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

By Katherine A. Keith, David D. Jensen, Brendan T. O'Connor. 2020

132 citations recorded by the source corpus.

Many applications of computational social science aim to infer causal conclusions from non-experimental 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 data-processing and evaluation decisions. Despite increased attention on adjusting for confounding using text, there are still many open problems, which we highlight in this paper.

Subfields: Artificial Intelligence, General Social Sciences, Statistics and Probability, Natural Language Processing, 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 Artificial Intelligence

1. Brandon M Stewart
2. Margaret E. Roberts
3. Richard A. Nielsen
4. In Song Kim

In General Social Sciences

1. Brandon M Stewart
2. Margaret E. Roberts
3. Richard A. Nielsen
4. In Song Kim
5. Dustin Tingley

In Statistics and Probability

1. Richard A. Nielsen
2. Margaret E. Roberts
3. Brandon M Stewart

In Natural Language Processing

1. Richard A. Nielsen
2. Margaret E. Roberts
3. Brandon M Stewart

In Causal Inference & Discovery

1. Richard A. Nielsen
2. Margaret E. Roberts
3. Brandon M Stewart

Citations in the corpus

• Adjusting for Confounding with Text Matching (2020, background): “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).”
• How to make causal inferences using texts (2018, unknown): “In related work, Roberts et al. ( 47 ) introduce techniques for matching on text to adjust for confounders, and recent papers ( 48 , 49 ) consider other text-based confounding approaches.”