Quote:
Originally posted by kappaloo
Well, let's think about it logically. The best studies are usually valid within about 3 percentages points 19 time out of 20 (95% of the time). The other 5% of the time, the results could be absolute crap.
How do we ensure that a particular study isn't that one time out of 20? We do more studies to show that the results are not unique. It's just common sense.
|
Those are confidence intervals. It doesn't mean that 5% of the time it will be wrong.
-Rudey