMM is not guaranteed to be consistent -- and in fact can be severely biased -- when values are missing at random (when the probability a value is missing depends only on values that are observed). We demonstrate the bias in a simple situation where a complete variable $X$ is both strongly correlated with $Y$ and strongly predictive of whether $Y$ is missing. The bias in the estimated regression slope can be as large as 80 percent, and persists even when we reduce the correlation between $X$ and $Y$. To make the bias vanish, the sample must be large ($n$=1,000) \emph{and} $Y$ values must be missing independently of $X$ (i.e., missing completely at random). Compared to other imputation methods, it seems that PMM requires larger samples and is more sensitive to the pattern of missing values. We cannot recommend PMM as a default approach to imputation.
format Preprint
id arxiv_https___arxiv_org_abs_2506_22981
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Imputing With Predictive Mean Matching Can Be Severely Biased When Values Are Missing At Random
von Hippel, Paul T.
Methodology
62N02
Predictive mean matching (PMM) is a popular imputation strategy that imputes missing values by borrowing observed values from other cases with similar expectations. We show that, unlike other imputation strategies, PMM is not guaranteed to be consistent -- and in fact can be severely biased -- when values are missing at random (when the probability a value is missing depends only on values that are observed). We demonstrate the bias in a simple situation where a complete variable $X$ is both strongly correlated with $Y$ and strongly predictive of whether $Y$ is missing. The bias in the estimated regression slope can be as large as 80 percent, and persists even when we reduce the correlation between $X$ and $Y$. To make the bias vanish, the sample must be large ($n$=1,000) \emph{and} $Y$ values must be missing independently of $X$ (i.e., missing completely at random). Compared to other imputation methods, it seems that PMM requires larger samples and is more sensitive to the pattern of missing values. We cannot recommend PMM as a default approach to imputation.
title Imputing With Predictive Mean Matching Can Be Severely Biased When Values Are Missing At Random
topic Methodology
62N02
url https://arxiv.org/abs/2506.22981