nce again Charles Murray has caused a national sensation
with a new book, The Bell Curve, co-authored with Richard
Herrnstein (unfortunately deceased in September, 1994).
Murray's earlier book attacking social programs, Losing
Ground (1984), and his 1993 article in the Wall Street
Journal attacking the welfare system, also attracted
widespread media coverage, though this new work tops even the
earlier efforts in publicity. Murray has appeared on nearly
every national media outlet from Larry King Live to National
Public Radio. Most newspapers and magazines have featured
stories, often with front page coverage. Where Murray's work
does not appear is in scholarly journals--significantly so.
If there were such a thing as social science malpractice, these authors would be at risk. Not because of the topics they address, but because their analyses are inadequate and their arguments biased, even deceptive at times. More technically, what Herrnstein and Murray do is infer from correlational data to unlikely explanations. They manipulate the strong collinearity and covariation of test scores with education, socioeconomic status (SES), and other variables to show that intelligence matters most by arbitrarily attributing the covariation to IQ. They eliminate competing explanations, such as education, through dubious arguments and propose a single-factor explanation, cognitive ability, to account for almost everything. Of course, this simple explanation, targeted at certain groups, enhances political viability. These analyses and findings would not be acceptable in scholarly journals.
Charles Murray himself has a history of such work. In his evaluation of Jesse Jackson's PUSH/Excel program, Murray collected the pre-data half way through the program, collected post-data at the end of the next year, subtracted the wrongly collected pre-data from the post-data, and declared that the program had no effects (based on some pre- scales registering in the ninety percentiles). Whatever one thinks of Jackson's program, this is no way to run an evaluation.
In Losing Ground Murray used a non-standardized ratio of raw means rather than the standard deviation as the key statistic in the educational analysis. When the standard deviation was calculated, his argument was destroyed, even on its own terms (House and Linn, 1986). Should someone who claims a degree in quantitative methods be able to tell a standard deviation, a most elementary statistic, from raw mean scores? In the jobs analysis Murray left out key data about male and female employment, which reversed the direction of the argument when included (House and Madura, 1988).
In other analyses he made similar egregious errors. In The Bell Curve, his style has not changed. The Bell Curve Arguments Since the book is long (845 pages), it is difficult to get its arguments straight. To be fair to the authors, the book is fairly complicated. Most reviewers in the media discuss it without having read all of it or without understanding the statistical analyses and technical reviews that underpin many of the arguments. Journalists and politicians tend to credit the book with scientific validity because of its many numbers and citations, a serious error. Here is a summary of the main arguments, followed by critiques of each point:
1. H&M Claim: A new cognitive-economic elite has developed in America based on intelligence and equal access to higher education. (Evidence: Higher SAT scores at Harvard and other elite universities; correlations between test scores and job productivity, average r= .4). iStarting in the 1950s, a handful of institutions became magnets for the very brightest of each yearQs new class. In these schools, the cognitive level of the students rose far above the rest of the college population. Taken together, these trends have stratified America according to cognitive abilityd (p. 29).
Critique: There is increasing stratification in American society, but it is more economic than intellectual. The rich have gotten richer and the poor poorer, but this has happened in most European societies over a similar period as well. Common economic policies are a clear explanation, not meritocracy. For 40 years after 1930 the gap between rich and poor in America narrowed. Since the end of the 1960s it has been widening, and is greater now than at any time since the creation of the modern welfare state....a similar trend is apparent in many other industrial countries....Why has inequality increased? A large part of the answer lies with the liberal economic policies adopted in many parts of the world in the past 15 years. Many governments have cut tax rates for the rich and restrained spending on benefits for the poor (Economist, 1994, p. 13). One does not need to invoke IQ to account for these changes. In any case the correlation between test scores and job productivity accounts for only 16% of the variance, even if taken at face value, leaving job success open to many influences other than IQ. Also, measured IQ varies hugely within occupations; there is no simple IQ cut-off score that determines success.
