No. 65
February 27, 2001
Lance Izumi

In his swan song as president of the University of California (UC) system, Richard Atkinson has proposed eliminating the use of SAT I test scores as part of the UC’s admissions process. Atkinson claims that the SAT I is unfair, damages students’ self-esteem, does not measure real student knowledge, and prevents the UC from having a racially-diverse student body reflective of California’s population. Many believe that it is this last reason, which has been a continuing battle cry of liberal educators and minority activists, that is the key motivation behind Atkinson’s radical proposal. Regardless, evidence contradicts Atkinson’s various claims.

The SAT Predicts Students’ Success in College

There are two SAT exams, the SAT I, which tests a variety of knowledge and skill areas, and the SAT II, which is subject-specific, and both are currently required by the UC. Atkinson would like UC to stop using both exams, but would like to eliminate the SAT I first. Because the students who take the SAT I and II change from year to year, thus making the quality of the annual student test-taking population variable, the exams are not a good indicator of general student performance. However, as a predictor of a student’s future success in college, research shows that standardized tests such as the SAT I are valuable.

Atkinson claims that students’ high-school grade-point averages (GPAs) are more accurate than SAT I scores in predicting students’ college grades. National evidence, however, contradicts this claim.

For example, the College Board, which oversees the SAT, gathered data on college grades in individual courses from 45 institutions of higher learning. According to the Board’s 1997 study of these data, “A significant finding of this analysis of data by course category is that the correlation is higher for the SAT than for high school record alone in most areas.”¹ In other words, the SAT I predicted grades in individual college courses better than high-school GPAs. In addition, the study found, “As with earlier validity studies, the combination of SAT scores and high-school record yields the best set of predictors.”² This is why most colleges and universities, including the UC, have used a combination of test scores and GPAs in their admission process. Indeed, regarding the SAT’s effectiveness in predicting college grades, the study notes: “Literally thousands of validity studies have been conducted for institutions for more than a half-century. The evidence is clear: The SAT works and it works very well in many different circumstances.”³

Further, although Atkinson implies that the SAT I discriminates against African-American and Hispanic students, the SAT is an especially good predictor of the college grades of ethnic groups. A data review of a 1993 College Board study of 46,379 students at 55 colleges and universities across the country found that “for most ethnic groups the SAT alone is a better predictor of course grades than are high school grades alone.”⁴ The review notes that the best predictor of students’ college performance is a combination of SAT scores and high-school GPAs, which, again, is exactly what is used by most colleges and universities in their admissions process.⁵

Further, the review points out that the SAT actually benefits most minorities: “For [African-Americans, Hispanics, and American Indians],… the SAT tends to predict a slightly higher GPA than the students actually earn.”⁶ It is, therefore, highly unlikely that minority students’ chances of admission are unfairly hurt by the SAT I.⁷ The data review concluded:

Students admitted to colleges or universities where the curriculum is beyond their level of academic preparation are unlikely to be academically successful, and placing students in such a situation does them a disservice. The best use of information from the SAT, and other academic performance measures, is to judge the likelihood that a student has the academic preparation necessary to succeed at a particular institution.… The SAT, in conjunction with high school grades, is the best means available to identify students who are likely to be academically successful.… However, eliminating the best tool available for judging the ability to succeed will not ensure equal access to all who are capable of succeeding, and runs the risk of increasing an already troublesome situation in which only 41 percent of Hispanics who enter college seeking a four-year degree end up with any degree at all.⁸

If SAT I scores are a good predictor of college grades, it should come as no surprise that tests like it are also better predictors of whether a student eventually completes a bachelor’s degree. A recent U.S. Department of Education study found that the correlation between SAT-like test scores and bachelor’s degree attainment was higher than the correlation between students’ high-school class rank/GPA and bachelor’s degree completion.⁹

The SAT and Minority Enrollment

It should be pointed out that, despite Atkinson’s claim that the SAT I is an obstacle to increasing underrepresented minority enrollment in the UC, getting rid of the SAT would not necessarily guarantee increased minority enrollment in California’s top universities. This is especially the case with regard to the University of California (UC) system.

The Master Plan for Higher Education in California recommends that the UC system set its freshman eligibility criteria so that the top 12.5 percent of the state’s public high-school graduates are eligible for admission. A 1997 UC study found that if SAT scores were eliminated as an admissions criteria (the UC currently uses both test scores and high-school grades in its admissions process), then the current GPA admissions standard of 3.3 in core high-school courses would have to be raised.¹⁰ Keeping the GPA standard at the 3.3 level would result in too many students becoming eligible, thus exceeding the Master Plan’s 12.5 percent limit.

