Abstract
he P value was introduced as a value to evaluate the results of statistical tests. The basic concept originated in the 1920s, and after the Second World War it was significantly expanded.
For about the last three decades there has been intense discussion about the problematic features of the P value concept and its use in science, and voices calling to abolish use of the P value are growing louder. In addition, suggestions have been made for alternative procedures that could replace or supplement the P value.
Statisticians have tried to invent an indicator similar to the P value, but without its weaknesses. There are many of these options.
Besides alternatives within the classical statistical testing paradigm, the use of an alternative statistical approach, so-called Bayesian statistics, is increasingly being discussed. An example of a moderate recommendation is that of using the Bayes factor, essentially an analogue of the P value in the Bayesian world.
The aim of this article is to present the Bayes factor in detail, to describe its similarities and dissimilarities with the P value, and discuss the possibilities of its calculation. In addition to computational procedures, a detailed discussion of the weaknesses of the Bayes factor is also included.