2. H&M Claim: Intelligence (Spearman's g) is essentially inherited and not subject to social or educational remediation. (Evidence: Correlational twin studies and Jensen's work, average r= .6). "Putting it all together, success and failure in the American economy, and all that goes with it, are increasingly a matter of the genes that people inherit....Still further, we know that the correlation between intelligence and income is not much diminished by partialing out the contributions of education, work experience, marital status, and other demographic variables" (p. 91, 97).
Critique: Intelligence is inherited to some extent no doubt. But that does not mean that environment does not play a critical role. Even the amount of variance accounted for by Jensen and identical twin studies, taken at face value, would indicate that there is much room left for environmental influences. The authors claim that environmental factors have been neutralized by 12 years of education for most students, leaving only genetic factors. But they are alone in claiming such equality of environment. The evidence is all on the other side.
Nor would psychologists working in this area accept the idea that Spearman's g is really all there is to intelligence. Intelligence is far more complex. Nor would test experts agree that tests of any kind measure mostly Spearman's g. Herrnstein and Murray use quite different kinds of tests-- SAT's, the Armed Forces Qualification Test, National Assessment, the Iowa tests, the Ravens Progressive Matrices, the Wechsler and others--interchangeably to make their case. However, achievement tests even in the same subject area measure different things, let alone IQ tests equated to achievement tests. The authors' use and intepretation of tests violates expert understanding of them.
3. H&M Claim: Since the population has been sorted out cognitively, intelligence accounts for individual successes and failures in poverty, unemployment, welfare, parenting, education, crime, and other social behaviors. (Evidence: Graphs of regression curves for two variables at a time, such as intelligence scores vs. poverty, with another variable on the same graph, such as SES vs. poverty, for white subjects only. IQ is defined by subscores from the Armed Forces Qualification Test, with other indicators from a Labor Department longitudinal survey. Similar analyses establish the centrality of IQ to several social behaviors.) "Whites with IQs in the bottom 5 percent of the distribution of cognitive ability are fifteen times as likely to be poor than those with IQs in the top 5 percent" (p. 127).
Critique: In relating intelligence to all kinds of social behavior, the authors use subscores from the Armed Forces Qualification Test, given to youths 14 to 22, as the independent variable for IQ. Other variables, such as SES and education, they dismiss as non-influential in various ways. For example, instead of using years of education as a variable in the multiple regression analyses, they split the group into high school graduates and college graduates, then show the influence of intelligence within these groups, thus eliminating education as a competing variable. If they had used years of schooling as a precise indicator, they would have discovered that education is quite important. The correlation of test scores with years of education is .64 and with SES .55 in this Labor survey.
A proper multivariate analysis would diminish the IQ relationship with other variables significantly. The authors said they did not enter both IQ and education variables into the regression equations together because they measure the same thing--another reason they should not be drawing the conclusions they do.
By visually comparing the lower tail of the regression curves at -2 standard deviations (where there is more error), they find that there is a higher proportion of low IQ than low SES people in poverty. However, this is true only at the extreme lower end of the curves. The difference is about 16% at the extreme and only 6% at -1 S. D. Above the mean the relationship is reversed, with SES counting more than IQ. If one turns to the footnotes to see what the precise quantitative relationships are, one finds, "We refrain from precise numerical estimates of how much more important IQ is than socioeconomic background" (p. 691). Actually, the R- Square is .10, meaning the relationship accounts for 10% of the variance, hardly a powerful influence, even with education excluded.
The authors attempt to explain why they do not report R- squares, estimates of the strength of the relationships, even though they are used elsewhere in the book. R-squares can underestimate strength in homogenous populations. However, the Armed Forces Qualification Test national sample is hardly likely to be highly homogenous. Furthermore, by using logistic regression analysis and chi-squared as the significance test, these very weak relationships are found to be significant because chi-squared is highly dependent on the large numbers in the survey (N=3367 for this particular analysis). Nor do the authors discuss measurement error in the AFQT test, SES, poverty, or other indicators. Nor do they use confidence intervals to give readers a sense of the extent to which they are operating in a gray area statistically.