Therefore, in the absence of SAT scores, the UC estimated that, in order to conform to the 12.5 percent limit, the minimum GPA for admissions would have to be raised to 3.65.¹¹ Under this scenario, white students, not underrepresented minority students, are the major beneficiaries. According to the study, “eligibility rates of white graduates increased by 17 percent (from 12.7 to 14.8) and the Latino eligibility rate increased by 5 percent (from 3.8 to 4.0).”¹² However: “In contrast, proportionately fewer African-American graduates within the fully and potentially eligible groups had achieved a 3.65 or greater GPA. The African-American eligibility rate falls from 2.8 percent to 2.3 percent, representing an 18 percent decline.”¹³

Further, the study found that eliminating the UC’s ability to use a combination of SAT score and GPA (a combination which UC terms its “index”) would have the greatest negative impact on African-American and Latino students. The study concluded that: “Had the index criteria not been a component of the eligibility policy, the statewide eligibility rate would have been lower by eight percent, yielding an overall eligibility rate of approximately 10 percent. The greatest impact of the index criteria is on African-American and Latino rates—index eligibility contributes 15 percent and 12 percent to these respective subgroup rates.”¹⁴

Judy Kowarsky, a top UC undergraduate admissions official, says that dropping the SAT does not level the playing field, partly because anything that supposedly helps minority students get into the UC also helps students of other races. According to Kowarsky, “When you adjust eligibility to accommodate more students from low-eligibility backgrounds, then the eligibility for other groups also adjust upwards.”¹⁵

The SAT and Racial Politics

Although Atkinson says that the UC must set high standards, he also says that since California has a highly diverse racial and ethnic population, the UC “must be careful to make sure that its standards do not unfairly discriminate against any students.”¹⁶ According to Atkinson’s logic, because he believes the SAT keeps African-American and Hispanic students out of the UC, the test thus discriminates against these groups, so therefore it must be eliminated. Of course, he does not say why poor Asian-American students, many of whom come from the same or similar neighborhoods and schools as African-American and Hispanic students, do just fine on the SAT. Addressing such a point would not be politically or racially correct.

Atkinson’s statements make it clear that his opposition to the SAT is motivated more by racial politics than by any genuine interest in the empirical evidence regarding student achievement. Indeed, David Murray, research director of the Washington, DC-based Statistical Assessment Service, states that “every major premise on which [opposition to the SAT] rests is false.”¹⁷ Murray observes that SAT scores correlate “with those on a whole range of other measures and assessments, including IQ tests, the National Assessment of Educational Progress, and the National Educational Longitudinal Study.”1¹⁸

Neither is the SAT somehow uniquely unfair to minorities. Murray points out that “Far from being idiosyncratic, the scoring patterns of whites, blacks, Hispanics, and Asians on the SAT and GRE are replicated on other tests as well.”¹⁹ Critics would retort that all standardized tests, not just the SAT, are biased. Murray refutes this charge by noting that “the National Academy of Sciences concluded in the 1980s that the most commonly used standardized tests display no evidence whatsoever of cultural bias.”²⁰ Murray notes that a University of California study showed that the SAT actually overestimates the first-year grades of African-Americans and Hispanics in the UC system.²¹

Murray also says that while the SAT may slightly underpredict the college success of women, for more selective universities, “the SAT predicts the grades of both sexes quite accurately.”²²

Finally, Murray points out that research shows that SAT scores do not depend heavily on the income of students’ families. Students of different races but whose families have similar incomes perform very differently on the exam.²³

Murray concludes that those who are trying to eliminate the SAT hope “to achieve the ends of affirmative action by other, more politic means.”²⁴ SAT critics, says Murray, want “more ‘nuanced’ measures of scholastic merit like ‘creativity’ and ‘leadership,’ tacitly understood as stand-ins for skin color.”²⁵ Atkinson, for instance, advocates a so-called “holistic” assessment of student merit. As Murray points out, however, “there is no reason to think that minority students possess these qualities in greater abundance than do their peers.”²⁶ Murray concludes, “The attempt to substitute them for test scores will thus only perpetuate the corrupt logic of affirmative action by piling deception upon deception.”²⁷

Grade Point Average

By eliminating SAT I scores from the admissions process, Atkinson’s proposal will most likely mean greater emphasis on students’ high-school GPAs. Giving increased weight to GPAs, however, opens up a can of worms for admissions committees. First, there is the issue of grade inflation and grade manipulation. In 1983, the overall average GPA of high-school graduates in California was 2.62. By 1996, the overall state high-school-graduate GPA had climbed to 2.78.²⁸

This GPA increase affected students of all ethnic groups. From 1983 to 1996, average African-American high-school-graduate GPAs increased from 2.26 to 2.41. Latino-graduate GPAs increased from 2.42 to 2.55. The GPAs of whites increased from 2.69 to 2.90, and Asian-graduate GPAs increased from 2.96 to 3.19.²⁹

According to the California Postsecondary Education Commission, “Increased participation in honors and AP classes probably account for some of the increase.”³⁰ Advanced placement (AP) courses are college-level high-school classes that allow students to earn college credits. The reason why increased participation in these classes increases GPAs lies in the fact that students can earn grades with a higher point score than the normal 4.0 (up to 5.0). It is thought that the opportunity to earn higher than a 4.0 grade gives students the incentive to take the more difficult AP classes.

