Hey Morphius,
Yeah, I would have been interested in the numbers also. Of course, the VNS numbers would have been just an estimate, subject to random sampling error and all that. They'd have to have a very large sample to be able to have a reasonable chance of detecting a difference of a few tenths of a percentage point. In a way, it might have been less misleading for them to just leave it the way they did--with a statement that the results probably (i.e. >50% chance) went this way or the other. Still, the statistician in me still would rather see the actual numbers, even though I know that they'd be misinterpreted by folks who aren't used to dealing with lies...uh, I mean...statistics. [img]http://www.chiefsplanet.com/ubb/smile.gif[/img]
|