SAT-Based Scholarship Programs Exclude African Americans

Status: 
Archived
Subject: 
University Testing

As more colleges and universities deemphasize standardized test scores and revamp their methods for making admissions decisions, a new study affirms that scholarship programs based on SAT scores can unfairly exclude high-achieving African American students.

 

In "Incorporating Noncognitive Variables in the Scholarship Selection Process of African American Students," presented this October at the annual National Association for College Admissions Counseling convention, Diane Warmsley of Queens College at the City University of New York (CUNY) examined African American sophomores at five colleges within the CUNY system, all of whom had achieved first-year grade-point averages of 3.0 or higher and none of whom were merit scholarship recipients.

 

Warmsley found that these students' first-year college academic achievements were significantly better predicted by their responses to a Noncognitive Questionnaire (NCQ) than by their SAT scores. The NCQ was designed by University of Maryland Professor William Sedlacek and T.J. Tracey and assesses such characteristics as positive self-concept, realistic self-appraisal, leadership, and academic interest and familiarity. According to the CUNY study, the five most significant NCQ variables accounted for about one-third of the variance of the study group's first-year grades; for the SAT, the figure was just over seven percent.

 

Warmsley concludes that the students she studied "may have been overlooked for scholarship consideration as their cognitive scores [SAT and high school grade-point average] were not exceptional," but "their ability to perform at a high caliber level is apparent." She recommends that "scholarship selection criteria should include noncognitive measures as a primary criterion." The study's sample size was small (24 students), so the conclusions are not "readily generalizable," according to Warmsley, but "the study [should] be replicated on a broader scale, to see if similar levels of validity are obtained."