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

Right - I'm charitably assuming that their study supported causality with p=0.02. In reality it doesn't.


My statement above but you are imputing the proposal of a causative radiation mechanism. is poorly phrased. A better way to put what I was getting at is that they are not claiming that radiation effects are the only possible causative mechanism. They found some observation in some data and published it, probably in the service of a future grant proposal. Saying they didn't prove anything for this publication is like saying water is wet.

Should hn spend much time discussing the result (or really, observation)? Absolutely not. Should they get grant funding for this? I dunno, I hope there are better proposals out there, but really have no idea.


Isn't it the case that we can expect roughly one in fifty studies with p<0.02 to be the result of statistical noise?


As dramatized at https://xkcd.com/882/




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

Search: