High School Grade Point Average (HSGPA) is a reliable and consistent indicator of first year college success. The predictive advantages HSGPA holds over standardized tests has been well documented through numerous “Predictive Validity” studies. Such studies have been conducted by post-secondary academic institutions assessing possible correlations between admissions data and academic “success” among first year students, whereas such success being defined as students receiving a passing grade in enrolled courses. Standard grading procedures determine students receiving a grade of C or better within enrolled courses are deemed as passing. HSGPA has been shown to have the highest positive correlation with freshman grades; however, the accuracy of placement predictive measures increases upon the integration of both HSGPA and standardize placement test such as the SAT. For example, a study titled “UC and the SAT: Predictive Validity and Differential Impact of SAT1 and SAT2 anh the University of California. Geiser with Studley 2003.) concluded that HSGPA, and more specifically HSPA in college prep courses had the highest predictive power of freshman GPA of students who gained admission to UC. The study also produced findings that second to HSGPA, “achievement-type” standardized test such as the SAT 2 were the best predictors of freshman GPA.
In the research study “Predicting College Grades Using Multiple Regression Equations” (Bridgeman, Burton and Pollack) examined the added variance of SAT scores in in predictive models’ vs HSGPA alone. For impact analysis purposes the SAT scores were divided into 5 categories. Each category reflects a numerical score that was derived through the combination of verbal and math scores and is defined as follows: Category 1 = 200 – 800, Category 2 = 810 – 1000, Category 3 = 1010 – 1200, Category 4 = 1210 – 1400, Category 5 = 1410 – 1600. Students with comparable HSGPA’s (3.3 – 3.7 and 3.7 – 4.0) were assigned an SAT category based on individual SAT results. Following standard protocol, multi-variate corrections were implemented for range restriction to estimate correlations in the potential applicant population, not just in the restricted group of students enrolled. By applying these corrections implications are made, furthermore assuming future applicants will have previously taken the SAT. Such implications were the basis for calculating a standard correction using standard deviations for the SAT and HSGPA.
Individual regression algorithms were derived per institution to calculate the correlation co-efficient. After the selectivity of each institution was assessed, a weighted average of correlations was calculated. Average correlations suggested that HSGPA was a slightly better predictor than the SAT, though the relationship was not consistent throughout individual institutions. Predictive results showed consistency between colleges in all selectivity groups. Regression analysis results implicate that SAT scores only explained a variance of less than 10% in college grades overall. This should not be interpreted as having a low impact on potential student success. Despite the low variance explanation, SAT categories had a profound effect