tag:blogger.com,1999:blog-5841910768079015534.post6193858091641094053..comments2024-03-25T17:14:36.888+00:00Comments on BishopBlog: Three ways to improve cognitive test scores without interventiondeevybeehttp://www.blogger.com/profile/15118040887173718391noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-5841910768079015534.post-61689515594472784812011-10-01T17:07:37.747+01:002011-10-01T17:07:37.747+01:00One non-technical, common sense approach is to des...One non-technical, common sense approach is to describe it as the get-better-anyway effect. If you go for treatment when you are at your worst, there is only one way to go next, you get better (or die, but that's rather rare).David Colquhounhttp://dcscience.net/noreply@blogger.comtag:blogger.com,1999:blog-5841910768079015534.post-76509790711529473772010-08-17T13:13:42.916+01:002010-08-17T13:13:42.916+01:00An excellent overview. I'd just quibble with o...An excellent overview. I'd just quibble with one thing:<br /><br /><i>"Regression toward the mean is a horribly difficult concept to understand, even for those with statistical training."</i><br /><br />I agree that RTTM is often hard to understand but I don't think this is because it's an inherently hard concept, rather it's hard because we tend to talk about it as a <i>thing</i> whereas it is in fact just an inevitable consequence of selection on the basis of a random score. With my students, I've avoided using the term regression-to-the-mean and just asked them what would happen if you rolled 100 dice, picked out all the "1"'s, and rolled those dice again. Of course they'd score higher on average. Only then do I say that this "effect" has a technical name...Neuroskeptichttps://www.blogger.com/profile/06647064768789308157noreply@blogger.com