ITHACA, N.Y. -- Intelligence test scores of Whites compared with African Americans, and of the members of high compared with low socio-economic groups, are not growing ever wider. This is contrary to often-reported arguments that Americans are getting dumber because low-IQ parents are outbreeding high-IQ parents. Rather, upon closer look, these scores point to a growing convergence, report two Cornell University developmental psychologists who are experts in intelligence assessment and types of intelligence.
In comprehensive analyses of national data sets of mental test scores (including tests containing verbal analogies, vocabulary, mathematics, science, writing and spatial reasoning) for American students, Wendy M. Williams and Stephen J. Ceci, both in the Department of Human Development at Cornell, write in the November 1997 issue of the scholarly journal American Psychologist that "there is no compelling evidence supporting the hypothesis that a dysgenic (negative) trend is at work, undermining Americans' intellectual capital." Williams is an associate professor of human development, and Ceci is the Helen L. Carr Professor of Developmental Psychology, both in the College of Human Ecology at Cornell.
Such dysgenic trends received wide publicity in 1994 with the publication of The Bell Curve by R.J. Herrnstein and C. Murray, which argued that demographic pressures were exerting such strong downward influence on cognititve ability in this country that dire negative social consequences would ensue.
Williams and Ceci, however, re-examined the data and looked at a broader base of tests that reflect cognitive ability; they focused on batteries of mental tests that are statistically comparable to IQ tests, as well as on IQ tests themselves. Their findings do not support suggestions put forth in The Bell Curve. Specifically, they report:
-- Racial differences in intelligence narrowed by about half between 1970 and 1988 and have since remained fairly constant;
-- Socio-economic (SES) class differences between the upper and lower thirds have continued to decline gradually since 1932. IQ scores between these two groups have, in fact, converged about 25 percent in the past 50 years or so.
-- PSATs (Preliminary Scholastic Assessment Tests), which are given to representative samples of high school juniors (as opposed to the self-selected high school seniors who take the SATs) reflect relatively stable score differences between the top and bottom quarters since 1961.
"It's true that SAT scores have declined rather precipitously in recent years," Williams concedes. "Yet, there are good reasons for the decline having nothing to do with dysgenic trends. Specifically, more high school seniors take the SATs now because more colleges require it. In the past, the group of seniors who took the test were self-selected, smaller and more elite."
And she points out, while SAT scores were declining, PSAT scores were climbing and were not influenced by sampling fluctuations due to self-selection. (Since 1965, virtually all high school juniors have taken the PSAT.)
"The most dramatic convergence of scores, however, has been in racial differences in IQ during the 1970s and 1980s, which we think were due to non-genetic factors," Williams said. "During this period, our country spent substantially more for education targeted at minorities, and at the same time, the educational attainment of minority parents increased enormously and their families grew smaller."
Williams and Ceci point out that the closing of the racial gap was due to gains in test scores by black students, not from reductions in white students' scores. The stability of the racial gap since the late 1980s is of concern since the gap has not closed further; but there is no evidence for "the suspicion that dysgenesis is taking place," the Cornell psychologists write. "If there are dysgenic pressures at work in society, then they are being more than offset by the positive pressures," says Ceci.
Williams and Ceci also examined changes in the average scores between brightest and dullest students to determine whether the gap between students at the top and bottom of the distribution is growing. Despite the more diverse pool of students taking the SAT today -- i.e., a less selective, more economically diverse population -- the gap between the top and bottom groups has not widened. Neither the top nor the bottom has varied significantly in either direction since 1976.
"In sum, we have found no compelling evidence supporting the hypothesis that dysgenic pressures are at work, undermining Americans' intellectual capital," the Cornell intelligence experts conclude. "In the battle to control the hearts and the minds of American educational policy-makers and opinion shapers in the mass media, it seems imperative that the combatants temper their passions with a close inspection of the data."