Multiple Imputation for Nonresponse in Complex Surveys Using Design Weights and Auxiliary Margins

TL;DR

This paper develops an advanced multiple imputation method for survey data with missing values, incorporating design weights and known auxiliary margins to improve the accuracy of nonresponse adjustments.

Contribution

It introduces a novel approach that integrates design weights into multiple imputation for complex surveys with nonresponse, enhancing prior methods that used fabricated weights.

Findings

Simulation studies demonstrate improved imputation accuracy.

Method effectively utilizes known auxiliary margins.

Enhanced estimates for variables with known distributions.

Abstract

Survey data typically have missing values due to unit and item nonresponse. Sometimes, survey organizations know the marginal distributions of certain categorical variables in the survey. As shown in previous work, survey organizations can leverage these distributions in multiple imputation for nonignorable unit nonresponse, generating imputations that result in plausible completed-data estimates for the variables with known margins. However, this prior work does not use the design weights for unit nonrespondents; rather, it relies on a set of fabricated weights for these units. We extend this previous work to utilize the design weights for all sampled units. We illustrate the approach using simulation studies.