> From: "Steve Simmons" <scs at di.org> > At least when I was in school, you pretty much had to be at least two > standard deviations from the norm to be extreme. Beyond three deviations > the IQ test ceases to have much qualitative meaning except 'extreme.' Another small problem with using IQ tests at extreme ranges is that a lot of tests have 'celings'. (I don't know if they have 'floors'; I'd suspect so, but I havn't looked into it.) That is, there is a certain score which is the highest possible score for a person to get on the test; anyone with an IQ above that score can't prove it with the test. Thus, IQ *scores* can cease to be meaningful above a certain point simply because they weren't measured accurately. One might look at this as a good thing - an acknowledgement that the 'extreme' values aren't very meaningful - except it isn't a case of an artifically imposed celing; it's a flaw in the test design. I was tested as a child. I don't know my IQ, only that it is somewhere above 130: I hit the test celing. Of course, this doesn't even address the issue of validity, but it's an interesting point.