SAT

• 1. Introduction

  ›

  The SAT has been shaping the futures of American students since its debut in 1926. A single score on a scale from 400 to 1600 can determine which college a teenager attends, which scholarship they receive, and, according to some researchers, how their life unfolds decades later. But the test that millions of students dread or celebrate is far stranger and more contested than any of them might guess. Its name has changed at least four times. Its scoring system has been overhauled. Its paper booklets were retired in favor of software run on laptops and tablets. And the question of what it actually measures has generated a century of argument among psychometricians, sociologists, educators, and politicians. What does a number between 400 and 1600 really tell us about a person? And who benefits most from a system that ties so much to a single morning's performance?

• 2. A Century of Name Changes

  ›

  For much of its life the test went by the name Scholastic Aptitude Test, a title that carried a particular promise: that it was measuring innate mental potential, not accumulated knowledge. That promise proved controversial enough that the College Board eventually dropped the word "Aptitude" entirely, renaming the exam the Scholastic Assessment Test. The name changes did not stop there. At various points the test became the SAT I: Reasoning Test, then the SAT Reasoning Test, and finally just the SAT. Each renaming reflected a shift in how its creators wanted the public to understand what they were selling.

  The College Board, which wholly owns, develops, and publishes the SAT, brought in David Coleman as president to oversee a significant overhaul introduced in 2016. Coleman said he wanted the test to reflect more closely what students actually learn in high school, aligning it with Common Core standards. That was a direct reversal of the original design philosophy, which had intentionally kept the test separate from high school curricula. The optional essay section, introduced to measure writing ability, was quietly discontinued after June 2021. The College Board cited other ways students could demonstrate writing skill, and acknowledged that the COVID-19 pandemic had accelerated what it called "a process already underway."

  The Subject Tests that had traveled alongside the SAT since around 1937, known at various times as SAT Achievement Tests and then SAT II: Subject Tests, were also dropped after June 2021. In January 2022, the College Board announced the move from paper to digital. International testing centers switched to the digital format on the 11th of March 2023, and the last paper-based SAT in the United States was administered in December 2023. The digital switch brought a significant structural change: the test became adaptive, with the difficulty of the second module in each section determined by the student's performance in the first.

• 3. Inside the Test Itself

  ›

  The current digitally administered SAT contains two main sections: Reading and Writing, and Mathematics. Each section is divided into two equal-length modules. The Reading and Writing modules run 32 minutes each and contain 27 questions apiece, built around short passages of 25 to 150 words, each followed by a single multiple-choice question. The Mathematics modules run 35 minutes each with 22 questions, covering algebra, advanced high school math, problem solving and data analysis, and geometry and trigonometry. Algebra accounts for 13 to 15 questions; geometry and trigonometry just 5 to 7.

  Roughly 75% of math questions are multiple-choice with four options. The remaining 25% are student-produced responses, requiring test-takers to type in a numerical answer rather than select from choices. Importantly, there is no penalty for guessing; scores reflect only the number of questions answered correctly. Calculators are permitted throughout the mathematics section, and a Desmos-based graphing calculator is built directly into the testing software. A study of calculator use found that students who used calculators on about one-third to one-half of the items scored higher on average than those who used them more or less frequently. Researchers concluded this likely reflected how able students use tools differently, not that calculator use itself was the direct cause.

  Scores for each section are reported on a scale of 200 to 800, always as a multiple of ten. The two section scores are added together to produce a total between 400 and 1600. The average Reading and Writing score in 2025 was 521; the average Mathematics score was 508. The test as a whole takes 2 hours and 14 minutes, plus a 10-minute break between sections, and costs US$68 as of 2024, with fee waivers available to low-income students. Results typically arrive two to four weeks after the exam.

• 4. The Coaching Industry

  ›

  Stanley Kaplan pioneered systematic SAT preparation in 1946 with a 64-hour course, and the preparation industry has flourished ever since. Books, in-person classes, online courses, and one-on-one tutoring are offered by a wide range of companies. The College Board itself announced a partnership with Khan Academy in the 2015-16 academic year to provide free preparation materials, explicitly aiming to level the playing field for students from lower-income families.

  The effectiveness of paid preparation is far more modest than its marketing suggests. The College Board and the National Association of College Admission Counseling found that tutoring courses produce an average gain of about 20 points on the mathematics section and 10 points on the verbal section. A 2012 systematic literature review estimated coaching effects of 23 and 32 points for math and verbal respectively. A 2016 meta-analysis put the effect sizes at 0.09 and 0.16 for the two sections, with considerable variation across studies. An early meta-analysis from 1983 concluded that a gain of 10 points was "too small to be practically important."

  Researchers Ben Domingue and Derek Briggs found, using data from the Education Longitudinal Survey of 2002, that coaching's statistically significant effects were limited to mathematics and were greater among students already taking rigorous courses and those from higher socioeconomic backgrounds. In other words, those who needed the boost least tended to gain the most. Among all ethnic groups studied, a 2011 study found that the effects of one-on-one tutoring were minimal. Test-preparation scams remain a genuine problem, and the College Board has long maintained that the SAT is essentially uncoachable. Cognitive scientist Sian Beilock has suggested that the sharpest practical gains may come not from commercial courses but from practice questions, proctored exams to build procedural memory, writing about one's anxieties on the day of the exam, and consistent sleep in the days leading up to it.

