Big Data meets Causal Survey Research: Understanding Nonresponse in the Recruitment of a Mixed-mode Online Panel

TL;DR

This paper introduces the double machine learning approach to address nonresponse bias in high-dimensional survey data, enabling more accurate causal inference in digital survey research.

Contribution

It applies double machine learning to survey statistics, providing a novel method for causal analysis of nonresponse in high-dimensional online panel data.

Findings

Double machine learning yields less biased causal estimates.

Method effectively handles high-dimensional survey data.

Improves understanding of nonresponse mechanisms.

Abstract

Survey scientists increasingly face the problem of high-dimensionality in their research as digitization makes it much easier to construct high-dimensional (or "big") data sets through tools such as online surveys and mobile applications. Machine learning methods are able to handle such data, and they have been successfully applied to solve \emph{predictive} problems. However, in many situations, survey statisticians want to learn about \emph{causal} relationships to draw conclusions and be able to transfer the findings of one survey to another. Standard machine learning methods provide biased estimates of such relationships. We introduce into survey statistics the double machine learning approach, which gives approximately unbiased estimators of causal parameters, and show how it can be used to analyze survey nonresponse in a high-dimensional panel setting.