Redefine statistical significance

With data generation, high false positive rates, and inability to reproduce findings all on the rise in science, Benjamin et al. propose lowering the standard threshold of the p-value to 0.005. The p < 0.05 standard would be termed “suggestive” under this scheme. Some of the caveats are that this applies to new hypotheses, not the retesting of old hypotheses and obviously does not apply to fields where the standard is already under p < 0.005. It’s a good reminder to scientists to learn statistics and data analysis well. They do point out that this won’t necessarily solve all the issues science has, but is an accessible step. More on this pre-print was reported in Nature and Science. PsyArXiv Preprints 10.17605/OSF.IO/MKY9J (OA)

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *