Claim Missing Document
Check
Articles

Found 21 Documents
Search

Beyond coastal and mainland differences: A multilevel item response theory analysis of mathematical literacy in an archipelagic region Dwinata, Alona; Mariyanti Elvi; Linda Rosmery Tambunan
Al-Jabar: Jurnal Pendidikan Matematika Vol 17 No 2 (2026): Al-Jabar: Jurnal Pendidikan Matematika
Publisher : Universitas Islam Raden Intan Lampung, INDONESIA

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24042/ajpm.v17i2.29899

Abstract

Purpose: This study aimed to examine the mathematical literacy ability of junior high school students in Karimun Regency using a Multilevel Item Response Theory (IRT) framework and to investigate whether differences existed between students attending coastal and mainland schools. Method: A quantitative research design was employed involving 307 Grade IX students from five public junior high schools in Karimun Regency, Indonesia. Participants were selected through stratified two-stage cluster random sampling to represent both coastal and mainland school contexts. Data were collected using a 12-item mathematical literacy test developed based on the PISA mathematical literacy framework and local contextual situations. The instrument demonstrated satisfactory psychometric quality, with corrected item-total correlations ranging from 0.478 to 0.829 and a Cronbach’s alpha coefficient of 0.921. Data were analyzed using a Multilevel IRT two-parameter logistic (2PL) model to estimate item characteristics, students’ latent abilities, and school-level variation. Findings: The results indicated that item difficulty estimates ranged from -2.37 to 1.63, while discrimination parameters ranged from 0.39 to 2.90, suggesting that the instrument covered a broad range of difficulty levels and generally functioned well in differentiating students with varying abilities. Formulate items tended to be easier than interpret and evaluate items. The multilevel analysis revealed substantial school-level variation in mathematical literacy ability, with an intraclass correlation coefficient (ICC) of 0.617, indicating that 61.7% of the variance was associated with differences between schools. However, no significant difference was found between students from coastal and mainland schools (p = 0.906). Significance: This study contributes to the growing literature on mathematical literacy assessment by demonstrating the usefulness of Multilevel IRT in capturing both psychometric characteristics and hierarchical educational structures. The findings suggest that school-level factors play a more important role than geographical location in explaining students’ mathematical literacy ability, providing valuable evidence for designing targeted educational improvement programs in archipelagic regions.