Valor P: Correcta e incorrecta interpretación

Abstract

Thanks to the predominance of the frequentist school, nowadays, inference from the data obtained in clinical trials, is mainly done by using the P value and the alpha probability of error. These two, despite being useful and important statistical tools, have limitations that are not widely understood, leading to confusion about its meaning and utility. When doing clinical research, there must be good comprehension of the problem at hand, the outcomes and the optimal statistical tools to employ, in order to get conclusions as true as possible. The preference of some statistical methods over another, generates a limited vision and interpretation of the evidence.