Educability-and-Group-Differences-1973-by-Arthur-Robert-Jensen

Equating for Socioeconomic Variables 239
if the premises are questionable. But the logic suggests an interesting
comparison. What if we compared (a) Negro children reared
in upper-middle-class homes by Negro parents whose educational
and occupational status and income were well above the average of
the white population with (b) white children reared in the lowest
SES category, whose parents are well below the average in
intelligence, have less than an average education, and are either
in unskilled work or on welfare
Even if parental IQs were not measured, there would be little
doubt in such a case that the high SES Negro parents would have
higher IQs in general than the low SES white parents. If these
SES factors are more important determinants of IQ than genetic
factors, there can be no doubt that the predicted result should be
a much higher mean IQ for the upper SES Negro children than
for the lower SES white children. An ideal study along these lines
has not yet been done; it would involve obtaining IQs of both
parents of every child and making a prediction of the child’s IQ
based on a genetic model. Since there is some regression toward
the population mean from parent’s IQ to child’s IQ, a genetic
theory of the racial intelligence difference would predict that
Negro and white children should regress toward different population
means. In the two SES groups we are considering here, the
regression would be in opposite directions: the children of the
high SES Negro parents would on the average regress to some
degree downward toward the Negro population mean, and the
low SES white children would regress upward toward the white
population mean. Because the Negro population mean is about
one standard deviation below the white population mean, the
mean IQs of our two hypothetical groups of children would be
much closer together than if we compared the mean IQs of low
and high SES white children, and this should be so, according to
our genetic hypothesis, even if the high SES Negro and white
parents were perfectly matched on IQ. The conformity of actual
data to the predictions from this genetic model will, of course, be
attenuated to the degree that the parents’ and offsprings’ environments
have been dissimilar with respect to factors influencing
mental development.
Facts relevant to this hypothesis have been summarized by
Shuey (1966, pp. 519-20): ‘Where Negro pupils have been compared
with whites of the same occupational or socioeconomic class