Three aspects of IQ scores

There is a lot of debate about IQ scores and intelligence and whether the former measures the latter.  There’s also a lot of debate about what intelligence is.  Of course, if we don’t know what intelligence is, we can’t really tell if IQ scores measure it.  But one problem with IQ scores is that they are usually reported as a single number.  That’s not adequate.

IQ tests such as the WISC (Wechsler Intelligence Scale for Children) or WAIS (Wechsler Adult Intelligence Scales) consist of a number of subtests.  How many subtests there are, and what they cover, varies among IQ tests, and, even within the Wechsler scales, varies from one version of the test to another.  But there are subtests.

On any test that consists of subtests, there are three aspects to consider: Elevation, scatter and shape.

Elevation is a measure of how high the score is.  It is usually operationalized as a mean, but sometimes as a median.  It is a central tendency.  This is the number that is usually reported.  If someone says “His IQ was 110” this is what they mean.

Scatter is a measure of how varied the subtest scores are.  It is commonly operationalized as a standard deviation, or sometimes a range or interquartile range.  It is a measure of spread.

Shape is a measure of which subtest scores are high or low.  There are, as far as I know, no generally accepted measures of shape. If shape is looked at all, it is looked at on a case by case basis.

Now, if we ignore the controversy over what intelligence is, then elevation is a measure of how smart you are, scatter can be an indication of whether you are learning disabled, and shape can show where those disabilities are.

This is a whole lot better than a single score, but it is still not adequate because IQ tests, regardless of their other attributes or problems, are not complete.  Anything that a newborn baby doesn’t know and that an adult does know has to be learned.  You can be LD in any of those skills, and  not even the strongest proponent of IQ tests would claim that they measure all of those skills.

Comments

  1. Peter, my daughter who was assessed with NLD had the WISC as part of tests. I found it very interesting as the psychologist did look at all the aspects you mention here.

    Is it fair to say that no matter the scatter or shape a high elevation is a high elevation…but that if the tests were done differently they could be higher for someone who is learning disabled ? I am referring to your last point about one having to learn something to be measured.

    I mean the scatter, to your point does not affect the elevation or does it ?

  2. Hi Michelle,
    I’m not Peter, but perhaps I can give a bit of insight? My son’s most recent WISC gives him an IQ of 128-respectable. But in that 128 is a Visual Processing score of 17%. On some of his subtests he would get a 99.9% and another a 4%. This is scatter-he has a huge margin between his good areas and his bad ones because he is LD. AND if he had a VP score that was more in line with the most of his tests, his IQ would go up to something like 145-150 or higher, (because some of the other subtests also take into account VP as a component.)

    Does this help?

  3. Hi Michelle

    I am not sure what you mean by “if the tests were done differently”. All people will get different scores on different types of IQ tests. For LD people, these scores will vary more.

    But scatter doesn’t DIRECTLY affect elevation. In theory, the range could be 30 to 130 or 100 to 200. It’s more complex, though, because there are floor and ceiling effects. It is not possible to get a score above a cerrtain number on any section

Speak Your Mind

*