How to improve inferential statistics reporting in health sciences

Abstract

Statistic techniques for inference are essential for health sciences. Those techniques are useful to identify, for example, risk factors. However, the scientific communication process can be biased when inferential statistics are wrongly used.  Here we provide seven guidelines to help readers to use the p-value and Bayes factor, two inferential statistics. Although the Bayes factor is less known than the p-value it is also prone to be misinterpreted and misused. A better scientific communication of research output would lead to a better understanding of scientific discoveries. As a result, this improvement in the information process would affect positively public health. We hope our guidelines to be helpful for researchers, reviewers, editors, policy decision makers and the general public.