by Dr. Mark H. Shapiro
Commentary of the Day - March 11, 2001: Un-SAT-isfactory:
The birdseed really hit the fan when University of California President Richard Atkinson suggested dropping the SAT-I test as a condition for admission to U.C. campuses. To hear it from some commentators (Virginia Postrel is a good example), this was tantamount to the end of western civilization and surely would cause Berkeley and UCLA to tumble from the ranks of the nation's elite public universities.
The SAT-I is regarded by many to be a test of "aptitude" for college-level work. The test is supposed to measure a student's verbal and quantitative reasoning skills. In fact, at one time the SAT was called the "Scholastic Aptitude Test", although that name now is downplayed by the College Board that sponsors the exam. Many students also take the SAT-II tests, which are subject-matter-based achievement tests. A careful reading of Atkinson's proposal shows that he wants the University of California to place more emphasis in the admissions process on the SAT-II subject-matter tests as well as other measures of actual student achievement.
The Irascible Professor thinks that this is an idea whose time has come, and agrees with Atkinson for the following reasons. First, SAT-I does not measure "aptitude" for college in any meaningful sense. It is simplistic in the extreme to assume that the qualities needed for success in college can be reduced to two numbers -- a SAT verbal score and a SAT quantitative score. One need only read Gould's treatise on the misuses of "intelligence" testing (The Mismeasure of Man, Stephen Jay Gould, W.W. Norton & Co., 1996) to understand that it is impossible to reduce complex human attributes to a single number or small set of numbers. The correlation between SAT-I scores and measured success rates in the first year of college is low. By some estimates there is less than a 15% correlation between grades in the freshman year and SAT-I scores.
Second, SAT-I scores are not "objective" in the sense that one cannot say with any certainty that if student A's total SAT-I scores are 150 points higher than student B's SAT-I scores then student A is better qualified for college than student B. The reason for this is that a student can improve his or scores on SAT-I by taking "prep" courses and practice exams. In fact, many of the major "SAT-prep" services offer a money-back guarantee stating that students who take their services can improve their scores by at least 150 points. This means that more affluent students who can afford to take "SAT-prep" courses are likely to achieve higher scores than the less affluent students who can't afford to the $1,500 needed for the typical "SAT-prep" course. On a test like the SAT-I, where scores are more or less normally distributed, a 150-point increase in raw score will push the "prepped" student into a much higher percentile rank than his or her "non-prepped" counterpart who has essentially the same "aptitude" for college work. Thus, excessive reliance on SAT-I works to the detriment of students from the less affluent segments of society.
Third, far too much time is spent by high school students preparing for SAT-I. Because very little of what is covered on SAT-I is directly relevant to college work, this is time wasted. On the other hand, the SAT-II exams - while far from perfect - encourage students to learn basic subjects in high school that will serve as a foundation for college work. In addition, while the locus of SAT-I preparation generally is outside the school classroom, the locus of SAT-II preparation is in the classroom. This shifts the emphasis of college preparation work back to the high schools, where it rightfully belongs.
Fourth, the SAT-II exams allow a student to present a portfolio of measurements that have been taken over a period of time rather than a snapshot that was taken in a single three-hour sitting. Thus, college admissions officers can obtain a better profile of the student's overall strengths and weaknesses. In addition, the SAT-II exams encompass essentially all of the "reasoning" skills contained in SAT-I, but do so in a context that also values the development of study skills and actual learning.
Atkinson called for a more "holistic" approach to college admission. One should not dismiss this as just a subterfuge for "affirmative action", as some have claimed. There are many attributes that are important for success in college (and in life) that neither the SAT-I nor SAT-II exams measure to any great extent. These include -- among others -- ambition, drive, persistence, originality, artistic talent, and leadership qualities.
This is not to imply that the SAT-I and SAT-II tests are without value in college admissions. It's just that they tell only a small part of the story. The SAT-I probably is a better indicator of failure than of success. In other words, a student with very low SAT-I scores probably is not going to do well in a highly competitive college or university. However even this is not a sure-fire indicator. For example, in a recent series of articles on the SAT Time Magazine points out that former U.S. Senator and Rhodes Scholar Bill Bradley scored only 485 points on the SAT-I verbal test, and well-known author Amy Tan's total SAT-I scores were in the unimpressive 1100-range. However, one could make a good argument that both Bradley and Tan are every bit as creative and intelligent as talk-show host Ben Stein who scored a near-perfect 1573; and, that college was not wasted on either Bradley or Tan.
Finally, the IP would be remiss if he did not remind his readers of the work of Clifford Adelman, which showed that the single best predictor of success in college is not SAT scores but rather the quality and rigor of a student's high school experience as measured by the number of advanced placement (AP) and International Baccalaureate classes taken by the student.
© 2001 Dr. Mark H. Shapiro - All rights reserved.