From these very inadequate analyses the authors conclude, "Cognitive ability is more important than parental SES in determining poverty" (p. 135), and "The high rates of poverty that afflict certain segments of the white population are determined more by intelligence than by socioeconomic background" (p. 141). Need we say that correlation does not mean causation, not even in such flawed (and deceptive) analyses? There are eight chapters like this, set up to show that intelligence is the important variable in much the same misleading fashion.
One can also dispute these findings from common sense. If crime is caused primarily by low IQ, as the authors contend, why is the murder rate in Britain less than one-tenth ours? Does this mean that the British are that much smarter than we are? Fixed intelligence should have the same results, according to Herrnstein and Murray. If we pursued their method of argument, we could scale how much dumber we are than Europeans by aligning our crime rates with theirs. Perhaps only the less intelligent emigrated from those countries to the US.
Or, conversely, the illegitimacy rates for Britain and France are higher than in the US at the moment. Perhaps the most intelligent people emigrated after all. These are the kinds of arguments that the authors make throughout the book. How ridiculous they are is evident when one moves away from black-white comparisons, where differences traditionally have been attributed to intrinsic personal characteristics. Furthermore, given fixed intelligence, crime rates should not vary much. But crime rates do vary. The alternative explanation that environmental factors play a major role in intelligence, crime, and other social behaviors is far more plausible.
4. H&M Claim: Intelligence is not only inherited but differentially distributed in ethnic and racial populations, with Asians on top, followed closely by whites, and blacks at the bottom. (Evidence: Scores on tests showing different ethnic and racial distributions from Jensen's work and test surveys like the SAT and NAEP.) The authors posit a gap of one standard deviation between white and black scores and concede that the gap has narrowed somewhat over the past two decades, but contend that convergence of scores will be limited because of innate genetic differences. "Given the cognitive differences among ethnic and racial groups, the cognitive elite cannot represent all groups equally...." (p. 315).
Critique: Claiming that blacks are inherently inferior to whites by showing differences in various kinds of test scores has garnered the most media attention, though Murray disingenuously suggests that he is surprised by the publicity since only a few chapters are devoted to the topic. Actually, in chapters comprising 239 of the 552 pages (43%) of the main text, black-white comparisons are made extensively. Even when blacks are not mentioned, the book is set up to support the genetic argument.
As for evidence on the genetic link to IQ, there is nothing new in the Herrnstein-Murray argument. As they say, there is a difference in test scores among racial groups, though the gap has narrowed. The genetic issue cannot be settled by correlations among tests scores. We note the opinion of Robert Linn, a leading test expert: "Tests are indeed culture dependent. This culture dependence clearly makes some inferences inappropriate. For example, a mean difference between groups on a test provides no basis for inferring that one group is inherently inferior to another. The results merely describe a current state of affairs and may be useful in drawing certain inferences regarding likely behaviors in other situations" (Linn, 1982, p. 366).
In our opinion, the permanent resolution of the genetic issue must await the development of biological theory so that there is an empirical basis, a truly scientific explanation. Speculations about test correlations by psychologists who have no biological basis for making these judgments can never resolve the issue. The current opinion of geneticists about intelligence is in the opposite direction from the Jensen- Herrnstein-Murray thesis.
5. H&M Claim: When one controls for IQ, blacks (and sometimes Latinos) are better compensated in wages, education, and over-represented in illegitimacy, poverty, etc. (Evidence: Graphs showing blacks and Latinos have better wages, jobs, etc., when one controls for equivalent intelligence, e. g., blacks with IQs of 105 have better salaries than whites with the same IQ).
