How we analyzed data about Holyoke’s ethnic studies program

We obtained anonymized data from Holyoke Public Schools on the 473 students enrolled in ninth grade in 2016-17 for this project.

We obtained anonymized data from Holyoke Public Schools on the 473 students enrolled in ninth grade in 2016-17, including their race, gender, English learner status, special education status, low-income status, grades, attendance, and standardized test results. We excluded 207 students from the analysis: those who took their first high school ethnic studies courses in or after 10th grade; those who had already been in ninth or 10th grade prior to the pilot year; students who transferred out of Holyoke schools; and students who were not enrolled for the full academic freshman year (less than 180 days). For the groups remaining, we conducted a multiple regression analysis. The groups studied were the ethnic studies (ES) cohort — those who opted to take at least one ethnic studies class their freshman year or took both English and History ethnic studies classes that school year – compared with those who never took an ethnic studies class in high school.

To predict the impact of taking ethnic studies courses, we controlled for the following factors: students’ race, gender, English learner status, special education status, low-income status, eighth-grade attendance, and standardized test results. We ran regression models to assess relationships between taking ethnic studies classes and whether students graduated/dropped out within five years, whether they earned meeting expectations or higher in the 10th grade English and math MCAS test, and the percentage of their letter grades of A, B, Cs and attendance rate each year in high school. We found a robust impact on increased graduation and decreased dropout among students who took ethnic studies courses. Even accounting for margin of error, the analysis suggests a positive impact for graduation, negative impact for dropout among the ES cohort. We also found some evidence of a positive impact from taking the ES courses on grades and attendance rates in ninth or 10th grades, but no clear impact in later years.

— Yuriko Schumacher, graduate student at Northeastern University, with guidance from Arindrajit Dube, professor of economics at University of Massachusetts, Amherst.