The IQ Controversy: The Media and Public Policy.
by Mark Snyderman and Stanley Rothman.
Transaction Books. 310 pp. $24.95.
Not many Americans know a great deal about intelligence testing, but strong opinions about it are rampant. The opinons are overwhelmingly negative. Reflexive hostility to IQ tests is the norm among humane and liberal-minded members of the educated classes, which means it is also the preferred wisdom of the American media.
The source of the hostility is fairly obvious. Anyone accepting the validity of IQ tests is instantly in the grasp of an unlovable proposition: people are not equal in mental ability. Worse, groups are not equal in average ability. The rich have more mental ability than the poor. Especially troublesome in an age of affirmative action, blacks and Hispanics score significantly lower than whites on average. Whites in turn score a bit lower than Asian Americans, which also counts as bad news in many precincts. Men and women have the same average scores, but men tend to be somewhat overrepresented at the tails of the distribution curve, meaning that more of them are in the high-talent ranges above, say, IQ 140. (Men are also overrepresented among mental defectives, a fact that does not mollify feminist critics of IQ testing.) An irresistible implication of these data: many of the advantages flowing to the most privileged members of American society may have been legitimately “earned.”
If you are committed to egalitarian values, this is exactly what you do not want to hear. So you yearn to believe that IQ data are somehow fraudulent, or that they really have nothing to do with intelligence, or that the group differentials they point up reflect only environmental (and therefore readily changeable) effects. The media work hard at satisfying the demand for these conclusions, and in the process tell us a lot of things about the IQ that are not so.
A major purpose of The IQ Controversy is to talk back to the media about intelligence testing. Stanley Rothman, professor of government at Smith College, is also the coauthor of The Media Elite and thus a combat veteran of media back-talk. Mark Snyderman is a Harvard-trained psychologist now studying law at the University of Chicago. They say their research grew out of a concern “that the views of the relevant expert community are reported inaccurately to the attentive public by the elite media.” Snyderman and Rothman are not, of course, the first scholars to express dismay about this reporting. In 1982, Richard Herrnstein complained in the Atlantic about his own infuriating experiences with editors and producers determined to misrepresent intelligence testing. The IQ Controversy has developed this case much more systematically.
The authors have undertaken several related tasks. They have performed a huge content analysis of media coverage of the IQ controversy during 1969-83. They have also taken an exhaustive survey of “the experts.” Along the way, they have produced a moderately technical, moderately readable history of the IQ and a guide to several issues central to the controversy. Their discussion rambles all over the lot, but the book’s main thesis is clear enough: the media point of view about intelligence testing is very much at odds with the beliefs of knowledgeable scholars.
The authors’ content analysis covers the New York Times, the Washington Post, the Wall Street Journal, the weekly newsmagazines, and the network news programs. Most of the news coverage is scored as “neutral,” i.e., nobody was pushing a point of view. But where judgments were offered, certain themes surfaced quite regularly. One staple of media analysis is the “cultural bias” of IQ tests and the possibility that they “are largely a measure of exposure to white middle-class culture.” This thought, which implies that the tests are totally fraudulent, was found 24 times in the Times, 12 times in the Post, 3 times in the Wall Street Journal, 15 times in the newsmagazines, and 9 times on the networks. The alternative view—that the most popular tests are basically fair—ran a very distant second.
Another major media preoccupation has to do with the way students are classified after they take IQ tests. A central purpose of the testing is of course to ensure that students work at appropriate levels and get remedial instruction if necessary. But the media seem magnetized by an opposing thought: that students may be misclassified or stigmatized because of their IQ scores. The analysis found 39 articles or broadcasts arguing this proposition and none resisting it. There were also 26 articles arguing that tests promote racism or inequality, none opposing this view.
The media have done worst of all with the nature-nurture issue. They have repeatedly argued that differences in IQ mainly or totally reflect environmental factors. Typical were Time’s comments after Japanese Prime Minister Nakasone’s disparaging remarks in 1986 about the intelligence of American minority-group members. Said the magazine: “Most scholars today believe that so-called intelligence and achievement differences stem largely from environmental factors.”
Summarizing what they have learned about media coverage of this issue, Snyderman and Rothman write:
Journalists have . . . clearly misrepresented the views of the relevant scientific community as to the interaction between genetic and environmental factors in explaining differences in IQ. . . . One would be forced to conclude from reading the newspapers and newsmagazines and watching television that only a few maverick “experts” support the view that genetic variation plays a significant role in individual or group differences, while the vast majority of experts believe that such differences are purely the result of environmental factors.
Scholars who have elaborated the genetic case for a wide audience—the most famous is Arthur Jensen—have not only been characterized as extremists but smeared as racists. The content analysis turned up 17 references to Jensen as a racist. Meanwhile, Leon Kamin of Princeton keeps being cited (in the Time article on Nakasone, for example, and in a CBS special report on “The IQ Myth”) as a respected authority figure. Kamin, it happens, really is an extremist. A Marxist who views the IQ as an instrument of class oppression, he is almost alone among academics in arguing that environmental differences may well explain all IQ variation.
The book’s survey of the experts is a major event. The survey is based on responses from 661 scholars, all members of professional organizations specializing in IQ-related issues, who filled out the questionnaires with varying degrees of completeness. (They had been instructed not to answer questions outside their own areas of expertise.) The tabulated responses are valuable not because experts are always right, which they obviously are not, but because their views give non-experts a baseline difficult to ignore. From now on, it will presumably be much harder, or at least more disgraceful, for the average uninformed anchorman to pass along as received truth the environmental extremism that has always come naturally to him when he gets around to intelligence testing.
