Here's a pretty interesting article that argues that "most claimed research findings are false."
http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.0020124
The statistics get a little bit dense, but scanning through the main points is still sufficiently alarming. What I mostly take away from this article is that academia needs to move past its fetishization of positive results. However, if we start publishing all negative results, the sheer amount of published material would be enormous. This brings us back to the topic of "information overload" that we've been discussing since the beginning. Anyway, it's quite a relief to know that we can all still get published without actually discovering anything!
Thoughts anyone?
What this paper presents isn't controversial - the calculations are inescapable. How the models presented map to reality can be argued, but it's at least clear that thinking about your results in this way can provide an important check against fooling yourself.
ReplyDeleteIt is no fun to convince yourself that something is true when it is not. In fact this is much worse than deciding you really don't know.
So let's talk about this a bit when we get a chance.