Author finds Black-White test score gaps at kindergarten entry in 2010 in reading (SD = .32), math (SD
= .54), and working memory (SD = .52 among children with valid scores).
Math and reading gaps widened by approximately .06 standard deviations over kindergarten, but the
working memory gap was constant.
Multivariate regressions show that student SES explained the reading gap at school entry, but gap
decompositions suggest that school quality differences were responsible for the widening of the
reading gap over kindergarten.
SES explained much of the math gap at school entry, but the widening of the math gap could not be explained by SES, school quality, or other hypotheses.
Standardized Black-White math and reading gaps widened, not because Black students learned less
than White students but because test score variance shrank from fall to spring.
SES related controls explain about 75 percent of the Black-White math gap.
The same controls dramatically reduce the Hispanic-White fall reading (-.57 to 2.13) and math (-.70
to 2.23) gaps but reduce the WM gap less (from 2.53 to 2.26
However, Black students lost their adjusted reading advantage by spring, with the adjusted gap
going from .08 to 2.05, and this loss was statistically significant.
SES did not explain the widening gaps; in fact, SES- adjusted gaps widened more than unadjusted gaps. Adjusted math and reading gaps widened primarily because Black students learned less math and reading over kindergarten than did White students from similar backgrounds. Additionally, author tested whether the gap widening in math could be explained by (1) racial differences in school readiness or prekindergarten experiences, (2) racial differences in parental beliefs about the importance of math, (3) differential math learning rates stemming from the fall working memory gap, (4) student-teacher racial mismatch, or (5) Black students being assigned to less effective teachers than White students within the same schools. These exploratory analyses did not produce convincing supporting evidence; however, data for these analyses are limited and further research is needed.
Student SES could not explain why math and reading gaps widened over the school year; in fact, controlling
for background, Black students learned less math and reading than did White students.
School proportion Black negatively predicts students’ math scores in the fall, but positively predicts
students’ math scores in the spring (controlling for background variables)