Causal Mediation Analysis

Follow this topic Ask about this topic

Causal mediation analysis asks how much of an intervention’s effect on an outcome operates through an intermediate variable, such as a biological marker or a financial factor. It distinguishes natural direct and indirect effects from interventional effects, which can be more practical when treatment changes confounders of the mediator–outcome relationship. Researchers use regression and g-computation alongside weighting, doubly robust estimators, influence-function methods, and representation or generative models for complex, high-dimensional data. A central challenge is that pathway effects often require strong, difficult-to-check identification assumptions, so flexible estimation must be balanced with careful sensitivity analysis and clear interpretation.

Latest papers

15 papers2 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

When Reconstruction Is Not Enough: Doubly Robust Counterfactual and Interventional Mediation Analysis with Learned Multimodal Representations
Oct 21, 2026Cong CaoCausal Effect Estimation Causal Mediation Analysis
Department of Biostatistics, Yale School of Public Health, New Haven, CT 06520, USA · Department of Surgery, Stanford University School of Medicine, Stanford, CA 94305, USA
Backtracking Direct, Indirect and Total Effects
Oct 20, 2026Jose M. PeñaCausal Effect Estimation Causal Mediation Analysis
STIMA, IDA, LiU
Path-specific harm decomposition: A partial identification framework
Sep 24, 2026Ruizi Yan, Dennis Frauen, Maresa Schröder +1Causal Effect Estimation Causal Mediation Analysis
Department of Statistics, University of Oxford · LMU Munich · Munich Center for Machine Learning (MCML)
Causal Attribution for Agentic Decisions: Estimators, Coupling, and a Traceability Specification
Sep 6, 2026Ajay Pravin MahaleCausal Attribution Causal Effect Estimation
Hochschule Trier
Representation Learning for Semiparametric Causal Mediation Analysis under No Essential Heterogeneity
Jul 12, 2026Roberto Faleh, Sofia Morelli, Holger BrandtSemiparametric Inference Parameter Estimation
Methods Center University of Tübingen, Germany
Causal-RetiGraph: Cross-Cohort Retinal Support and Same-Subject Pathway Analysis for Diabetic Retinopathy
Jul 6, 2026Inam Ullah, Imran Razzak, Shoaib JameelCausal Mediation Analysis Diabetic Retinopathy Grading
University of Southampton Southampton, United Kingdom · Mohamed bin Zayed University of Artificial Intelligence United Arab Emirates
Fixed-Confidence Best-Arm Identification for Causal Mediation Analysis
Jul 5, 2026Harsh Shrivastava, Yuta Kawakami, Junpei Komiyama +1Multi-Armed Bandits Causal Effect Estimation
Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), Abu Dhabi, UAE · RIKEN AIP, Tokyo, Japan
RetiSEM: Generalising Causal Models for Fragmented Biomedical Data
Jun 23, 2026Inam Ullah, Imran Razzak, Shoaib JameelLatent Variable Models Structural Causal Models
University of Southampton, Southampton, SO18 1PB, Hampshir, United Kingdom · Mohamed bin Zayed University of Artificial Intelligence, United Arab Emirates
MediEncoder: Nonlinear Representation Learning for High-Dimensional Causal Mediation Analysis
Jun 5, 2026Shi Bo, Debarghya Mukherjee, AmirEmad GhassamiCausal Effect Estimation Representation Learning
Department of Mathematics and Statistics, Boston University
Causal Fairness for Survival Analysis
May 12, 2026Drago PleckoCausal Mediation Analysis Survival Analysis
Department of Statistics & Data Science UCLA
Distributional Causal Mediation via Conditional Generative Modeling
May 3, 2026Jinlun Zhang, Haoneng Huang, Zishu Zhan +1Causal Effect Estimation Causal Mediation Analysis
State Key Laboratory of Multi-organ Injury Prevention and Treatment, Department of Biostatistics, Guangdong Provincial Key Laboratory of Tropical Disease Research, School of Public Health, Southern Medical University, Guangzhou, China
Decomposing Discrimination: Causal Mediation Analysis for AI-Driven Credit Decisions
Mar 29, 2026Duraimurugan RajamanickamFinancial Services Causal Effect Estimation
VP, Artificial Intelligence, Hudson Valley Credit Union, Poughkeepsie, NY · PhD Candidate, Causal Machine Learning, University of Arkansas at Little Rock
Sequential Transport for Causal Mediation Analysis
Mar 16, 2026Agathe Fernandes Machado, Iryna Voitsitska, Arthur Charpentier +1Causal Mediation Analysis Causal Attribution
Université du Québec à Montréal (UQAM), Canada · Ukrainian Catholic University, Ukraine · Kyoto University, Japan +2
Imagination Helps Visual Reasoning, But Not Yet in Latent Space
Feb 26, 2026You Li, Chi Chen, Yanghao Li +4Causal Mediation Analysis Visual Reasoning
School of Computer Science and Technology, Beijing Jiao-tong University · Key Laboratory of Big Data & Artificial Intelli-gence in Transportation (Beijing Jiaotong University), Ministry of Education · Tsinghua University
Multiply Robust Causal Mediation Analysis with Continuous Treatments
May 19, 2021Yizhen Xu, AmirEmad Ghassami, Numair Sani +1Causal Effect Estimation Causal Mediation Analysis
Division of Biostatistics, Department of Population Health Sciences, University of Utah, USA · Department of Mathematics and Statistics, Boston University, USA · Department of Computer Science, Johns Hopkins University, USA

Oct 21, 2026J/K move · Enter open · S save

When Reconstruction Is Not Enough: Doubly Robust Counterfactual and Interventional Mediation Analysis with Learned Multimodal Representations

Cong Cao

Department of Biostatistics, Yale School of Public Health, New Haven, CT 06520, USA · Department of Surgery, Stanford University School of Medicine, Stanford, CA 94305, USA

Causal Effect Estimation Causal Mediation Analysis Causal Representation Learning Doubly Robust Estimation

Open paper AskPDF arXiv

Learned representations of high-dimensional multimodal health records are increasingly used to adjust for confounding, yet they are usually trained to reconstruct the data rather than preserve the information that causal identification requires. We study counterfactual outcomes under alternative treatments, including phenotype-specific effects and patient-level counterfactual risk profiles of the kind digital twins provide, and interventional direct and indirect effects under exposure-induced intermediate confounding. We give conditions under which a representation preserves identification and show that interventional mediation is more demanding than counterfactual comparison. We prove that no covariate-only training objective, whether reconstruction, contrastive, or physics-informed, can certify that a representation is sufficient, and that, in a linear structural model, a representation can lose arbitrarily little reconstruction accuracy yet induce arbitrarily large bias. We decompose the error of cross-fitted doubly robust estimators into a first-order representation bias, which neither double robustness nor cross-fitting removes, and a second-order remainder of nuisance-estimation errors, so coverage worsens as the sample size grows when the representation is insufficient. We propose a workflow combining modality-specific variational autoencoders trained outside the estimation sample, cross-fitted doubly robust estimators, locked phenotypes, select-then-confirm pathway analysis, portable representations, and diagnostics for representation insufficiency, exposure leakage, and sensitivity. In simulations, discarding a rare, moderately strong confounder gave indirect-effect intervals with 2% coverage at n = 3,000; reconstruction loss missed the problem, and a simple predictive diagnostic flagged it in every replicate.