Directed acyclic graphs in clinical research
Olaf M. Dekkers, Kristina Laugesen, Rolf H. H. Groenwold
Abstract
Directed acyclic graphs (DAGs), or causal diagrams, are graphical representations of causal structures that can be used in medical research to understand and illustrate potential bias, including bias arising from confounding, selection, and misclassification. Further, they provide guidance for researchers about how to address a potential bias.
Topics & Concepts
Directed acyclic graphConfoundingSelection biasGraphical modelCausal structureSelection (genetic algorithm)Causality (physics)Causal modelComputer scienceMathematicsStatisticsAlgorithmMachine learningQuantum mechanicsPhysicsAdvanced Causal Inference TechniquesHealth Systems, Economic Evaluations, Quality of LifeStatistical Methods in Clinical Trials