Such results are confirmed also by a study reported two years earlier, in which over 100 children had been re-measured with Army Alpha 10 to 12 years after their initial Stanford-Binet measurements at ages 7 to 9 years. [2] All these children had IQ's over 130, and half of them were over 150, ranging up to 190. From this study the following conclusions had been drawn:
"Of 116 children testing in the top centile of the distribution of school children by Stanford-Binet, 82 per cent were found when near maturity, ten years later, to rate in the top centile of the military draft by Army Alpha. The remainder rated in high centiles. No individual of either sex regressed to or nearly to the average. . . . Girls regressed from the top centile somewhat more frequently than boys, this regression being in part but not fully accounted for by the known sex difference between medians on Army Alpha.
"This result affords a validation, by means of elapsed time, of the predictive power of available mental tests on the one hand; and on the other, a proof of the constancy of the intellectual development of gifted children in terms of centile status."
CRITIQUE OF THE CONCEPT OF "GENIUS" AS APPLIED IN TERMS OF IQ
The term "genius" has been used by Terman—and following him by many others—to denote children testing at or above 140 IQ (S-B). In the light of the developmental data herein presented, it would appear that the term "genius" is thus misapplied, unless we wish to define as "geniuses" persons who represent approximately the best fourth of all students being graduated from American colleges.
Of individuals here followed to early maturity, those who test at about 140 IQ (S-B) are found to define approximately the 75th percentile of college graduates, taking the country over. They are far from "genius," if by that term is to be meant the degree of mental ability that is capable of outstanding original intellectual achievement. It is only when we have an IQ (S-B) of at least 160 in a child, that we may begin to expect mildly noteworthy accomplishments, such as winning "honors" in a first-class college. Very rarely are "honors" won in first-class colleges by those who test below this status in childhood. The small sample of college graduates here presented is truly representative of the much larger sample in our files (not tested by our end tests) in this respect.
Of primary interest to the present investigators is the subsequent history of those who in childhood have achieved the extremely infrequent rating of 180 IQ or higher. At maturity will these persons still stand out from their contemporaries in mental tests and in achievement?
This question is answered affirmatively by our data. The five children here included, [3] who achieved IQ's (S-B) on first test in childhood of more than 180, are they who "find the tops" on CAVD at maturity. Every one of these top-rank persons is noteworthy among contemporaries. Before the age of 22 in all cases, one had prosecuted research in history, one in mathematics, one in chess, and two had become established in learned professions. One stood high in the national ranking for chess. A long list of medals and prizes had been won by them. All but one of those graduated from college had been elected to Phi Beta Kappa.
These unusual achievements show how children testing above 180 IQ rise above the generality of the college populations in adolescence and in early maturity. None of those who tested in childhood around 140, 150, or 160 IQ (S-B) approaches these others at maturity in honors and prizes won, or in test scores.
This is, perhaps, the most significant fact to be derived from our data: that the children who test at and above 180 IQ constitute the "top" among college graduates. They are the students of whom one may confidently predict that they will win honors and prizes for intellectual work.