Armor, David J., & Duck, Stephanie
The Effect of Black Peers on Black Test Scores
George Mason University
Estimate the effect of peer characteristics on student achievement using state testing data from NC and SC and national data from ECLS.
Journal Name or Institutional Affiliation:
- None of the analyses presented in the study find the magnitude of Black peer effects that Hanushek, Kain and Rivkin found in their Texas study.
- Despite the fact that these three data sources involve different populations of students, different achievement tests, and different measures of student backgrounds and school characteristics, the results for Black peer effects are remarkably consistent with one another. They all start with moderate relationships between racial composition and black reading and math achievement, but after controlling for student background, lagged achievement, selected school and teacher characteristics, and (for the state analysis) fixed effects for year by grade and school by grade, only one of the six Black peer coefficients is statistically significant.
- It does not appear that Black students who spend most of their elementary and middle school years in predominantly Black schools are adversely affected by high concentrations of Black peers, once we take into account their family background.
- Results are more consistent with the theory that states that family background is the primary cause of lower Black scores in schools with higher concentrations of Black students.
Academic Achievement, Math, Peer Effects, Racial Composition, Reading, SES
Secondary Survey Data
Method of Analysis:
Two Stage Least Squares
Students in NC & SC
Unit of Analysis:
- North Carolina "End of Grade" Tests and South Carolina Palmetto Achievement Challenge Tests (PACT). ECLS data.
- This study uses NC state testing data for grades 3 to 8 and from 1997 to 2005. The data used for this analysis includes four cohorts of students who attended grades 3 to 8 between 1997 and 2005. The first cohort attended grade 3 in 1997 and grade 8 in 2002; the last cohort attended grade 3 in 2000 and grade 8 in 2005.
- Total NC analysis sample consists of nearly 500,000 different students.
- Total SC sample is nearly 230,000 different students.
- Analyzed results of models with fixed effects and without fixed effects, replicating HRK analysis.
- Two basic forms of the analysis, one using aggregate data and one using individual student data.
- Aggregate data was obtained by averaging all variables by race within the groups formed by the cross-classification of cohort (year), school, and grade. This yields approximately 16,800 observations for Black students and 17,800 for White students.
- For the individual student analysis using the stacked gain and lag scores, there are approximately 500,000 observations for Black students and 1.1 million observations for White students.
- DV: math and reading test scores
- IV: % black students in each grade level at each school, % free lunch, parent education, years in same school, % highly qualified teachers, % certified teachers, pupils per teacher,
- Then models include controls for lagged math & reading scores
- Proceeds to compare NC and Texas black peer effects for math and reading
- NC (reading and math test scores, grade, student race/ethnicity, free lunch status, and parent education), SC (the same as NC but no parent education).