<p>
[quote]
...Christopher Berry of Wayne State University and Paul Sackett of the University of Minnesota...pulled 5.1 million grades, from 167,000 students, spread out over 41 colleges. They also got the students SAT scores from the College Board, as well as the list of schools each student asked the College Board to send their SAT scores to, an indicator of which colleges they applied to. By isolating the overlapswhere students had applied to the same colleges, and taken the same courses at the same time with the same instructorthey extracted a genuine apples-to-apples subset of data.</p>
<p>It turns out that an SAT score is a far better predictor than everyone has said. ...
<p>This type of apples-to-apples comparison that attempts to minimize the effect of self-selection is important in trying to consider how the SAT is a predictor of college GPA. When you consider that SAT 1400-scorers and their 1,100 counterparts tend to go to different colleges, it would seem that using high school GPAs as predictors of college GPAs becomes less meaningful.</p>
<p>If you play a mind game of imagining that Harvard were to choose its freshman class randomly from all high school graduates across the country, rather than by merit, students there with higher SATs would, as a rule, outrank academically those with lower scores. With the widespread grading inconsistencies and grade inflation among secondary schools, I doubt that high school GPAs would predict college grades for this mythical Harvard freshman class as well as their SAT scores would.</p>