Black-White Summer Math Gap Trends: The Sensitivity of Estimates to Various Reasonable Representations

• In Event: Poster Session 14
  In Poster Session: Division L Section 7 Poster Session

Abstract

The estimation of racial test score gap trends plays an important role in monitoring educational equality. Gap trend estimation is complex, however, and estimates can vary depending on test metric, modelling strategy, consideration of test measurement error, and whether the test scale is assumed to be interval. Sensitivity to these factors has not been examined in research on racial summer gap trends. In this study, I use data from the nationally-representative ECLS-K to examine the sensitivity of black-white summer math gap trend estimates to modelling and psychometric assumptions. I find that estimates differ meaningfully across various reasonable representations. I discuss lessons for researchers, as well as contributions these findings make to our understanding of the black-white summer math gap trend.

Author

• David Quinn, Harvard University