The Pervasive Avoidance of Prospective Statistical Power: Major Consequences and Practical Solutions

Abstract

The estimation of the prospective statistical power (PSP) is mandatory when using a classical Neyman-Pearson statistical method that together with the one by Fisher, represents one of the pillar of the so-called frequentist statistical approach. Unfortunately in the course of the history of their application, these two methods have been mixed, adopting the Fisher approach for hypotheses or model comparisons and their differences ignored in the classical Null Hypothesis Significant Testing (NHST) approach, which at present, represents the most used statistical approach in many research fields, from psychology to medicine, from neuroscience to ecology.

The uncritical application of the NHST statistical approach in ignoring its assumptions, strengths and weakness, has been considered one if not the principal cause of the “crisis of confidence” in scientific evidence.

In this paper we offer practical solutions on how to overcome this situation.

Keywords: prospective statistical power; NHST; replicability