Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Medicine and life sciences adopted bad statistics. Every single paper claims something along the lines of "what we observed is so incredibly unlikely (<5%) assuming complete randomness, it must mean our theory is true, our new drug works, the bad pollutant causes cancer, etc." Papers that should claim "what we observed is completely ordinary and there is nothing to see here" never get published.

Even if this was done correctly (it isn't, as John Ioannidis keeps pointing out), you'd get lots of false positive results (Ioannides again). The moment you look at a collection of statistics, the only available conclusion is "They all contradict each other!!1"

Until they ask physicists for help, medicine, life sciences, and many other fields will not be evidence based. We can call them "argument based", but that's not the same...



Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: