Answers are grounded in this paper and cite their sources.
Most medical researchers still believe that statistical testing is important to determine whether an estimated effect is important or not. Unfortunately, this believe is based on several misunderstandings regarding statistical significance. Therefore, the purpose of this paper is to highlight these misunderstandings and to convince medical researchers to refrain from statistical testing. The biggest problem with statistical testing is that it is based on an arbitrary cut-off value and a non-existing dichotomy. Therefore, statistical testing should not be used to evaluate the results of medical research. Instead of using statistical significance, the results should be evaluated by its clinical relevance. It is clear that refraining from statistical testing will be a challenge, but it will definitely improve the evaluation of the impact of medical research.
No takes yet. Share an insight, caveat, or question.
Jos WR Twisk (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: