tag:blogger.com,1999:blog-6976144462093297473.post5633824509623181707..comments2024-03-26T19:10:00.791-04:00Comments on West Coast Stat Views (on Observational Epidemiology and more): General versus particular casesJosephhttp://www.blogger.com/profile/10760453165301871031noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-6976144462093297473.post-15903589370100917112014-07-06T17:35:47.735-04:002014-07-06T17:35:47.735-04:00Multiplying p-values as Tracy and Beall suggest do...Multiplying p-values as Tracy and Beall suggest does not get you another p-value. In fact, it is unclear what it gets you besides a product of p-values. The easy way to see that is to take a study, divide the data in half at random, so that we have, in effect, two independent studies, and then mulitply their p-values together.<br /><br />P-values measure results that did not happen as well as results that did. Their product does not have the same meaning as the product of the probabilities of two independent results that did happen. If it has any meaning at all.Billikinnoreply@blogger.com