Policy Implications of a Predictive Validity Study of the Specialized High School Admissions Test at Three Elite New York City High Schools

Policy Implications of a Predictive Validity Study of the Specialized High School Admissions Test at Three Elite New York City High Schools By Jonathan Taylor Adviser: Dr. Herbert Saltzstein Although admission to New York City's elite public high schools has been controversial because of the disproportionate representation by ethnicity and gender of students admitted, there has been no research on the Specialized High School Admissions Test (SHSAT), the sole admissions criterion. This dissertation had four primary questions: (1) What is the predictive validity of the SHSAT? (2) Would the use of NYC achievement test scores improve prediction? (3) Does the SHSAT exhibit equal predictive validity across gender? and (4) Do disadvantaged students admitted to the Discovery Program with test scores below the cutoff earn grades comparable to regularly admitted students? These research questions were analyzed using the following data provided by the Department of Education: SHSAT scores for 27,905 students; school grades for 2921 students enrolled in the three largest of these elite schools, Stuyvesant High School, Bronx High School of Science, and Brooklyn Technical High School; NYS achievement test and AP exam scores for all students in the sample who could be linked using de-identified ID numbers. This study found the predictive validity of the SHSAT was lower than that of NYC achievement tests. The combination of achievement test scores and SHSAT yielded better predictions than either variable alone. Prediction by the SHSAT was least precise in the range of scores around the cutoffs for admission. Gender bias was found in the form of under-prediction of girls’ grades by SHSAT scores. Discovery Program students achieved grades equal to those of regularly