A 1979 study conducted by researchers at the University of Virginia looked at the SAT scores and high school grades of first-year chemistry students. Researchers found that SAT math scores were a better indicator of the students' success than their high school GPAs. A 2001 Educational Testing Service (ETS) report noted that SAT II and GRE Subject Tests in math and science disciplines were valid predictors of future performance within those same academic areas, but that the predictive validity received a boost when considered in combination with the overall SAT and GRE General Test, respectively. A 2010 study conducted by researchers at Capella University concluded a moderate to high relationship between science and math classroom grades and standardized test scores on the Virginia State Standards of Learning (SOL), a standardized test required of third, fifth and eighth grades in Virginia public schools.
The 2001 study by ETS found that SAT II and GRE Subject Tests had less predictive validity when it came to the correlation between students' scores and classroom grades in history, foreign language and literary subjects. A 2010 study by Capella University that analyzed SOL test scores reported the same observation.
Findings relating to standardized tests in general, such as the SAT, suggest that the correlation between grades and standardized tests varies according to the demographic in question. Studies, including one conducted in 2000 by the College Entrance Examination Board, have reflected a strong correlation between standardized test scores, high school GPAs and college freshman grades for white students. However, for some groups of minority students, such as African Americans, and non-traditional students, including those returning to school after age 30, standardized test scores have not shown a correlation with high school grades or college freshman grades.
Researchers have consistently demonstrated that predictions of future academic performance become more valid when using both high school grades and standardized test scores to make the assessment, rather than one or the other. One 1992 study conducted by researchers at the University of Pennsylvania found that the SAT was the weakest predictor of college freshman year academic performance, that SAT II scores were only slightly better predictors and that class rank was the most useful predictor. However, combining the three factors yielded the greatest predictor of freshman year performance. A 2007 study by researchers at the University of California Berkeley found high school GPA to be the strongest predictor of academic performance in college, across all academic disciplines. It also found that the correlation was less negatively impacted by variations in demographic factors than were standardized test scores.