In the education world, the quality of the data underpins the quality of the research, and if there was ever an example of that, it’s in the National Center for Education Statistics’ report released this morning, looking at achievement gaps between white and Hispanic students.
NCES’s analysis of trend data in the National Assessment of Educational Progress presents a pretty grim picture of stubborn academic gaps between white and Hispanic students, who now make up one in five students in American classrooms (and a majority in many classrooms in the West and South.) I’m already waiting for the slew of secondary research, from the Institute of Education Sciences and others, to parse out why, after decades of interventions and a federal law primarily focused on closing the achievement gap, we have made so little progress.
Yet the report itself is an interesting look at what progress has been made in the data on Hispanic students, and particularly Hispanic English-language learners. For much of NAEP’s history, English-language learners were not automatically disaggregated in student subgroup reporting, and in the years after No Child Left Behind was passed, the National Assessment Governing Board, which oversees the tests, developed new working groups to revisit how NAEP looked at ELLs. In the process, the group specifically split those with low English proficiency from students with disabilities, according to Kathleen Leos, the former director of the U.S. Department of Education’s office of English-language acquisition, who consulted in those early discussions.
“This laser-like focus from NAEP on the subgroups is extraordinary, and a long time coming,” said Kathleen Leos, now the president and chief executive officer of the Washington-based Global Institute for Language and Literacy Development.
In 2009, NAGB started a serious review of how states include English-language learners in their NAEP testing, finding widely disparate practices that led to anywhere from 6 percent to 70 percent of ELLs and students with disabilities left out of the exams, depending on the state. In 2010, the advisory group approved new policies intended to ensure the “nation’s report card” included a bigger and more representative sample of ELLs.
Yet in a briefing with reporters prior to the Hispanic-student report’s release, NCES Commissioner Sean P. “Jack” Buckley said the ELL data in the report is still hard to compare across states because of continued variation in the ways states provide accommodations for English-language learners taking the NAEP.
Ms. Leos told me the Education Department and NAGB still has a lot of work to do for ELL data quality within different student groups, particularly since state-level NAEP data must be included in NCLB accountability report cards and is so frequently used for research.
“The districts still get to cherry-pick the students they think will do well,” she said. “We really want to move beyond that to a really good solid sampling of the students in the school system who can take the assessment. We need to find out how to get the broadest range of student participation and the most valid scores possible, so you really understand the snapshot of student achievement for all student groups.”