<![CDATA[To address this gap, this study aims to evaluate the temporal stability and reliability of the Local Wisdom-Based Religious Moderation Attitude Questionnaire for Islamic Junior High School (Madrasah Tsanawiyah/MTs) students using the Rasch Modeling paradigm. This study employed a quantitative approach, specifically utilizing a cross-sectional survey design combined with a test-retest method over a three-week interval. Involving a cohort of 375 students from six MTs in South Kalimantan Province, the research evaluated a 16-item instrument utilizing a 4-point forced-choice Likert scale. Data were computationally analyzed using the Rating Scale Model (RSM) via Winsteps software to extract interval reliability metrics, temporal stability, scale diagnostics, unidimensionality, and item fit. The empirical analysis confirmed that the instrument fulfilled the unidimensionality assumption, with a raw variance explained of 48.5%, substantiating that the dimensions integrate into a single construct. The test-retest assessment demonstrated robust temporal stability, evidenced by the logit correlation of the Person Measure (0.89) and Item Measure (0.94), indicating the instrument's resilience against temporary cognitive fluctuations. Furthermore, the internal reliability proved to be excellent (Cronbach's Alpha 0.88; Person Reliability 0.86; Item Reliability 0.95), complemented by a person separation index (2.54) that accurately differentiates students' affective strata. Scale diagnostics verified that the exclusion of a neutral option allowed the instrument to operate hierarchically without threshold disordering, and all 16 items were proven to fit the model. In conclusion, this local wisdom-based moderation instrument exhibits exceptional logical consistency and temporal stability. Therefore, it is highly recommended as an early warning system and a standard for affective curriculum evaluation for educational policymakers]]>
Copyrights © 2026