Critique: Evidence showing the overcompensation of blacks and Latinos on salaries and their overrepresentation in higher education is based on comparing whites, blacks, and Latinos with the same test scores. Since the authors insist on intelligence influencing SES and education, not the other way around, the covariation attributable to IQ, education, SES, and other things is usually attributed to IQ alone. Equating salaries and other behaviors on the basis of IQ only makes sense if one believes that intelligence is the determinant of merit, which the authors never prove.
6. H&M Claim: The US is experiencing "dysgenic" pressures in that low IQ women are having more children, and many immigrants entering the country have low IQs. Latinos play a major role, with 64% of children born to Latino mothers having IQs less than 90, according to the authors. The Latino-white IQ gap is about the same as the black-white gap. Hence, there is downward pressure on average intelligence. (Evidence: Graphs showing numbers and percentages of children born to women of less education and lower IQs. Mean IQ of those born in US is .4 S. D. higher than foreign born.) "...evidence indicates that the self-selection process that used to attract the classic American immigrant--brave, hard working, imaginative, self-starting, and often of high IQ-- has been changing, and with it the nature of some of the immigrant population" (p. 341). "So far, we have been treating the distribution of intelligence as a fixed entity" (p. 342). "But within the scholarly community, there is little doubt about differential fertility or about whether it is exerting downward pressure on cognitive ability" (p. 348).
Critique: Arguments that the US is suffering from low IQ immigrants and the over-breeding of the less intelligent hark back to early in the century when psychologists met immigrants at Ellis Island, gave them IQ tests, and declared many of them--Jews, Italians, Greeks, and Hungarians-- mentally retarded (Gould, 1981). Many were deported and some states passed sterilization laws, during a time of immigration and economic pressures similar to today's. Herrnstein and Murray never argue for such drastic measures. They just want to take away social benefits from the genetically inferior. We can only express surprise and dismay that biological racism has emerged once again.
7. H&M Claim: Education and early training cannot change intelligence to any significant degree. "Formal schooling offers little hope of narrowing cognitive inequality on a large scale in developed countries because so much of its potential contribution has already been realized with the advent of universal twelve-year systems.... Head Start...does not improve cognitive functioning" (p. 389). (Evidence: Review and dismissal of educational programs showing gains in test scores, e. g., Head Start and the Perry School Project). Hence, education offers no way out for those at the bottom. "For many people, there is nothing they can learn that will repay the cost of the teaching" (p. 520)
Critique: Education and educational research are critical to the argument of the book. If test scores can be raised by education, then cognitive ability is not fixed and efforts should be made to improve education, the opposite of what the authors conclude. Essentially, they argue that every study that shows significant test gains is flawed technically or non-definitive. They assume that educational environments are uniform across the country. Of course, the readers of this journal know there are any number of well conducted studies that show substantial test score gains, that the cognitive gap between whites and blacks has narrowed, and that K-12 educational environments are far from uniform.
To take only a few cases, the Husen-Tuijnman (1991) study tracing the effect of schooling over many years is summarized by Herrnstein and Murray as saying that IQ at the age of 10 was five times more important than years of schooling. In fact, the study concludes, "The results provide support for the thesis...that IQ as measured by group intelligence tests is not stable but changes significantly between 10 and years of age....[The] argument that the amount and quality of schooling experiences to which children are exposed are implicated in the observed changes in measured IQ is also confirmed" (Husen and Tuijnman, 1991, p. 22). The study's statistical analysis methods are far superior to those in The Bell Curve.
Head Start is dismissed because the test gains disappear by sixth grade, which is quite different than saying there are no gains. The Perry Preschool Program is similarly dismissed for disappearing test gains, even though there are other lasting social outcomes, such as less illegitimacy. As to gains from the Carolina Abecedarian Project, these are dismissed because perhaps the experimental and control groups were not initially equivalent. Herrnstein and Murray suddenly discover lack of methodological rigor--but only in other people's work.
