Testing the exceptionally gifted - where most tests break down...

There is only one broadly recognised IQ test that can measure these extreme ranges - the Stanford-Binet Form LM normed in 1972.

The most commonly used IQ tests, the more recent Stanford-Binet IV and the Weschler Intelligence Scale for Children - III, and the Weschler Adult Intelligence Scales all start to seriously break down somewhere between IQ 130 and IQ 140 and fail completely around the 150 range.

The SB-LM, if used on someone young enough can measure up to around 220+ - although it really has to be used before a child turns 12 to have any particular value and in the case of the profoundly gifted, prior to the age of 9.

The Stanford-Binet LM is generally accepted by the experts on the Exceptionally Gifted and Profoundly Gifted (Kearney and Silverman among them) as being useful for measuring IQ at these levels. The publisher of the test also regards it as useful - even though they more or less abandoned it when the Revision IV came out (my opinion).

There are known problems with the SB-LM - one of the major ones is that there is something wrong with the probability predictions for high IQs (those over 160 or so). The prediction was for example that the number of people with IQs at the 200 level should have been exceedingly small - around one in a billion. That isn't so - they are far more common than that (at least 35 times more common - probably more than that, IMHO, but that's the best figure I've personally seen). The 1 in a million level should be reached at around 178 - that isn't true either. So the probabilities at those levels are skewed - why is the subject of some debate.

But it really doesn't matter. It's also known that even though the probabilities are skewed, that comparisons between scores are still valid. Someone with an IQ of 200 does exceed the level of someone with an IQ of 180 who does exceed the level of someone with an IQ of 160. The fact that the scores may be more common than expected does not diminish the value of the test in drawing comparisons and judging what someone is likely to need.

Now, as for people gaining higher scores today than in the past, this is certainly true - known as the Flynn Effect. The average IQ has increased around 15 points. But the thing is - the Flynn effect only seems to apply around the average. It doesn't seem to apply in the higher IQ ranges - 50 years ago, the top 2% of people had a measured IQ of 130 - and today that figure is still around 2% even though the average IQ has increased 15 points. These changes have not been shown to have any relevance to giftedness. They have occurred - and why they have occurred could involve a number of factors. People are better fed today than they were in the past, they are better educated, they are better informed. What causes the change? Maybe the better general health? Maybe the better education? Maybe the testing is better? In all probability a number of factors apply - individually their effects may be small but taken together a 15 point increase could easily result. But this increase is not seen among the gifted.

Taken with permission from an OG message by Shaun Hately

 © NSWAGTC 2007

Jump to... Top | Why IQ testing matters | WPPSI R | WISC 3 | Characteristics of gifted children | Identification | Information Centre

Suggestions? | Like to Help? | Disclaimer | Webmaster | Site maintained for the NSWAGTC by Austega