Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimates

Abstract

Many applications of computational social science aim to infer causal conclusions from nonexperimental data. Such observational data often contains confounders, variables that influence both potential causes and potential effects. Unmeasured or latent confounders can bias causal estimates, and this has motivated interest in measuring potential confounders from observed text. For example, an individual's entire history of social media posts or the content of a news article could provide a rich measurement of multiple confounders. Yet, methods and applications for this problem are scattered across different communities and evaluation practices are inconsistent. This review is the first to gather and categorize these examples and provide a guide to dataprocessing and evaluation decisions. Despite increased attention on adjusting for confounding using text, there are still many open problems, which we highlight in this paper.

Adjusting for Confounding with Text Matching

“…Mozer et al. (2020) and Veitch, Sridhar, and Blei (2019) directly build on our framework to propose alternative text adjustment approaches, and the related literature is reviewed in Keith, Jensen, and O'Connor (2020).…”

Unsupervised Discovery of Implicit Gender Bias

“…While their goals differ greatly from ours, our framework is generally consistent with their recommendations. Keith et al (2020) provide a more complete overview of using text to reduce the influence of confounding variables.…”