Looking to the Past, to help the Future: Using Retired NAEP Variables to Improve ELL Identification

Growing interest in the U.S. immigrant population and subsequent generations, particularly as it relates to student performance and educational opportunity, necessitates a clear understanding of how well foreign-born, and often non-native-English-speaking, students are performing academically on the National Assessment of Educational Progress (NAEP). NAEP reports on this population using an English Language Learner (ELL) classification status, a variable that is collected and reported independently by jurisdiction, which might differentially capture and describe members of this group based on idiosyncratic criteria. To assess the extent to which this occurs, we use two additional variables – Length of Residence in the U.S. and Language Use at Home – that have been collected in the past by NAEP to inspect and validate ELL-status. As of 2003, however, NAEP only collects one of these: how often the child reports a language other than English being spoken at home.1 NAEP data show that ELL students perform less well than non-ELL students; moreover, on average students who have not lived in the U.S. all their lives or students who live in households where a language other than English is spoken do not perform as well as their counterparts. While the three variables are undoubtedly associated, it is not clear how or to what extent they are related to each other in the context of NAEP scores. Thus, we investigate the comparability of Length of Residence in the US (LRUS)2 and Language Use at Home (LUH)3 in relation to NAEP mathematics achievement scores, net of ELL-status, while controlling for other important factors.