A Culturally Grounded Data Literacy Assessment for Middle School Students: Development and Psychometric Evaluation

Abstract

Data literacy is increasingly important in middle school mathematics, yet many assessments capture only isolated skills and rely on contexts far removed from students’ everyday experiences. This study developed and psychometrically evaluated a culturally grounded data literacy assessment for Grade 8 students. The instrument contained 30 four-option multiple-choice items covering six dimensions: data knowledge; data management and preparation; analysis and interpretation; critical evaluation and reflection; communication and presentation; and ethical-situational reasoning. Six experts reviewed the instrument, and 128 students from two junior secondary schools in Medan, Indonesia, completed it. Evidence was examined using Aiken’s V, classical item analysis, Cronbach’s alpha, McDonald’s omega, and theory-guided exploratory factor analysis based on tetrachoric correlations. Mean Aiken’s V ranged from 0.919 to 0.924, item difficulty from 0.438 to 0.617, and point-biserial discrimination from 0.259 to 0.528. Fifteen items were retained, 10 were flagged for review, and five were prioritized for revision. Alpha ranged from 0.628 to 0.768, while omega ranged from 0.770 to 0.843. Data knowledge, analysis and interpretation, and critical evaluation and reflection showed the strongest psychometric performance. Communication and presentation and ethical-situational reasoning require refinement. Overall, the instrument supports low-stakes formative diagnosis, although broader or high-stakes use requires further independent empirical validation.