The other reason why GPAs have increased is grade inflation. While some students really earn their grades by working hard and taking challenging courses, others receive grades higher than their efforts warrant or receive high grades in less difficult courses. As a University of California report notes, GPAs are subject to “biases resulting from variations in school curriculum and grading practices.”³¹ A 1997 College Board report notes that, nationally, the first-year college GPAs of students are much lower than their high-school GPAs. The report concluded that “high school grades depend as much on the difficulty of the curriculum to which students are exposed, and on the grading standards of the individual high school, as they do on what students actually learn and how prepared they are for college-level classes.”³²

One stunning example of grade inflation and manipulation occurred at Balboa High School in San Francisco in 1998. In an official memorandum, administrators at the high school told teachers to increase the combined total of all As, Bs, and Cs at the school by five percent over the previous year’s total.³³ In other words, administrators ordered teachers to stop giving out so many low or failing grades. Given such grade manipulation, it is no wonder that college admissions counselors rate SAT scores a more reliable measure than GPAs.³⁴

Conclusion

Richard Atkinson’s proposed elimination of the SAT I as an admissions requirement for the UC is foolish and wrongheaded. In order to create his utopian racially-balanced student body, Atkinson claims that admissions should be based on an evaluation of the whole student that includes greater emphasis on a student’s life experiences. The trouble with such an admissions strategy is that it is ultimately based significantly on subjective factors and will result in arbitrary decisions—decisions which will hurt students. As David Murray observes:

But gaining a fuller picture of a particular student’s promise is a difficult business, especially in an admissions process that very often involves sorting through thousands of individuals. Moreover, it can only go so far before it ceases to have anything to do with education. What a student is like outside the classroom is surely significant, but until we are prepared to say outright that the heart of the matter is something other than fitness for academic work, a crucial gauge of whether a student is going to be able to pass a biology final or write a political-science research paper will remain that old, much-maligned SAT score.³⁵

Atkinson should, therefore, listen to the UC’s own previous recommendations. After release of the 1997 UC report that showed that elimination of the SAT I would actually hurt minority admissions, the faculty committee that helps set UC admission requirements publicly stated its support for keeping the SAT I.³⁶

In sum, the SAT I is a useful predictor of student performance in college. Further, given the comparison problems inherent in high-school grading, the SAT I gives colleges a single yardstick by which to measure students’ potential. Dropping the SAT I, therefore, would open a Pandora’s box of evils that would end up harming California education. The prudent course, then, would be for the UC Board of Regents to ignore Richard Atkinson’s proposal and retain use of the SAT I.

Notes

1 Gretchen W. Rigol, “Common Sense About SAT Score Differences and Test Validity,” The College Board, Research Notes, June 1997: 5.

2 Ibid.

3 Ibid., 4.

4 Greg Perfetto, “Understanding the SAT scores of Hispanic students in the context of educational opportunities, performance and outcomes,” The College Board, September 23, 1997: 3.

5 Ibid., 2.

6 Ibid., 3.

7 Ibid.

8 Ibid., 4.

9 Clifford Adelman, “Answers in a Tool Box: Academic Intensity, Attendance Patterns, and Bachelor’s Degree Attainment,” U.S. Department of Education, June 1999.

10 University of California, Office of the President, Student Academic Services, “University of California Follow-up Analyses of the 1996 CPEC Eligibility Study,” December 1997: 19.

11 Ibid.

12 Ibid., 20.

13 Ibid.

14 Ibid., 6.

15 ”Move to Drop Entrance Tests Could Hurt Minorities, Report Says,” San Jose Mercury News, January 30, 1998. Available at www.ccsf.cc.ca.us/Guardsman/s980130/cps02.htm.

16 Richard Atkinson, “Standardized Tests and Access to American Universities,” 2001 Robert H. Atwell Distinguished Lecture, delivered at the 83rd Annual Meeting of the American Council on Education, Washington, D.C., February 18, 2001. Available at www.ucop.edu/ucophome/pres/comments/satspch.html.

17 David W. Murray, “The War Against Testing,” Commentary, September 1998: 36.

18 Ibid.

19 Ibid.

20 Ibid.

21 Ibid.

22 Ibid.

23 Ibid.

24 Ibid., 37.

25 Ibid.

26 Ibid.

27 Ibid.

28 California Postsecondary Education Commission, “Performance Indicators of Higher Education in California 1998,” December 1998: 45.

29 Ibid.

30 Ibid.

31 University of California, Office of the President, Student Academic Services, op. cit., 17.

32 Perfetto, op. cit., 2.

33 ”Teachers Told to Pump Up Grades,” San Francisco Chronicle, April 4, 1998.

34 Richard P. Phelps, “Why Testing Experts Hate Testing,” Fordham Foundation, January 1999: 9, citing National Association for College Admission Counseling, “Members Assess 1996 Recruitment Cycle,” 2, 4.

35 David W. Murray, op. cit., 37.

36 ”Move to Drop Entrance Tests Could Hurt Minorities, Report Says,” op. cit.

* Lance Izumi is the Director of the Center for School Reform at the California-based Pacific Research Institute for Public Policy. This article is based on research published in his study, California Index of Leading Education Indicators 2000.