Last month, tens of thousands of 12-year-olds sat next to high school juniors and seniors in classrooms and gymnasiums across the country and took the SAT. With disturbing frequency, the SAT I: Reasoning Test is being used to select 7th grade students for university-based academic-enrichment programs across the country. Designed for college-admission purposes, the SAT is being misused as the determining criterion for young gifted students to gain access to many campus-based summer courses, international study trips, college-credit courses, mentors or master teachers, scientific field studies, supplementary learning materials, advanced-placement courses, career-planning conferences, and even high school credit courses.
Duke University’s Talent Identification Program, or TIP, operates a project targeting academically gifted adolescents in 16 states: Alabama, Arkansas, Florida, Georgia, Iowa, Kansas, Kentucky, Louisiana, Mississippi, Missouri, Nebraska, North Carolina, Oklahoma, South Carolina, Tennessee, and Texas. Its stated target group is “academically talented youth...our nation’s most valuable intellectual resource.” With the help of teachers and counselors in the schools, TIP last year targeted 70,000 academically gifted 7th graders who scored at or above the 97th percentile on school-administered standardized tests. For access to the TIP enrichment programs, these 7th graders are “invited” to take the regular SAT (or ACT) and to post an SAT math score of more than 570 and an SAT verbal score of more than 570 (or a combination of more than 1040).
Northwestern University’s Center for Talent Development runs three similar enrichment summer programs for academically gifted youths. These programs give students the opportunity to study with master teachers and complete entire yearlong high school courses during a three-week camp. In Northwestern’s Spectrum program, 7th graders seeking credits in algebra, geometry, and trigonometry must obtain an SAT math score of more than 550 or a combined math and verbal score of more than 920 (with an SAT math score of more than 520).
In addition, Stanford University’s Education Program for Gifted Youth allows access to participation in its math enrichment programs based solely on SAT performance. Only students scoring in the top 3 percent of the population on the math portion of the SAT or the Preliminary SAT are eligible.
Johns Hopkins University’s Institute for the Academic Advancement of Youth only admits 7th graders with SAT math scores above 530 and SAT verbal scores above 510. For its Center for Talented Youth’s center for academic achievement, a summer enrichment program, SAT math and verbal scores each above 430 are required. Higher cut-off scores are required for other programs of the institute.
At Johns Hopkins and Duke universities, like the other university-based enrichment programs, the sole criterion for being identified as “talented” and eligible for participation is the single standardized-test score. Such a talent-identification practice raises validity and equity issues that are troublesome for educators seeking viable and defensible enrichment programming for children with academic interests and gifts.
An important validity issue is that the SAT was designed to predict first-year college grades, not to identify academically gifted subteens. The enrichment programs’ materials, however, justify the sole use of the SAT score for selection. They report that the SAT is “intended to predict academic success” or “predicts scholastic achievement for very gifted 7th graders,” or that the score “provides separate information about a number of academic skills.” But the particulars of the touted outcomes predicted by the SAT score given at such a young age are not identified by the programs. The only outcome that an SAT score of a 12-year-old can likely predict is that child’s score on future tests similar to the SAT.
The key equity issue raised by this “talent identification” practice, according to mounting research evidence, is that the SAT significantly underpredicts the actual academic performance of girls. One consequence of the sole use of the SAT is that more boys gain access to these enrichment programs than girls. Depending on the cut-off scores, which differ for enrichment programs, the boy-girl ratios for program eligibility also vary. In the 1980s, there were as many as 13 boys for every girl who qualified. By 1993, some programs using the lower cut-off score of 500 per subtest were finding three boys for every girl qualified. But the Duke, Johns Hopkins, and Northwestern programs use math cut-off scores, for example, from 530 to 570, to qualify for most enrichment programs. Despite some positive trends in girls’ scores, still many fewer girls than boys meet the cut-off-score requirement.
The resulting underrepresentation of girls in final-eligibility ranks for these programs is what causes test-reform advocates to raise the specter of possible legal challenge of the practice.
| When students were given the SAT under an untimed condition, girls performed significantly better than under timed conditions. | 
|
One such legal action was begun in 1995 by the National Center for Fair & Open Testing. It has precipitated a study by the U.S. Department of Education’s office for civil rights of the National Merit Scholarship Corp., which dispenses $25 million worth of prestigious scholarships annually. FairTest’s suit claimed that of the 15,600 National Merit semifinalists in l996, only 39.5 percent of those finally eligible for the scholarship were female, even though 56 percent of the initial scholarship competitors were female. Because scores on the PSAT were the governing criterion for scholarship eligibility, FairTest successfully argued that the use of a single test score which advantaged boys resulted in less access to the scholarship money for girls, and was therefore biased. As a result of this legal challenge, the National Merit Scholarship Corp. and the Educational Testing Service have agreed to some changes in PSAT testing practices that will begin to help equalize the male-female score differential.
Some university-based enrichment programs are acknowledging problems with comparing boys and girls using the SAT score, even if their eligibility practices do not yet offer solutions. One researcher with Duke’s TIP program, observing a 40-point gender differential on SAT math scores, had 7th grade TIP participants retake the SAT under experimental conditions. When students were given the SAT under an untimed condition, girls performed significantly better than under timed conditions. More important, however, girls scored as well or better than boys under the untimed condition. The researchers reasoned that girls may have as much knowledge about mathematics as boys, but need more time to answer math questions.
Other organizations weighing in on the “testing and talent identification” debate include the American Association of University Women. The AAUW’s report “How Schools Shortchange Girls” warned that these “talent searches,” using the SAT math score, promote a cycle of under-education of girls in these and related fields and under-identify mathematical or scientific talent in girls. Moreover, the AAUW report alerted parents to the possible deleterious effects of 7th graders’ reading in some enrichment-program materials that boys are expected to outperform girls on the SAT.
The Center for Women Policy Studies and researcher Phyllis Rosser have examined the SAT gender gap and long argued that the perceived underachievement of girls, specifically in math, could be reduced or eliminated by changing testing and talent-identification practices. Ms. Rosser surveyed enrichment programs for gifted adolescents at Johns Hopkins, Duke, the University of Denver, the University of California at Sacramento, the University of California at Berkeley, and the ROGATE New Jersey Talent Search program in 1990. She found that only the New Jersey program, which used multiple measures for the identification of talent (such as additional test scores besides the SAT or PSAT), had comparable numbers of eligible girls and boys.
The practice of using one SAT score to determine which 12-year-olds qualify for some of the nation’s most advantageous scholarly riches should be eliminated. Even the College Board and the Educational Testing Service, the makers of the SAT, warn against sole reliance on a single test score for making high-stakes educational decisions. These university-based academic-enrichment programs, by using SAT scores that underpredict girls’ achievement, are reinforcing inequities of educational opportunity and a gender gap in math and science. True talent searches should alternatively look at a range of factors such as grades, course rigor, academic interests, multiple test scores, and teacher nominations. By intensifying the search for academically capable adolescents, we create a more valid and defensible definition of “talent.” And by broadening the definition of talent, we communicate to American boys and girls that adults in charge are wise and fair-minded.
