157 – Translational Statistics: Problems in Preclinical Studies and a Call for Change
Statistical Reporting Must Improve in Preclinical Research: A Call for a Mentality Shift
Romain-Daniel Gosselin
Lausanne University Hospital
It is well-established that faulty handling of statistics contributes to the reproducibility crisis in preclinical research. The widespread adoption of the myriad of existing guidelines is hard to achieve because researchers often protest against the costs of these reforms, in terms of money, workforce or ethics. However, such justifications are invalid in the case of statistical reporting since more rigorous data presentation would simply require a cultural change with no further resources. Precisely, a quantification of the existing disregard for statistical reporting in the published literature shows that bad habits (such as failure to disclose sample sizes, statistical procedures or software/code) are still ubiquitous and have largely outlived all guidelines. The ACcess to Transparent Statistics (ACTS) call to action is presented, assembling four simple pragmatic measures that are rapidly achievable by journals to enhance the quality of statistical reporting in preclinical research through a global cultural change. The ACTS call to action is a plea for concrete top-down action from publishers of scientific journals, which should spearhead the battle against irreproducibility.