• 5. What the Score Predicts

  ›

  In 2009, education researchers Richard Atkinson and Saul Geiser from the University of California system argued that high school grade point average is a stronger predictor of college grades than the SAT, regardless of the type of high school attended. The UC academic senate's 2020 report pushed back, finding that SAT scores were at least as good as high school GPA at predicting first-year retention and graduation, and better at predicting first-year GPA. Crucially, the report found this predictive power held across demographic groups, noting that scores were actually better predictors of success for underrepresented minority students, first-generation students, and students from low-income families.

  Geiser responded by calling the academic senate's findings "spurious," arguing that student demographics had been omitted in a way that distorted the analysis. Jesse Rothstein, a UC Berkeley professor of public policy and economics, countered that the academic senate had overestimated the value of the SAT and that admissions policies had not adequately "compensated" for group differences in test scores. By contrast, Brown, Yale, and Dartmouth, after analyzing their own institutional data, concluded that SAT scores were more reliable predictors of collegiate success than GPA and helped them identify more potentially qualified students from disadvantaged backgrounds.

  Psychometricians Thomas Coyle and David Pillow showed in 2008 that the SAT predicts college GPA even after statistically removing the general factor of intelligence, with which the test is highly correlated. A 2019 study with roughly a quarter of a million participants found that SAT scores and high school GPA together offered an excellent combined predictor of freshman GPA and second-year retention. Psychologist John Mayer called the SAT's predictive powers "an astonishing achievement" in a 2014 opinion piece, cautioning against making the test optional. Research by David Lubinski, Camilla Benbow, and their colleagues has further shown that SAT scores can predict life outcomes well beyond university, including income and occupational achievement.

• 6. Gender, Race, and Score Gaps

  ›

  A 2013 report from the American College Testing Board found that boys outperformed girls on the mathematics section of the SAT, a gap that had persisted for over 35 years. As of 2015, boys averaged 32 points more than girls on that section. Among students scoring in the 700-800 range, the male-to-female ratio was 1.6:1. Jonathan Wai and colleagues, analyzing three decades of data from over 1.6 million intellectually gifted seventh graders in the Duke University Talent Identification Program, found that in the 1980s the gender gap among students scoring in the top 0.01% was 13.5:1 in favor of boys, but had narrowed to 3.8:1 by the 1990s.

  Researchers who study this gap frequently point to greater male variability in cognitive abilities across cultures. Because males are overrepresented at both the highest and lowest ends of the performance distribution, more boys end up at the extremes of mathematical testing. In the late 2000s, for every girl who scored a perfect 800 on the mathematics section, there were two boys. On the verbal section, both sexes appeared broadly at parity among the top 5%, though girls developed a slight edge beginning in the mid-1980s.

  On race and ethnicity, a 2001 meta-analysis drawing on results from more than 6.2 million participants found a difference in average scores between black and white students of around 1.0 standard deviation, with comparable findings in the SAT's own data from 2.4 million test-takers. In the mathematics section, the black-white gap measured 0.91 standard deviations in 1996 and fell to 0.79 by 2020. In 2013, Asian Americans scored 0.38 standard deviations higher than white students in mathematics. In 2020, according to the College Board, 83% of Asian students met the college-readiness benchmark in reading and writing and 80% did so in mathematics; for black students those figures were 44% and 21% respectively. Sociologist Christopher Jencks has argued that this does not reflect a flaw in the tests themselves but rather that the tests measure skills that certain groups are systematically less likely to develop through their schooling and socialization.

• 7. Beyond College Admissions

  ›

  Certain high-IQ societies have long treated SAT scores from specific eras as qualifying entry criteria. Intertel accepts combined verbal and mathematics scores of at least 1300 on tests taken through January 1994. The Triple Nine Society accepts scores of 1450 or greater on tests taken before April 1995, and at least 1520 on tests taken between April 1995 and February 2005. Mensa accepts qualifying scores earned on or before the 31st of January 1994. Each society's cutoff reflects the era in which its norms were calibrated, since scoring scales changed over time.

  Researchers studying intellectual precociousness have used the mathematics section specifically to identify subjects for longitudinal studies. A 2005 study by Jonathan Wai, David Lubinski, and Camilla Benbow found that among the top 1% of intellectually precocious individuals, those with higher SAT mathematics scores at age 12 were more likely to earn a doctorate in STEM, publish research, hold a patent, or secure university tenure. Wai further showed that the average SAT or ACT scores of the college someone attended predicted individual differences in income, even among the wealthiest Americans, including Fortune 500 CEOs, billionaires, federal judges, and members of Congress.

  Frey and Detterman, in their 2004 study, found that SAT scores correlated with scores on the Armed Services Vocational Aptitude Battery at r=.82 in their sample, adjusted to .857 when corrected for non-linearity. Psychometrician Linda Gottfredson noted that the SAT is effective at identifying intellectually gifted college-bound students in particular. The study by Wai, Lubinski, and Benbow that administered the SAT to twelve-year-olds through programs like the Study of Mathematically Precocious Youth and the Duke University Talent Identification Program found that those in the top quartile of composite scores were markedly more likely, decades later, to hold doctoral degrees and to have at least one STEM publication.