In predicting college grades, both high school grades and SAT/ACT scores matter. Why shouldn't high school grades count?
There is no consensus on this among the researchers or administrators. Noncognitive variables play increasingly more important role in admission processes at many universities and in predicting college success.
Which researchers have found that high school grades and SAT scores are not predictive? Below is one of many studies confirming what I wrote. High school grades have a larger noncognitive component than SAT scores -- smart but lazy students may have good scores but mediocre grades.
http://professionals.collegeboard.com/data-reports-research/cb/validitySAT08-predicting_1st_yearValidity of the SAT for Predicting First-Year Grades: 2008 SAT Validity Sample
Krista Mattern; Brian Patterson
02/11/2011
The findings for the 2008 sample are largely consistent with the previous reports. SAT scores were found to be correlated with FYGPA (r = 0.54), with a magnitude similar to HSGPA (r = 0.56). The best set of predictors of FYGPA remains SAT scores and HSGPA (r = 0.63), as the addition of the SAT sections to the correlation of HSGPA alone with FYGPA leads to a substantial improvement in prediction (Δr = 0.07). This finding was consistent across all subgroups of the sample, by both institutional characteristics and demographics (Δr ≥ 0.06). All correlations presented here have been corrected for restriction of range, but the same basic patterns hold for the raw correlations.
