Related Experiment Videos

Simpson's paradox: an example from hospital epidemiology

R Reintjes1, A de Boer, W van Pelt

  • 1Department for Infectious Diseases Epidemiology, National Institute of Public Health and the Environment (RIVM), Bilthoven, The Netherlands.

Summary

Simpson's paradox, a statistical anomaly, is demonstrated using real data from a multicenter nosocomial infections study. This extreme confounding phenomenon is explained intuitively.

Related Concept Videos