Most people are like Kansas, LD people are like Switzerland: Some thoughts on intelligence, tests and LD people

I am learning disabled.  I also have a PhD in psychometrics.  Psychometrics is the study of psychological measurement, including measurement of intelligence.  So, I have thoughts on this from two angles.  I’ll organize these in a list.

  1. Is intelligence a myth?  No.  It’s not.  How can I make this statement so strongly?  Because, if intelligence were really a myth, then statement such as “Einstein was intelligent” would have no meaning.  But they do have meaning.  A statement like “Einstein was intelligent” would be agreed to by just about everyone.  But even if we make a statement about intelligence that might find less universal agreement, such as “Obama is intelligent”, people who argue the different sides would bring in similar ideas.  If someone said “I can tell Obama is intelligent because his last name starts in a vowel” very few people would think that valid.
  2. If intelligence isn’t a myth, what is it?  That’s a MUCH harder question, and there is no universal agreement on any specific answer.  But we can say a few things.  First, intelligence is a latent trait.  That means that it cannot be measured directly (unlike, say, height).  One usual way of measuring latent traits is to measure a bunch of things that we think make up the latent trait.  This is problematic, because there is not agreement on exactly what things should go in to that list.  But there is some agreement.  Another problem is that, in the nature of latent traits, the individual things in the list only correlate imperfectly with intelligence.  That means that, even if we agree that (say) vocabulary should be part of an intelligence test, it will be possible to find people who have small vocabularies but who strike most of us as intelligent.
  3. Is intelligence unidimensional or multidimensional?  That’s a tricky question because the best answer I can come up with is both.  But here’s an illustration.  Suppose we wanted to know how tall ancient people were, but all we had was a bunch of fossils.  Suppose further that the fossils had been sorted so we were virtually certain that the bones of particular individuals were together.  How could we tell how tall they were?  We could measure each individual bone, and determine length of shin and hip, and spine and whatever else (I am not an anatomy expert) and then get an idea of total height.  For most people, the lengths of the different bones would match up.  People with long legs tend to have long torsos.  But there are exceptions.  Manute Bol (a former basketball player) was well over 7 feet tall, but nearly all of the “extra” height was in his legs.  His waist was 5 feet off the ground.  His upper body was big, but not extraordinarily so.  Similarly, with intelligence (if we could agree on exactly what makes it up) for most people the traits go together.
  4. What about tests of intelligence?  There are various tests of intelligence.  One of the best known sets are the Wechsler scales.  They are made up of a series of subtests.  The theory is that these subtests measure things that correlate with intelligence (whatever that is).  No one pretends this correlation is perfect.  But, for most people, there’s a fairly strong correlation among the different subtests.  Not perfect, but strong.  For many learning disabled people, this is not so.  For example, when I was 9 years old I took the WISC.  I got subtest scores from 60 to 160.  These averaged out to a bit over 100.  But that’s a meaningless measure.  Here’s an analogy:
  5. What’s the average elevation of a region?  Suppose we want to know the average elevation above sea level for different regions. Sometimes, this is sensible.  Sometimes it is not.  There was an Englishman from one of the flatter parts of that country who, on returning from Switzerland, said that “if you threw the mountains into the lakes you would be rid of two nuisances at once”.  Whatever your opinion of mountainous places, you could measure the height above sea level of (say) 10,000 places in Switzerland, and you could average them, but what would it tell you?  On the other hand, if you measured the average elevation of 10,000 places in (say) Kansas, and averaged them, the result would be meaningful.  Most people are like Kansas.  We LD folk are like Switzerland.

Comments

  1. Very good article.

  2. Nina Liakos says

    I love this analogy!

  3. Peter, great analogy !!

Speak Your Mind

*

This site uses Akismet to reduce spam. Learn how your comment data is processed.