American Scientist

The Big Test: The Secret History of the American Meritocracy. Nicholas Lemann. 406 pp. Farrar, Straus and Giroux, 1999. $27.

In 1933, Harvard's new president, chemist James Bryant Conant, informed two assistant deans of his plan to create a genuinely academic undergraduate environment at Harvard. Conant wanted serious scholars, not the idle rich young men who skipped classes, whose grades were "gentlemen's Cs" and whose lives revolved around parties and sports. Conant's first step was to establish scholarships based on academic merit as opposed to financial need. But how could one identify academically outstanding young men from the nation's public schools? Existing entrance exams consisted mainly of essays on classical subjects (Latin and Greek) that dominated the eastern boarding school curriculum, but which were barely taught in public high schools.

One of the deans was Henry Chauncey, who 12 years later would become the first president of the Educational Testing Service and whose ancestor, Charles Chauncy, was president of Harvard in the 1650s and 1660s. Henry Chauncey realized that the then young field of mental testing—testing for ability as opposed to knowledge—provided a solution to the selection problem. Thus began the chain of events that would lead to the pivotal role of the Scholastic Aptitude Test (SAT) in admissions at selective colleges.

Chauncey's and Conant's goals went well beyond the reform of Ivy admissions. They sought to forge a system in which the most talented people from all backgrounds could rise to the top. They wanted to nurture Jefferson's "natural aristocracy of talents," to (in Conant's words) "use the powers of government to reorder the haves and have-nots every generation to give flux to our social order.? Abilities must be assessed, talents must be developed, ambitions guided.? This is the task for our public schools." Although SATs prospered as higher education was expanding, Conant's concern was not that too few students went to college but that too many of the wrong ones did so.

Conant did not foresee in the 1930s that America would ever need the talents and skills of millions of college graduates and holders of advanced degrees. His goal was to identify and nurture quasi-Platonic guardians, to seek the best from among the masses, not to expand education and opportunity.

Conant and Chauncey were in some respects men of their times with limited prescience. The Big Test could still have been an essentially admiring account of how a small group of elite Eastern Episcopalians promoted a scientifically based conception of merit that recognized and encouraged talent from immigrant groups, from non-Europeans and from those who lived in the American hinterlands, and unintentionally facilitated both higher education for women and the expansion of higher education.

Instead, Lemann's tone and theme are set by his conspiratorial subtitle: The Secret History of the American Meritocracy. As described by Lemann, the SAT was promoted by a small, critically placed group of descendants of 17th-century colonial clergy for whom it was "the moral equivalent of religion." These righteous men were animated by their ancestors' Puritan impulse to establish and control a seemingly virtuous but oppressively strict social order. They created a "new kind of class system even more powerful than the old."

Lemann often acts as if SAT scores determine who can go to college as opposed to which colleges a student can attend. He exaggerates the lifetime benefits of attending an elite college as opposed to a moderately selective one. The Big Test does not consider the empirical evidence that justifies ability testing. (Lemann's other writings show that he is aware of validity issues.) He grudgingly concedes (though often forgets) that SATs are useful predictors of college grades, but he sees academic ability as only a small aspect of merit. Even if SATs were relevant only to college, not to life, it would still be reasonable for colleges to select students mainly on the basis of the best available indicators of their ability to do college work.

Lemann distrusts elites and argues that they will inevitably seek to perpetuate their own kind. But his main objection to testing is suggested by his focus on affirmative action. Non-Asian minorities do get appreciably lower average scores than whites. But Lemann fails to acknowledge that when used to predict grades, SATs not only do not discriminate against African Americans, they predict higher average grades than are received.

Lemann includes an account of the contemporary anti-testing movement that was spearheaded by the Nader-affiliated Public Interest Research Groups in the late '70s. Although visceral opposition to testing has grown since then, selective colleges have become more dependent on SATs, because high school grade inflation has reduced the reliability of grades and class rank. Until very recently, high school grades have been better predictors than SATs of college grades. (Colleges, of course, use both.) According to the latest data, SATs now predict better than high school grades. Even if SATs had less predictive validity than they do, national colleges would need some means of comparing students under uniform conditions.

Lemann would like to replace the SAT with subject-matter tests of a yet-to-be-determined national curriculum—a proposal that is politically unrealistic and probably also technically so. We cannot agree on a national high school curriculum. Even if we could, a politically acceptable test based on a least-common-denominator education would be (like the Regent's exams in New York) too easy to reliably distinguish the most talented from the above average.

Still, there is a serious downside to the SAT. Time spent on practicing SAT-type questions would be better spent studying subject matter. The curriculum can be distorted by a focus on SATs. An early tester, Carl Brigham, wrote in 1938:

If the unhappy day ever comes when teachers point their students toward these examinations ... we may look for the inevitable distortion of education [by] tests ... mathematics will be ... broken into disintegrated bits ... the sciences will become highly verbalized ... languages will be taught for linguistic skill without reference to literary values ... practice and drill in the writing of English will disappear.

Although Lemann's test based on a national curriculum is unworkable, selective colleges could give less weight to SATs if they gave more weight to subject-matter tests like the Advanced Placement tests that enable students to get college credit for high school classes. Many more students today can take AP classes than had access to the classical subjects that dominated Harvard's pre-1930s entrance exams. Unfortunately, AP classes are more available in well-to-do suburban schools than in inner-city or rural schools.

On balance, SATs serve indispensable functions and will do so for the foreseeable future.—Malcolm J. Sherman, Mathematics, The University at Albany, SUNY