How to make causal inferences using texts

By Naoki Egami, Christian Fong, Justin Grimmer, Margaret E. Roberts, Brandon M Stewart. 2018

185 citations recorded by the source corpus.

Text as data techniques offer a great promise: the ability to inductively discover measures that are useful for testing social science theories with large collections of text. Nearly all text-based causal inferences depend on a latent representation of the text, but we show that estimating this latent representation from the data creates underacknowledged risks: we may introduce an identification problem or overfit. To address these risks, we introduce a split-sample workflow for making rigorous causal inferences with discovered measures as treatments or outcomes. We then apply it to estimate causal effects from an experiment on immigration attitudes and a study on bureaucratic responsiveness.

Subfields: General Social Sciences, Political Science and International Relations, Sociology and Political Science, Causal Inference & Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In General Social Sciences

In Political Science and International Relations

In Sociology and Political Science

In Causal Inference & Discovery

Citations in the corpus

CausaLM: Causal Model Explanation Through Counterfactual Language Models (2020, background): “In areas such as medicine, the causal graph can built with the assistance of a doctor that is aware of potential confounders and their relationship, but in NLP this is more of a challenge.”
Unsupervised Discovery of Implicit Gender Bias (2020, background): “Egami et al. (2018) do consider a similar setting where text is both an outcome and a confound.”
Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyond (2021, background): “This position paper follows a small body of surveys that review the role of text data within causal inference (Egami et al., 2018; Keith et al., 2020).”
Adjusting for Confounding with Text Matching (2020, methodology): “Finally, Egami et al. (2017) provide a framework for causal inference with text as treatment or outcome, complimenting our discussion of text-based confounding.”
A Survey on Causal Inference (2020, methodology): “Moreover, a series of methods have been proposed [44, 156, 169, 173] to handle the case when the treatment, confounder, or outcome contains textual data.”