@techreport{ai26-1486, title = {{Hmong but not Asian, Sāmoan but not Pacific Islander: Tracing the ECLS-K Racial Data (Mis)Classification Journey}}, author = {Wendy Castillo and Daranee Taychachaiwongse Teng and Kristine Jan Cruz Espinoza}, institution = {Annenberg Institute at Brown University}, type = {EdWorkingPaper}, number = {26-1486}, year = {2026}, month = {May}, doi = {10.26300/wk9y-sm74}, url = {https://edworkingpapers.com/ai26-1486}, abstract = {Race is a socially and politically charged concept that remains contested in the United States. We examine racial data (mis)classification in the Early Childhood Longitudinal Studies (ECLS-K) dataset. Centering the racial data journey of Asian American, Native Hawaiian, and Pacific Islander (AA\&NHPI) students, we find two types of racial data (mis)classification: (1) racial reformation related to the reconfiguration of parent/caregiver-reported racial data and (2) categorical friction when ethnicity was parent/caregiver-reported and race was not. Educational data practices and datasets like ECLS-K play a role in obscuring differentiated educational outcomes by operationalizing the myth that AA\&NHPIs are a monolith. We offer recommendations for addressing racial data (mis)classification and engaging a critical race research praxis.}, }