A Review of Generalizability and Transportability

Abstract

When assessing causal effects, determining the target population to which the results are intended to generalize is a critical decision. Randomized and observational studies each have strengths and limitations for estimating causal effects in a target population. Estimates from randomized data may have internal validity but are often not representative of the target population. Observational data may better reflect the target population, and hence be more likely to have external validity, but are subject to potential bias due to unmeasured confounding. While much of the causal inference literature has focused on addressing internal validity bias, both internal and external validity are necessary for unbiased estimates in a target population. This paper presents a framework for addressing external validity bias, including a synthesis of approaches for generalizability and transportability, the assumptions they require, as well as tests for the heterogeneity of treatment effects and differences between study and target populations.

arXiv comment30 pages, 3 figures

References in corpus (19)

• Matching Methods for Causal Inference: A Review and a Look Forward
• Recursive Partitioning for Heterogeneous Causal Effects
• BART: Bayesian additive regression trees
• Ethical Machine Learning in Health Care
• External Validity: From Do-Calculus to Transportability Across Populations
• Generalizing causal inferences from individuals in randomized trials to all trial-eligible individuals
• Sensitivity analysis for an unobserved moderator in RCT-to-target-population generalization of treatment effects
• Learning Weighted Representations for Generalization Across Designs
• Removing Hidden Confounding by Experimental Grounding
• Likelihood reweighting methods to reduce potential bias in noninferiority trials which rely on historical data to make inference
• Causal Inference for Comprehensive Cohort Studies
• Combining Non-probability and Probability Survey Samples Through Mass Imputation
• Propensity Score Methods for Merging Observational and Experimental Datasets
• Generalizing trial findings using nested trial designs with sub-sampling of non-randomized individuals
• Combining Observational and Experimental Datasets Using Shrinkage Estimators
• Know your population and know your model: Using model-based regression and poststratification to generalize findings beyond the observed sample
• On variance estimation for generalizing from a trial to a target population
• A Sequential Significance Test for Treatment by Covariate Interactions
• Is Big Data Sufficient for a Reliable Detection of Non-Technical Losses?

Cited by in corpus (13)

• Improving trial generalizability using observational studies
• Transportability without positivity: a synthesis of statistical and simulation modeling
• Causal effect on a target population: a sensitivity analysis to handle missing covariates
• Transportability of model-based estimands in evidence synthesis
• On adjustment for temperature in heatwave epidemiology: a new method and toward clarification of methods to estimate health effects of heatwaves
• Entropy Balancing for Causal Generalization with Target Sample Summary Information
• Minimax Rates and Adaptivity in Combining Experimental and Observational Data
• Who Are We Missing? A Principled Approach to Characterizing the Underrepresented Population
• Towards a unified approach to formal risk of bias assessments for causal and descriptive inference
• Uncertainty in Lung Cancer Stage for Outcome Estimation via Set-Valued Classification
• Conditional Cross-Design Synthesis Estimators for Generalizability in Medicaid
• Multi-Study R-Learner for Estimating Heterogeneous Treatment Effects Across Studies Using Statistical Machine Learning
• Regression and Causality