Hacker News new | ask | show | jobs
by snydly 3828 days ago
> P-values should be used with strong caution.

FiveThirtyEight (and Scientific American, and others) did some pretty interesting articles about this recently if you haven't seen it:

http://fivethirtyeight.com/features/science-isnt-broken/

Just from personal experience, the use of p-values is really broken in biology/chemistry. The things I've heard principal investigators say...