What is the SAT Good For? Absolutely Nothing
Photo Credit: Laurence Gough | Shutterstock.com
Stay up to date with the latest headlines via email.
A certain Groundhog Day quality accompanies fall with the opening of schools and the inevitable release of SAT scores. While the Bill Murray comedy made light of the existential hell of a life lived like Sisyphus, the College Board's data dump each fall leaves little to provoke even a smile.
In 2000, David Grissmer confronted the publication and uses of SAT data:
"There is a unique issue in public policy associated with the publication and use of aggregate SAT test scores. The publication of aggregate score averages for schools, school districts, states, and the nation need not be done to carry out the primary purpose of the test: the sorting of individuals in the college admission process. However, annual SAT average scores have been routinely published at all levels since the 1960s. These scores receive widespread media coverage at every level because they provide easy and, in many cases, the only available comparative score data among local schools and school districts and among states. Because SAT scores are given at the end of K-12 education, they are often seen as providing indicators of the quality of the K-12 education system....More than a decade later, little has changed. Some erosion in the power of the SAT to determine college entrance has occurred (see growing list of colleges that allow admissions without SAT scores). But more and more students take the SAT, the test remains linked powerfully to scholarships and NCAA eligibility, and the media continues to fuel the College Board's data dump and the concurrent public and political conclusions drawn inaccurately from the numbers.
"Such use of SAT scores would be beneficial if the scores reflected accurately the quality of schools and school systems. Unfortunately, the aggregate SAT scores, at best, convey no useful information about educational quality and, at worst, convey highly misleading information about educational quality. Research beginning in the 1980s has pointed out the flaws in the aggregate SAT scores stemming from the ever changing, self-selected sample of students taking the tests. Yet aggregate SAT scores continue to be published and widely misinterpreted as indicators of educational quality." (pp. 223-224)
The SAT, What Is It Good For? Absolutely Nothing
The SAT, like war, isn't good for anything, even the single purpose of the test—to determine college freshman success.
Each year when SAT data are released, we should be reminded of some essential facts:
• The College Board itself cautions against using the SAT for any comparative purposes: "Educators, the media and others should...not rank or rate teachers, educational institutions, districts or states solely on the basis of aggregate scores derived from tests that are intended primarily as a measure of individual students." Average SAT scores for any state reflect the affluence of the test takers and the relative percentage of test takes—but certainly not the quality of the schools or the teachers.
• The College Board's own research repeatedly confirms that SAT scores are less predictive of freshman college success than GPA. (See Table 5, p. 5)
• SAT scores historically and currently are most strongly correlated with parental income and level of education for parents. Select any year from the archived data, and these facts are confirmed. In short, the SAT is a metric that confirms privilege more so than identifying academic achievement or academic readiness for college (except in which ways those are inextricably tied to privilege).
• In every year the SAT has been administered, males outperform females (except for the more recent writing section). But, looking carefully at the 2005 SAT data (the last year I can find the data separated by gender), there appears to be affluence and poverty correlations buried in the gender gap; note the chart below and how the percentages of female takers compared to male takers shift from high percentages of low-income females to high percentages of high-income males:
Less Than $10,000
$10,000 - $20,000
$20,000 - $30,000
$30,000 - $40,000
$40,000 - $50,000
$50,000 - $60,000
$60,000 - $70,000
$70,000 - $80,000
$80,000 - $100,000
More than $100,000
This leaves us with one unanswered but essential question: Why do we remain committed to a test that perpetuates the unfair privilege of affluent males?