214 THE MISMEASURE OF MAN
literature already studded with absurdity—though Jensen (1979,
pp. 113 and 355) and others still take it seriously.*
Terman (1917) had already published a preliminary study of
Francis Galton and awarded a staggering IQ of 200 to this pioneer
of mental testing. He therefore encouraged his associates to pro-
ceed with a larger investigation. J. M. Cattell had published a rank-
ing of the i ,000 prime movers of history by measuring the lengths
of their entries in biographical dictionaries. Catherine M. Cox,
Terman's associate, whittled the list to 282, assembled detailed
biographical information about their early life, and proceeded to
estimate two IQ values for each—one, called Ai IQ, for birth to
seventeen years; the other, A2 IQ, for ages seventeen to twenty-
six.
Cox ran into problems right at the start. She asked five people,
including Terman, to read her dossiers and to estimate the two IQ
scores for each person. Three of the five agreed substantially in
their mean values, with Ai IQ clustering around 135 and A2 IQ
near 145. But two of the raters differed markedly, one awarding
an average IQ well above, the other well below, the common figure.
Cox simply eliminated their scores, thereby throwing out 40 per-
cent of her data. Their low and high scores would have balanced
each other at the mean in any case, she argued (1926, p. 72). Yet if
five people working in the same research group could not agree,
what hope for uniformity or consistency—not to mention objectiv-
ity—could be offered?
Apart from these debilitating practical difficulties, the basic
logic of the study was hopelessly flawed from the first. The differ-
ences in IQ that Cox recorded among her subjects do not measure
their varying accomplishments, not to mention their native intelli-
gence. Instead, the differences are a methodological artifact of the
varying quality of information that Cox was able to compile about
the childhood and early youth of her subjects. Cox began by assign-
ing a base IQ of 100 to each individual; the raters then added to
(or, rarely, subtracted from) this value according to the data pro-
vided.
•Jensen writes: "The average estimated IQ of three hundred historical persons...
on whom sufficient childhood evidence was available for a reliable estimate was IQ
155.... Thus the majority of these eminent men would most likely have been rec-
ognized as intellectually gifted in childhood had they been given IQ tests" (Jensen,
!979. P- H3)-