Reluctantly, they admit that there is some possibility of improvement: "As of now, the goal of raising intelligence among school-age children more than modestly, and doing so, consistently and affordably, remains out of reach" (p. 402). But "modestly" and "affordably" are quite different qualifications than saying cognitive ability is fixed. As for all the studies showing gains, they say "...in the technical journals, the good news is viewed with deep suspicion and discounted....The fact is we and everyone else are far from knowing whether, let alone how, any of these projects have increased intelligence. We write this pessimistic conclusion knowing how many ostensibly successful projects will be cited as plain and indisputable evidence that we are refusing to see the light" (pp. 409-410).
In this they were prophetic. Social scientists of every persuasion have contested the authors' refusal to acknowledge a large literature showing cognitive improvement under certain circumstances, including even economists who insist on direct economic benefits from schooling. James Heckman, a Chicago economist, said, "Murray and Herrnstein just ignored the evidence. They are 180 degrees off course" (Passell, 1994, p. B10). Michigan psychologist Richard Nisbett, noting dismissal of the narrowing of the black-white achievement gap, said, "This is not dispassionate scholarship. It is advocacy of views that are not well supported by the evidence...." (Nisbett, 1994, p. 15).
There is no need to belabor the point with an audience of researchers. What is true is that there are no guaranteed nor standard methods of cognitive improvement and that many of the best educational programs are more costly than the public is likely to pay for on a massive scale. What is not true is that improving cognitive ability is impossible. Thus this critical educational link in the overall Herrnstein- Murray argument is simply wrong in the face of overwhelming evidence.
8. H&M Claim: American education has been "dumbed down" in order to accommodate mediocre and low IQs. (Evidence: SAT score declines.) The schools have done about as well as could be expected, given the innate lack of ability of many students, but the schools have also slowed down the talented because of "multiculturalism." "The problem with American education is confined mainly to one group of students, the cognitively gifted" (p. 417). Those at the top should be provided a better education because they are the ones who produce the goods and benefits for society.
Critique: In their pitch for education for the gifted, the authors present an intellectual trickle-down theory. "It should be said openly: The people who run the United States-- create its jobs, expand its technologies, cure its sick, teach in its universities, administer its cultural and political and legal institutions--are drawn mainly from a thin layer of cognitive ability at the top. (Remember--just the top 1 percent of the American population consists of 2.5 million people)" (p. 418). The other 250 million will be surprised to hear this. (Throughout the book the authors assert that the readers themselves are in this cognitive elite).
In spite of such silliness, in this chapter there are actually some conclusions that fit the available evidence (NAEP, SAT, and Iowa test scores over time). The authors conclude that education is about the same quality for the average student as it has always been, contrary to common opinion--which is not to say that it is good enough. They even acknowledge changes in test construction, which may account for some changes in test scores. Apparently, a test is not a test after all. They attribute the drop in SAT- verbal scores to the educational system idumbing downd the elite.
This dumbing down occurred when standards were lowered because of "multiculturalism" and the necessity of minimizing racial differences in performance, according to the authors. What about the simultaneous increases in SAT-math scores? Apparently, math courses were added and toughened. If readers of this review have detected an inconsistency in that test scores measure intelligence accurately, which can't be changed much, yet the authors spend considerable time explaining changes in test scores, that is because this internal contradiction runs throughout the book. When the argument requires that cognitive ability can't be changed, as in genetic endowment, it can't be. When the argument requires explaining changes, it can be changed, such as dumbing down through inadequate education. In fact, the book is filled with inconsistencies throughout, assertions of different things in different places.
Actually, we agree with the assertion that we need better programs for the gifted, though not with the argument put forth by the authors, which is likely to give the cognitively gifted a bad headache. In general, academically talented students are not well challenged in American schools.
9. H&M Claim: Affirmative action policies and programs in higher education are ill-conceived and pernicious. Blacks are receiving grossly unfair advantages over other groups. (Evidence: Gaps in SAT and other entrance test scores between whites, Asians, and blacks at elite universities.)
