mlsanityData to 13 Oct 2026

Proximal Causal Inference With Text Data

arXiv 2401.06687 ↗First posted 12 Jan 2024cs.CL cs.LG stat.MEcited by 7

Not on the map. Why?

Abstract (arXiv)

Recent text-based causal methods attempt to mitigate confounding bias by estimating proxies of confounding variables that are partially or imperfectly measured from unstructured text data. These approaches, however, assume analysts have supervised labels of the confounders given text for a subset of instances, a constraint that is sometimes infeasible due to data privacy or annotation costs. In this work, we address settings in which an important confounding variable is completely unobserved. We propose a new causal inference method that uses two instances of pre-treatment text data, infers two proxies using two zero-shot models on the separate instances, and applies these proxies in the proximal g-formula. We prove, under certain assumptions about the instances of text and accuracy of the zero-shot predictions, that our method of inferring text-based proxies satisfies identification conditions of the proximal g-formula while other seemingly reasonable proposals do not. To address untestable assumptions associated with our method and the proximal g-formula, we further propose an odds ratio falsification heuristic that flags when to proceed with downstream effect estimation using the inferred proxies. We evaluate our method in synthetic and semi-synthetic settings -- the latter with real-world clinical notes from MIMIC-III and open large language models for zero-shot prediction -- and find that our method produces estimates with low bias. We believe that this text-based design of proxies allows for the use of proximal causal inference in a wider range of scenarios, particularly those for which obtaining suitable proxies from structured data is difficult.

Key facts. Cited by 7 papers in the map (data through 13 Oct 2026). Builds on 24 of its 61 references that are in the map. First posted on arXiv 12 Jan 2024.

Builds on

24 of 61 references.

References · 24

  1. 2020Language Models are Few-Shot Learnerscited by 32,320
  2. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,765
  3. 2022Scaling Instruction-Finetuned Language Modelscited by 3,007
  4. 2021Multitask Prompted Training Enables Zero-Shot Task Generalizationcited by 1,468
  5. 2021Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpuscited by 485
  6. 2019Benchmarking Zero-shot Text Classification: Datasets, Evaluation and Entailment Approachcited by 304
  7. 2021Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyondcited by 178
  8. 2016Identifying Causal Effects With Proxy Variables of an Unmeasured Confoundercited by 139
  9. 2017Automated versus do-it-yourself methods for causal inference: Lessons learned from a data analysis competitioncited by 106
  10. 2020An Introduction to Proximal Causal Learningcited by 94
  11. 2018How to Make Causal Inferences Using Textscited by 59
  12. 2020Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimatescited by 59
12 more
  1. 2021Proximal Causal Learning with Kernels: Two-Stage Estimation and Moment Restrictioncited by 55
  2. 2021Causal Inference Under Unmeasured Confounding With Negative Controls: A Minimax Learning Approachcited by 43
  3. 2020Causal Effects of Linguistic Propertiescited by 39
  4. 2018Benchmarking Framework for Performance-Evaluation of Causal Inference Analysiscited by 33
  5. 2018Challenges of Using Text Classifiers for Causal Inferencecited by 33
  6. 2019The Case for Evaluating Causal Models Using Interventional Measures and Empirical Datacited by 30
  7. 2019Estimating Causal Effects of Tone in Online Debatescited by 19
  8. 2022On Testability of the Front-Door Model via Verma Constraintscited by 8
  9. 2023RCT Rejection Sampling for Causal Estimation Evaluationcited by 7
  10. 2016On falsification of the binary instrumental variable modelcited by 5
  11. 2023Causal Inference With Outcome-Dependent Missingness And Self-Censoringcited by 5
  12. 2023Statistical and Causal Robustness for Causal Null Hypothesis Testscited by 3

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2025: 2; 2026: 5.

7 papers.

  1. 2026Towards Multimodal Time Series Anomaly Detection with Semantic Alignment and Condensed Interactioncited by 3
  2. 2025LLM-Driven Treatment Effect Estimation Under Inference Time Text Confoundingcited by 2
  3. 2025Annotation-Assisted Learning of Treatment Policies From Multimodal Electronic Health Recordscited by 0
  4. 2026Predicting Causal Effects from Natural Language Queries using Structured Representationscited by 0
  5. 2026Uncertainty-Guided LLM Semantic Augmentation for Heterogeneous Treatment Effect Estimationcited by 0
  6. 2026Exploring Sparse Autoencoders in Text-Based Causal Confounding Adjustmentcited by 0
  7. 2026Proximal Balancing for Causal Effect Estimation under Unmeasured Confoundingcited by 0