The following article appeared in the 1 August 1994 edition of Newsweek and is copyrighted by them. It is reproduced here only as a temporary measure for educational discussion purposes and will be removed as soon as Newsweek indexes it electronically for retrieval via the Internet. They have announced their intent to do so sometime in 1999, after which the same document will be available through their search engine.


Merchants of Mediocrity

The College Board nationalizes grade inflation

by Robert J. Samuelson

WE HAVE PLENTY OF RANDOM MEDIOCRITY IN AMERICA; we don't need to promote it. Now the College Board proposes to do just that. It recently an-nounced (in case you missed the news) that it would raise the average scores on its two basic SATs - Scholastic Assessment Tests - to 500, the midpoint on the 200-to-800 scale. In 1995, the average verbal SAT would rise from 424 to 500, while the average math SAT would increase from 478 to 500. Almost everyone gets a boost. This is nationalized grade inflation: the solution to low test scores is to jack them up.

Naturally, the College Board heatedly denies this. It says that the SATs aren't being weakened, only "recentered." Nor would students' percentile rankings - their place among all test takers - be affected, and these are also reported to colleges. So why change at all? "Clarity," says College Board president Donald M. Stewart. Students will quickly know where the average is (500) and can easily measure how well they did. Today's low averages create needless confusion, he argues.

Gulp. Why should anyone be confused? Students get personalized reports providing their test scores and percentiles. These are neatly printed in a big box that shows, for example, that a 470 verbal score is in the 65th percentile. For someone mystified by percentiles, the report explains: "The national percentile for your verbal score of 470 is 65, indicating you did better than 65% of the national group of college-bound seniors." Students who don't understand this aren't ready for college.

Up to a point. the "recentering" is mere bureaucratic make-work. It probably won't much affect college admissions. Students vying for spots at highly selective schools (the Yales and Stanfords) will still have to compete to get them. By contrast, many other colleges are eager to enroll almost anyone with a high-school diploma. Without students, colleges go out of business, admissions standards are lax.

But in a larger sense, the "recentering" hurts us all by sanctioning mediocrity. Like it or not, the SATs have become a barometer for, say, the top 30 to 40 percent of students. The "recentering" will obscure long-term trends and make them harder to explain. The College Board promises tables to convert the new test scores to the old. That is: to show that a 500 on the 1995 verbal test equals a 424 on the 1993 test or that a 670 used to he only a 600. Does Stewart really think he's advancing "clarity" by creating two sets of numbers that mean the same thing?

The practical effect will be to hide the lower scores and to reassure everyone. Worse, the "recentering" blurs the distinction between students' abilities in math and reading (and, by inference, writing). The trends on the two SATs contrast sharply. On the math test, the score dropped from a postwar high of 502 in 1963 to a low of 466 in 1980 and then rebounded to 478 in 1993. On the verbal test, there has been no rebound. In 1963, it was 478 (just below the postwar high of 479 in 1956). By 1980, the score was down to 424, where it's essentially stayed.

Schools ought to be asking why they have done better in math than in reading. One answer may be an increasingly "nonliterate culture," as historian Diane Ravitch of New York University puts it. But Ravitch - an assistant secretary of education in the Bush administration - also thinks that a toughening of high-school course requirements in the 1980s succeeded in math and not elsewhere. "In math, you can't fudge it," she says. "You can't turn algebra into 'fun with numbers.' When you get into languages and social sciences, you really don't know what's under the [course] labels."

By raising both the verbal and math averages to 500, the College Board would obscure the crucial fact that skills have deteriorated much more in reading than in math for college-bound students. It doesn't mean much to compare yourself to the average if the average student doesn't read or write well. But the College Board is preaching otherwise. On today's tests, Stewart writes, "identical verbal and math scores do not indicate equal levels of skills." This is simply wrong. Compared with past performance, similar scores do indicate similar skills. It's only against today's averages that identical scores rank differently.

Schools can't address problems that have been defined out of existence. If we don't expect much from students, we won't get much. If we bless low scores by inflating them, we undermine students' motives to work harder to do better. "Our kids don't read enough, and we're sending a message that things are fine," as Ravitch says.

Less elite: The marvel is that the College Board - composed of colleges and secondary schools - has justified this travesty. One argument is that the SATs were first "centered" at the 500 midpoint in 1941, when only 10,000 students took the tests. Since then, American college has become a lot less upper-crust. Students students span the economic spectrum; there are more women, blacks, Hispanics and other minorities. Bigger pools of students (the argument goes) depress average scores. The SATs, therefore, need to be "recentered" to reflect this "democratization." Sounds convincing. It isn't.

Much of the "democratization" preceded the drop in test scores. Between 1951 and 1963, the number of test takers went from 81,000 to nearly 1 million; test scores rose slightly. The fact that more blacks (who have lower average scores) and more women (who have lower math scores) now take the exam does explain a small part of the score decline. But alter a thorough analysis, social scientists Charles Murray and R. J. Herrnstein reported in The Public Interest magazine that the main declines occurred among whites and could not be explained by changes in students' gender, economic class or parental education.

The decline occurred, they argued, because the college track in high school was "mediocritized." Grades were inflated. Courses became less demanding. Less homework was required. Fewer term papers were assigned. The effects cannot be washed away by raising SAT scores. But they can be disguised. High schools do not want to be seen producing poor students; colleges do not want to be seen misspending on the ill-prepared. This is the real reason - conscious or otherwise - for "recentering." It is to make "educators" look better.


View index of teaching-related articles