It’s widely believed that racial gaps in test scores are just class gaps. And, if that’s not true, then it’s assumed that race is fading away in importance relative to class. But an important study shows that in multiracial California, race is becoming more influential in recent years.
THE GROWING CORRELATION BETWEEN RACE AND SAT SCORES: NEW FINDINGS FROM CALIFORNIA
Center for Studies in Higher Education
University of California, Berkeley
This paper presents new and surprising findings on the relationship between race and SAT scores. The findings are based on the population of California residents who applied for admission to the University of California from 1994 through 2011, a sample of over 1.1 million students. The UC data show that socioeconomic background factors – family income, parental education, and race/ethnicity – account for a large and growing share of the variance in students’ SAT scores over the past twenty years. More than a third of the variance in SAT scores can now be predicted by factors known at students’ birth, up from a quarter of the variance in 1994. Of those factors, moreover, race has become the strongest predictor. Rather than declining in salience, race and ethnicity are now more important than either family income or parental education in accounting for test score differences. It must be cautioned that these findings are preliminary, and more research is needed to determine whether the California data reflect a broader national trend. But if these findings are representative, they have important implications for the ongoing debate over both affirmative action and standardized testing in college admissions.
… The regression results show a marked increase since 1994 in the proportion of variance in SAT scores that can be predicted from socioeconomic background factors largely determined at students’ birth. After falling slightly from 25% to 21% between 1994 and 1998, the proportion of explained variance increased each year thereafter, growing to 35% by 2011, the last year for which the author has obtained data. Remarkably, more than a third of the variance in SAT scores among UC applicants can now be predicted by family income, education, and race/ethnicity. This result contrasts sharply with that for high school GPA: Socioeconomic background factors accounted for only 7% of the variance in HSGPA in 1994 and 8% in 2011. …
Nevertheless, even without being able to observe those intermediating experiences directly, regression analysis enables one to assess the relative importance of different socioeconomic factors in predicting test performance. Figure 2 provides standardized regression coefficients, or “beta weights,” for predicting SAT scores conditional on family income, parents’ education, and race/ethnicity. The coefficients show the predictive weight of each factor after controlling for the effects of the other two, thereby providing a measure of the unique contribution of each factor to the prediction.
In 1994, at the beginning of the period covered in this analysis, parental education was the strongest of the three socioeconomic predictors of test performance. (The standardized regression coefficient of 0.27 in that year means that, for each one standard deviation increase in parental education, SAT scores increased by 0.27 of a standard deviation, when income and underrepresented minority status were held constant.) The predictive weight for parental education has remained about the same since then. The weight for family income has shown a small but steady increase from 0.13 in 1998 to 0.18 in 2011. But the most important change has been the growing salience of race/ethnicity. By 2011, the predictive weight for underrepresented minority status, 0.29, was greater than that for either family income or parental education. When the regression results for the UC sample are pooled across applicant cohorts, race/ethnicity is the strongest predictor of SAT scores over the last four years.
A key implication of this finding is that racial and ethnic group differences in SAT scores are not simply reducible to differences in family income and parental education. At least for the UC sample, there remains a large and growing residual effect of race/ethnicity after those factors are taken into account.
As shown in Figure 8, the test score gap in California is greatest between black and white SAT takers but has oscillated up and down and shows no consistent trend since 1998. If one were to draw inferences about racial and ethnic differences from the black-white gap alone, one might conclude that there has been little change in this respect.
But that conclusion would be wrong. For all other racial/ethnic comparisons, test score gaps between underrepresented minority and other students have been growing. The Black-Asian, Latino-White, and Latino-Asian test score gaps have increased almost every year since 1998.