Affirmative action policies in the workplace are equally unfair. Cognitive ability has been shown to be linked to job performance. General cognitive tests are good predictors of job performance and should be used for hiring. (Evidence: Black-white IQ differences in job categories; test score and job productivity studies.) "...since the early 1960s blacks have been overrepresented in white collar and professional occupations relative to...IQ...." (p. 479).
Critique: In arguing against affirmative action Herrnstein and Murray base much of their argument on meta-analytic studies by Hunter and Schmidt, which show that IQ is the central determinant of job productivity. When we informed Gene Glass, inventor of meta-analysis, of these findings, he said the Hunter and Schmidt studies were not credible (Personal communication, 11/9/94). In fact, Murray makes much of meta-analysis as the new statistical technique which has led to ibreak-throughd understandings (p. 71). His explanation of the origins of meta-analysis is wrong (p. 675). Certainly, there are legitimate questions to be raised about affirmative action, but reconsideration is not helped by these claims.
10. H&M Claim: Given the meritocratic stratification of American society based on intelligence, what policies should be adopted? There are many possibilities but the central ones all involve the underclass. We fear that a new kind of conservatism is becoming the dominant ideology of the affluent...conservatism along Latin American lines, where to be conservative has often meant doing whatever is necessary to preserve the mansions on the hills from the menace of the slums below....The most likely consequence in our view is that the cognitive elite, with its commanding position, will implement an expanded welfare state for the underclass that also keeps it out from underfoot. Our label for this outcome is the custodial state....In short, by custodial state we have in mind a high-tech and more lavish version of the Indian reservation for some substantial minority of the nation's population.... (pp. 518, 523, 526).
Given the inevitability of this scenario, the authors recommend the following policies:
--A wide range of social functions should be restored to the neighborhood. Government can do much to foster the vitality of neighborhoods by trying to do less for them
--The federal government should support greater flexibility for parents to send their children to schools of their choosing. Federal scholarships should reward academic performance. Some federal funds focused on the disadvantaged should be reallocated to programs for the gifted.
--Social and legal rules should be made simple so the unintelligent can understand them, e. g., make simple rules about crime and punish offenders. This will help the less intelligent learn a simple morality they can grasp. Also, simplify rules about marriage and childrearing.
-- Supplement the income of the poor, since it is not their fault they are born less intelligent. No concrete plan is advocated.
-- End support for low IQ women to have babies, e. g., welfare programs that reward illegitimacy.
--Change immigration policies to allow only the competent into the country.
--If government programs are needed, target the small portion of the population that causes the social problems. -- Treat people as individuals, not as members of groups, e. g., in affirmative action.
And there is this final judgment: "Cognitive partitioning will continue. It cannot be stopped because the forces driving it cannot be stopped....Inequality of endowments, including intelligence, is a reality. Trying to pretend that inequality does not really exist has led to disaster. Trying to eradicate inequality...has led to disaster. It is time for America once again to try living with inequality...." (p. 551).
Critique: The final chapters on policies for dealing with meritocratic stratification amount to standard conservative prescriptions--schools of choice, educational vouchers, programs for the gifted, less for the disadvantaged. Abolish welfare, establish more severe penalties for crime, reduce government programs, and so on. There are other rationales for these policies, not based on the quicksand of The Bell Curve. We will not discuss the policies other than to note that many would make cognitive stratification worse, not better.
In summary, the book claims that the social ills of society are caused by intrinsic characteristics of unintelligent people, who cannot be helped very much. We all must learn to live with our innate inequalities, which also means the superiority of the cognitive elite. According to the authors, those of us in privileged positions should not feel guilty about what nature has dealt us, but perhaps feel a little sorry for the unfortunates who were not born with our brains. Ultimately, this book is a justification of the economic stratification that has developed over the past decades in American society.