Boring in on the nature-nurture issue, Snyderman and Rothman asked the experts to evaluate five different kinds of evidence that have been cited as reasons for believing in some genetic contribution to IQ differences. Of those who responded, 94 percent found one or more of the reasons persuasive. The single most compelling reason, checked by 84 percent of the respondents, was the barrage of studies on identical twins reared apart. Since identical twins are genetically indistinguishable, those who are separated at birth or soon afterward offer a unique opportunity to sort out the relative effects on IQ of heredity and environment. The studies typically show that the IQ’s of the separated twins correlate powerfully—far more so than, say, the IQ’s of ordinary siblings reared apart. Especially devastating to the extreme environmentalist position are studies contrasting the IQ’s of identical twins reared apart and fraternal twins reared together. The fraternal twins have similar environments but are different genetically (only about half their genes are shared); the identical twins have different environments and similar genes. Yet the studies show the IQ’s of the separated identical twins to be more powerfully correlated than those of the fraternal twins growing up together.
So there is an overwhelming case for some heritable component in IQ. How strong is the component? The average of the answers supplied by the experts: .596 for the white population, .571 for blacks. Within both groups, in other words, genetic factors are thought to account for almost 60 percent of the variability in IQ. What about the possibility that genetic factors also play a role in explaining the difference between blacks and whites (about 15 IQ points on average)? Arthur Jensen has been fiercely attacked, and not very loudly defended, for postulating that a genetic component probably explains part of the gap. It emerges, however, that Jensen has a surprising number of closet supporters among the experts. Snyderman and Rothman report that a plurality (45 percent) of all the experts, and a majority of those who responded to the question, believe that the black-white gap reflects genetic as well as environmental factors. A majority (55 percent) of the experts believe that genetic factors also help to explain socioeconomic differences in IQ.
A popular criticism of IQ tests is that they do not really measure intelligence—that, indeed, the tests have been constructed and administered in the absence of agreement about the nature of intelligence. To some extent, the experts accept this criticism. A 54 percent majority believe that IQ testing still lacks “an adequate theory of intelligence.” On the other hand, most of the experts seem to feel they are at least talking about the same kinds of behavior when they invoke the word “intelligence.” Asked to list its major constituents, virtually all agreed that these should include an ability to reason abstractly, to solve problems, and to learn. Most also agreed that the mental abilities they deemed crucial to intelligence are pretty well measured by IQ tests. Of the ten abilities checked as important by 60 percent or more of the experts, only two—“adaptation to one’s environment” and “creativity”—were held by a majority to be not adequately measured on the tests.
The experts are also inclined (58 percent) to believe that IQ tests measure something that can reasonably be labeled “general intelligence,” or g. Here again they appear to be mostly siding with Jensen, possibly the profession’s most vocal advocate of g. Those on the other side, who include Kamin and Stephen Jay Gould of Harvard, have argued that g is a mirage and that the various skills being tested cannot be meaningfully combined. (Gould, another Marxist who is cited endlessly by the media, has written that “the chimerical nature of g is the rotten core of Jensen’s edifice.”) The case for general intelligence rests heavily on the fact that all those skills being tested are positively correlated: if you score above average on one of them, you are predicted to score above average on any other, implying that some kind of general factor is at work.
Some of the experts’ answers seem surprising. Snyderman and Rothman were evidently startled by the extent to which they allow that IQ tests are biased against blacks. On a scale where 1 represents no significant bias and 4 represents extreme bias, the experts came up with a mean rating of 2.12. That cautious judgment is a long way from the persistent denigration of IQ tests in the media as totally invalidated by cultural bias. Still, the score is perplexing. In an academic environment swarming with researchers who would give an arm to prove test bias, studies persistently fail to find its existence at any significant level. The bottom line on cultural bias is predictive power: if a test does equally well at predicting the academic and job performance of blacks and whites, and if any given score predicts the same outcome for each race, then the test cannot reasonably be viewed as biased. In fact, IQ tests do well at meeting these conditions. Where there is a minor difference in predictive power, it tends to be in favor of blacks (i.e., it overpredicts their performance).
So why did the experts nevertheless insist on a certain amount of bias against blacks? Snyderman and Rothman hypothesize, plausibly, that the judgment is at bottom political. On most of the questions the experts tackled, their political perspectives (as revealed in responses to a battery of questions about social values) seem not to have influenced their conclusions. But on several of the most politically sensitive questions, especially those involving group differences, there were measurable differences in the responses of liberal and conservative experts. If I read the authors correctly, they are saying that the liberal experts were somewhat inclined to override all those studies of test validity.
A question not asked of the experts was whether group differences in IQ should be studied at all. Some highly vocal psychologists and sociologists now argue that they should not be. In some cases, the argument is advanced in a Marxist framework, with the testers cast as agents of a ruling class aiming to keep minorities and the poor in their place; in other instances, the objection seems less ideological, rooted mainly in feelings that publicizing test data is just too painful and insensitive. At its 1986 meeting, the American Psychological Association held a debate on the desirability of studying group IQ differences. One scholar on the panel argued firmly that it is wrong to engage in “research that serves anti-egalitarian purposes” and elaborately compared such studies to “hereditarian research” performed by the Nazis. Cheap shots featuring the Nazis have become a recurring problem for any scholar studying group differences in IQ.
It is hard to write a scenario in which such studies would actually disappear. But it is easy enough to see signs right now of self-censorship among the experts. Snyderman and Rothman point to indications that many are uncomfortable operating outside the restricted world of their professional journals. Many experts also seem to resent the small minority of their colleagues who have ventured outside that world and stirred up controversy in the process. Some of the survey findings in The IQ Controversy suggest that Arthur Jensen and Richard Herrnstein are resented on such grounds. Between the distortions of the media and the self-censorship of the experts, it may never be easy to overcome hostility toward intelligence testing.