Significance Test Controversy Revisited

Bruno Lecoutre  was director of research at the Centre National de la Recherche Scientifique (CNRS) in Paris and then in Rouen since 1996. He was a consultant in the pharmaceutical industry for 20 years. He retired in 2015. He obtained a PhD in experimental psychology from the University of Paris VIII in 1976 and a PhD in mathematics from University René Descartes in Paris in 1980. His main research interests have been in the foundations of statistics and in the development and applications of Bayesian procedures for the analysis of experimental data, including their methodological and computational aspects.Jacques Poitevineau taught experimental psychology and statistics at the university for 13 years and worked for 40 years at the Centre National de la Recherche Scientifique (CNRS) in Paris. He retired in 2013. He started programming in 1972, mainly in Fortran. He obtained his PhD in psychology from the University of Rouen in 1998. His research interests include computational statistics and statistical applications in experimental sciences.

• Introduction.- Preamble - Frequentist and Bayesian Inference.- The Fisher, Neyman-Pearson and Jeffreys Views of Statistical Tests.- GHOST: An Officially Recommended Practice.- The Significance Test Controversy Revisited.- Reporting Effect Sizes: The New Star System.- Reporting Confidence Intervals: A Paradoxical Situation.- Basic Fiducial Bayesian Procedures for Inference About Means.- Generalizations and Methodological Considerations for ANOVA.- Conclusion.- Index.