Personal experience and statistical truth

The question I want to raise is one that is a real puzzler for intelligence researchers. Why don't we get no respect? (Yes, I know that's a double negative. I'm just trying to imitate a TV comic.)

The statistics are pretty clear. Tests of general cognitive ability are by far the best tests for predicting performance in both academia and the workplace. If a job is cognitively demanding, the correlation between test performance and job performance will be somewhere in the .40-.55 range. This is about twice as high as any tests of personality or motivation. It is as lot better predictor of performance than a free-form interview...which is how an awful lot of people are hired.

BUT it's also pretty clear that people do not believe this. Over and over, I read and hear statements like "Personality is much more important than intelligence" or "Motivation is the important thing." or "
I knew someone who had real good (bad) test scores and they did as real bad (good) job."

I add that these statements are not limited to non-specialists. I have heard them from many Ph.D. Psychologists who don't study intelligence. Now why should this happen?

I think there are three reasons, all of which could be true to some degree. But I'd bet most on the third one.

The first is a sort of confusion about how we want things to be. Most of the people who talk about tests have pretty liberal social philosophies. They are concerned about the fact that there are differences between demographic groups in their test scores, so they worry that validating test scores would validate things like job discrimination. So, because desire directs belief, the validity of the scores is downgraded.

The second is a similar concern. US society and European liberal social thinkers are strongly anti-elitist. Some people are worried that because higher socio-economic groups show higher test scores, and because there is research showing that tests scores have a strong genetic loading, accepting the tests would amount to intellectual elitism and would reinforce present day social inequalities.

I think many intelligence researchers would say that the two reasons I have just given are the reason that intelligence tests are maligned. But saying this maligns the maligners! Both reasons have the hidden message "People know the test scores predict behavior, but they don't want to admit it.

I think there is a third reason. It seems more in accord with life, and it does not amount to a claim that people are denying things they know. (Truth in packaging: I borrowed this idea from Charles Murray.)

When people reason about things, their personal experiences have much more weight than abstract statistics. AND, second point, we live in a cognitively segregated society. (The reason is at least partly because we live in an educationally segregated society.) To see what I mean, ask yourself how many of your friends...people you deal with at least once a week, and in a setting that is not restricted to formal exchanges, like a passenger-bus driver exchange, have markedly different educational backgrounds than you do?. I would bet that not very many do. (This is not quite so much the case if you are in the military, but that's a special situation.)

Put these two tendencies together, and you see that most people do not get to observe the problem solving behaviors of people whose intelligence differs very much from their own. This is certainly true of professionals, college professors, executives, etc....The people who do the talking. Because the typical person sees only a small bit of the range of intelligence that is actually out there, the importance of intelligence...in the big picture, in the whole population just isn't appreciated.

So do you think this is true? Thought question for the day: What are possible problems in a cognitively segregated society and how could they be reduced?