Inference for partial correlation when data are missing not at random

TL;DR

This paper develops a method using uncertainty regions for inference on partial correlations in datasets with missing not at random, ensuring reliable coverage and demonstrated effectiveness through simulations and real data.

Contribution

It introduces uncertainty regions for partial correlation inference under non-random missing data, a novel approach in this context.

Findings

Uncertainty regions achieve desired asymptotic coverage.

Method performs well in finite samples as shown by simulations.

Real data example confirms practical applicability.

Abstract

We introduce uncertainty regions to perform inference on partial correlations when data are missing not at random. These uncertainty regions are shown to have a desired asymptotic coverage. Their finite sample performance is illustrated via simulations and real data example.