Data Fusion for Partial Identification of Causal Effects

TL;DR

This paper introduces a partial identification framework with sensitivity analysis for causal effects using data fusion, addressing scenarios where assumptions like no unobserved confounding and exchangeability fail simultaneously, and applies it to educational data.

Contribution

It proposes a novel partial identification method with interpretable sensitivity parameters and doubly robust estimators to assess causal effects under assumption violations.

Findings

Framework applied to Project STAR data shows robustness of causal conclusions.

Sensitivity analysis quantifies how assumption violations affect causal effect bounds.

Method enhances causal inference reliability when assumptions are uncertain.

Abstract

Data fusion techniques integrate information from heterogeneous data sources to improve learning, generalization, and decision making across data sciences. In causal inference, these methods leverage rich observational data to improve causal effect estimation, while maintaining the trustworthiness of randomized controlled trials. Existing approaches often relax the strong no unobserved confounding assumption by instead assuming exchangeability of counterfactual outcomes across data sources. However, when both assumptions simultaneously fail - a common scenario in practice - current methods cannot identify or estimate causal effects. We address this limitation by proposing a novel partial identification framework that enables researchers to answer key questions such as: Is the causal effect positive or negative? and How severe must assumption violations be to overturn this conclusion? Our approach introduces interpretable sensitivity parameters that quantify assumption violations and derives corresponding causal effect bounds. We develop doubly robust estimators for these bounds and operationalize breakdown frontier analysis to understand how causal conclusions change as assumption violations increase. We apply our framework to the Project STAR study, which investigates the effect of classroom size on students' third-grade standardized test performance. Our analysis reveals that the Project STAR results are robust to simultaneous violations of key assumptions, both on average and across various subgroups of interest. This strengthens confidence in the study's conclusions despite potential unmeasured biases in the data.