EPI's weekly illustration of education data trends — January 14, 2015
There has been a habitual argument that the SAT tests significantly defer to more affluent people, based on the belief that those who earn more (a) have a history of educational legacy surpassing those from lower earning strata; and (b) have greater access to high-quality learning for their children. This belief is supported by a plethora of research.
The 2010 cohort data utilized in the graph below is courtesy of The College Board. The colored bars represent the three SAT tests: Critical Reasoning (Blue), Mathematics (Red), and Writing (Green). The two lines represent High School GPA (Purple) and First-Year College GPA (Black).
As illustrated, SAT scores follow household income in a stepwise manner, correlating very closely, giving credence to the notion that income has a dynamic impact on the abilty to post better SAT scores. The positive differential in average scores in the Critical Reasoning test is 83 points between affluent (<200k) and low-income (<$40k) households. The gap is 88 points for mathematics and 95 points for writing.
The impact on first year GPA is also correlated with income, with a differential of 0.38 GPA points. Interestingly, the high school differential between income and GPA is almost nonexistant (0.09).
So, why is high school GPA the least impacted by household income? It could be because high school courses and grades are not normed and less elastic than the other indices. One could argue the same for first-year GPA, but those grades are arguably more informed by test scores and less on issues related to teacher bias and classroom grade work. This stated, it is interesting that high school GPA remains a better predictor* of first year GPA than any of the SAT tests (although marginally).
*The correlation of HSGPA is 0.55 to FYGPA. The SAT-CR correlates at 0.50; the SAT-M at 0.49; and the SAT-W at 0.54. When combined impact of the four items correlates to 0.63 on FYGPA.
SOURCE: Patterson, Brian F., and Mattern, Krista D. (2013). Validity of the SAT for Predicting First-Year Grades: 2010 SAT Validity Sample. New York, NY: The College Entrance Examination Board.