Abstract / Summary
Education for sustainable development (ESD) emphasizes critical and systems thinking, and both depend on students' ability to read and interpret data such as Sustainable Development Goal indicators. Statistics anxiety is a documented barrier to this kind of data literacy, yet few multidimensional measures of it are available for Indonesian higher education. This study reports a preliminary psychometric evaluation of an Indonesian version of the 51-item Statistics Anxiety Rating Scale (STARS), in which items referring to outdated technology were updated. Data came from 100 undergraduate psychology students at Universitas Sumatera Utara. Each of the six subscales was analyzed with the Rasch rating scale model to examine item fit, rating scale functioning, dimensionality, reliability and separation, and differential item functioning (DIF) by high school track and entry cohort. Correlations with statistics course grades (n = 69) provided criterion-related evidence. Forty-five of the 51 items fit the model. Person reliability ranged from .71 to .86, and thresholds were ordered in every subscale, although respondents barely distinguished between the two highest categories of Worth of Statistics. Four items showed DIF by entry cohort, which appears to reflect whether students had already taken a statistics course. Anxiety subscales correlated negatively with course grades (ρ = -.28 to -.40). The Indonesian STARS shows promising measurement properties, but several items need review, and the results remain preliminary until back translation and expert review are completed.
Keywords: statistics anxiety, Rasch model, data literacy, education for sustainable development, Indonesia