The United States is now the most economically unequal of all industrial countries. The ratio of the income of the richest 20% of households to the lowest 20% is about 11 to 1, compared to 4 and 5 to 1 in Japan and Germany, which are among the most equal. Furthermore, inequality of income is strongly negatively related to increases in labor productivity and growth in GDP per capita (Economist, 1994, pp. 20-21). That American society is developing a more rigid class structure, we would agree, though not about the causes nor solutions that Herrnstein and Murray advance.
Finally, what can one say about the book as a social
phenomenon which has swept American society? Clearly, the
book has a receptive audience. Under pressure from
immigration, crime, a declining economy, lower paying jobs,
and an unresponsive government, many Americans will find in
this book a ready explanation for their difficulties and the
unhealthy direction of their society. Putting the blame for
social stratification and deterioration on the unintelligent,
especially minorities and immigrants, is convenient
scapegoating. The social circumstances are similar to those
in past episodes of biological racism in American history.
As for Murray, in spite of his deceptive arguments, he has discovered one great truth. He knows that he can present these incorrect findings in the popular media, that the media will carry them extensively, and that the social science community cannot catch up to him. Although critics will strip his arguments of merit in scholarly journals, the critiques will be too technical, too complex, and too late to be given prominence in the media, so that his arguments will continue to be interpreted as one side of a liberal- conservative dispute, and his opinions will be given equal or more weight than the entire body of scientific opinion.
As former Governor Lamm of Colorado responded in 1985 when we told him of the critical statistical error that destroyed Murray's educational argument in Losing Ground, "Oh, you academics are always carping about something."
On November 9, 1994, Murray was interviewed by Mike Rosen, the right-wing talk show host in Denver, who said that the findings of Losing Ground are now conventional wisdom. Murray agreed and said that in ten years, those of The Bell Curve would be also, partly because he had used a sophisticated statistical technique called meta-analysis to show that what counted most in job productivity was IQ. Of course, he had done nothing of the sort.
Not all people are fooled. Some manage to penetrate the scientific veneer. For example, John McLaughlin, the bumptious, conservative TV journalist, called the book, "pseudo-science." And he is exactly right. But this is an exception. For the most part there is no effective way to counter these false claims in the media. If social science were physics, this book would not last longer than cold fusion claims from Utah. But, alas, it is not, and Murray has turned the trappings of social science against social understanding. This is the great truth that Murray has learned, and he has employed it again for his own fame and fortune, to our societal detriment.
Note 1: This paper will appear in Educational Evaluation and Policy Analysis, summer, 1995. The opinions in this review are our own, but we express our appreciation for suggestions to Ken Howe, Margaret Eisenhart, Phil Langer, Michael Meloth, Colby House, Lorre Shepard, and Steve Guberman.
Economist. (1994). Slicing the cake. 333: 7888, 13.
Gould, S. J. (1981). The mismeasure of man. New York: W. W. Norton.
Herrnstein, R. J. and Murray, C. (1994). The bell curve: Intelligence and class structure in American life. New York: Free Press.
House, E. R. and Linn, R. L. (1986). Book review of Losing ground: American social policy, 1950-1980. Educational Evaluation and Policy Analysis, 8: 3, Fall, 324-328.
House, E. R. and Madura, W. (1988). Race, gender, and jobs: Losing Ground on employment. Policy Sciences. 21: 351-382.
Husen, T. and Tuijnman, A. (1991). The contribution of formal schooling to the increase in intellectual capital. Educational Researcher, 20:7, 17-25.
Linn, R. (1982). Ability testing: Individual differences, prediction, and differential prediction. In A. K. Wigdor and W. R. Garner (Eds). Ability testing: Uses, consequences, and controversies. Washington, DC: National Academy Press. 335- 388.
Murray, C. (1984). Losing Ground: American Social Policy, 1950-1980. New York: Basic Books.
Nisbett, R. (1994). Blue genes. New Republic, Oct. 31, 15.
Passell, P. (1994). "Bell Curve" critics say early I. Q. isn't destiny. New York Times, Nov. 9, B10.