Saggio di ipotesi statistiche e di significatività. Problemi fondamentali e fallacie applicative di interesse psicologico

A critical review of methods of Null Hypothesis Statistical/Significance Testing (NHST) is developed focusing on psychological science and practice. Frequentist and Bayesian NHST paradigms are compared and contrasted. Issues at a fundamental level and application fallacies are pointed out, and state-of-the-art remedies, if any, are presented and discussed. In conclusion, it is shown that NHST is prone to theoretical and interpretation problems, seriously affecting the logical consistency of results attainable by use of such testing methods. To avoid NHST misuses and related misinterpretations is a matter of concern of scientists, practitioners, teachers, and students who are making a habit of implementing statistical inference techniques in experimental research fields.