An Evaluation of the Robustness of the National Assessment of Educational Progress Trend Estimates for Racial-Ethnic Subgroups

Abstract

The National Assessment of Educational Progress (NAEP) is the primary indicator of trends in the achievement of American students, and it has figured prominently in debate about the relative educational progress of racial and ethnic minorities. For about a decade, however, the NAEP assessment used to estimate long-term trends (trends of more than a few years' duration) has been different from the main NAEP assessment used for cross-sectional descriptions of performance. This article investigates the quality of performance estimates from the NAEP long-term trend assessment for Black and Hispanic students. NAEP's estimates of the trends for these groups are highly imprecise to the point where their usefulness for policymaking is arguable. In addition, the long-term trend assessment differs from the main NAEP assessment in numerous ways that could influence reported results (e.g., samples are defined in terms of age rather than grade, and a different method is used to classify the racial-ethnic identity of students). We recommend that an open debate on the assessment of minority group trends be undertaken, that design changes be considered to strengthen measurement of those trends, and that reporting of NAEP results be changed to make the differences between the NAEP main and trend assessments clear to key audiences.