When possible, report a Fisher-exactPvalue and display its underlying null randomization distribution
Authors:
Open Access
- 23 July 2020
- journal article
- research article
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 117 (32) , 19151-19158
Abstract
Significance: Statistical analyses of randomized experiments often rely on asymptoticPvalues instead of using the actual randomization procedure that led to the observed data. Fisher-exact and asymptoticPvalues can differ dramatically: The former should be preferred because it is calculated using the exact null randomization distribution, which, in small samples, can substantially differ from its approximating Student’stdistribution. Moreover, we may learn something scientifically interesting from examining the shape of the null randomization distribution.Keywords
Funding Information
- HHS | NIH | NIH Office of the Director (DP5O0D021412)
- HHS | National Institutes of Health (RO1 AI102710)
- National Science Foundation (IIS 1409177)
This publication has 12 references indexed in Scilit:
- B vitamins attenuate the epigenetic effects of ambient fine particles in a pilot human intervention trialProceedings of the National Academy of Sciences, 2017
- Causal Inference for Statistics, Social, and Biomedical Sciences2015
- Controlled Exposure of Healthy Young Volunteers to Ozone Causes Cardiovascular EffectsCirculation, 2012
- Design of Observational Studies2010
- For objective causal inference, design trumps analysisThe Annals of Applied Statistics, 2008
- Statistical Models for CausationEvaluation Review, 2006
- The design versus the analysis of observational studies for causal effects: parallels with the design of randomized trialsStatistics in Medicine, 2006
- Linear Statistical Models for Causation: A Critical Review2005
- Constructing a Control Group Using Multivariate Matched Sampling Methods That Incorporate the Propensity ScoreThe American Statistician, 1985
- The Future of Data AnalysisThe Annals of Mathematical Statistics, 1962