Human IQ and Increasing Intelligence

Abstract

The only real evidence we have that childhood intelligence is increasing is that scores on tests of intelligence are increasing. But what if the tests are wrong? What if tests are a poor measure of intelligence or do not measure it at all? Could it be that the standardized tests so widely used do not measure our ability to learn, but rather measure how much we have already learned? We have defined intelligence as the ability to solve problems, but virtually any problem that is posed assumes some degree of prior knowledge. For example, most adult IQ tests presuppose that test takers know how to read. Many intelligence tests use analogies (“Finger is to hand as X is to foot”) to test logical ability, but analogies indirectly test whether a subject has an adequate vocabulary. Similarly, no one could use mathematical reasoning skills unless they have prior knowledge about numbers, and how to use them. Even a brilliant person who is not a native English speaker might perform rather poorly in a timed test given in English – and many IQ tests are timed. We have postulated three technical hypotheses that could potentially explain the Flynn effect, all of which contend that intelligence tests somehow mismeasure intelligence. But before we explore these hypotheses in depth, it is important to shed some light on the tests themselves.