Thursday, July 14, 2005

Okay, so here's a thought about research...

Titled Research: Third of study results don't hold up, one has to wonder -- what about the accuracy of this study? What if it's off by a factor of 30% (or even total crap)? That would mean that up to .3 * .3 studies (or less than 10 percent) are wrong, um conservatively speaking. Or, we could take the positive approach, which would be that it's off in the other direction -- an additional 30 percent of the 30 percent that are wrong are wrong. That would be .30 + .70*.30, or a whopping 51% of all studies being crap.

Of course, this cheap shot analysis is wrong, as should be readily apparent -- it's probably off by more than .30, even. Wait, maybe it's off by 9% or 51%...


No comments: