An educational document explaining the impact of multiple testing and approaches to controlling it.
What the guide offers
This clinician-oriented educational document (14 pages) defines and describes different types of errors that can occur in null hypothesis significance testing. The increase in the Type I error rate when multiple hypotheses are tested is described and illustrated with a clinical example. The authors then explain different approaches to and methods for controlling the Type I error rate under multiple testing. The impact of multiple testing adjustments on significance findings and research conclusions is illustrated. Finally, recommendations for when and what methods to use to control for multiple testing are discussed.
Who created it
Ryan Simmons, MSc, and the Research Design and Analysis Core of the Duke University Global Health Institute (Durham NC).