Wednesday, March 03, 2010

Statistics and Medical Studies: Making False Positives a Certainty

No big surprises here, but it's good to see this problem getting more attention.

A friend of mine, Holly, and I used to discuss starting a "Journal of Negative Results," though we (as is my nature) never did anything but talk about it. Turns out it's a thing (here, here), which is good.

This is a problem in philosophy, too--though since we have little effect on anything important, it's less cause for concern. But you're not going to get a Ph.D. if the last line of your dissertation was "so I guess the argument I've been developing doesn't work after all." Science will figure this problem out, because it matters, and because self-correction is the essence of science. Philosophy will probably continue to wallow in its own crap for a couple thousand more years. (Here's the question I most commonly ask myself: if I were wrong, how could I tell? (This is not a rhetorical question.))

Previously: fishy science.


Post a Comment

Subscribe to Post Comments [Atom]

<< Home