PRETEND INTELLIGENCE 85to examine IQ, states that “the mea sure ment of intelligence— which has
been done primarily by IQ tests— has utilitarian value because it is a rea-
sonably good predictor of grades at school, per for mance at work, and
many other aspects of success in life.”^9
How accurate and meaningful are such associations and correlations?SELF- FULFILLED VALIDITYIt is undoubtedly true that IQ test scores predict school achievement
moderately well, with correlations of around 0.5. Th e prob lem lies in the
self- fulfi llment of this prediction. Some test items are devised to contain
knowledge that has been learned in school: “In what continent is Egypt?”
“Who wrote Hamlet?” “What is the boiling point of water?” and so on.
Or they use a lot of textlike rules and analogies. So it should come as
no surprise that per for mance on them is associated with school per for-
mance. As Robert L. Th orndike and Elizabeth P. Hagen explained in
their leading textbook, Educational and Psychological Mea sure ment,
“From the very way in which the tests were assembled [such correlation]
could hardly be other wise.”^10
Rather than in de pen dent mea sures confi rming an under lying cognitive
power or capacity, IQ tests and school tests are simply diff er ent mea sures
of the same learning. Th is is why correlations between IQ and achievement
tests tend to increase with age. Th is is why parental drive and encourage-
ment with their children’s school learning itself improves the children’s
IQ (as the American Psychological Association group reports).^11
And that raises another question. If it is genuine predictability we
need, why not just rely on teachers’ judgments? It has long been demon-
strated that teachers can predict with far greater accuracy, and in a
fraction of the time, the future achievements of their pupils. One review
suggested a correlation of around 0.66, which far surpasses the predic-
tive power of an IQ test.^12 Yet how much more seductive is the mystique
of a scientifi c- looking test of “general intelligence.”
Similar doubts arise about the use of occupational level, salary, and so
on, as validatory criteria. School achievement largely determines level of
entry to the job market. Th e frequently reported correlation of 0.5
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:52:12 UTC