<![CDATA[This study examines the relationship between age, gender, and educational level and self-efficacy in learning Arabic among Indonesian high school students. This non-experimental survey involved 1,008 students aged 12–18 years: 332 from junior high schools/madrasahs (SMP/MTs) and 676 from senior high schools/madrasahs (SMA/MA). Self-efficacy was calculated as the average of 16 items with four response categories. Linear regression with robust HC3 covariance estimated the main effects and three two-way interactions, with age centered at 15 years. The model explained 1.4% of the variance in scores (adjusted R² = 0.008); mean scores were nearly identical across gender (2.93 for males and 2.93 for females) and educational level (2.95 for SMP/MTs and 2.92 for SMA/MA). The age × education level interaction was small (B = 0.058, 95% CI [0.005, 0.110], p = 0.031), but was no longer significant after Bonferroni correction (p = 0.094) or when using standard errors grouped by school label (p = 0.119). Estimates for the 14–16-year-old subgroup showed the same direction but were not precise.The three-factor scale indicated insufficient model fit, so construct validity requires further examination. Additionally, the overlap in age ranges across educational levels limits the separation of the effects of age and education. These findings suggest that demographic characteristics do not sufficiently explain self-efficacy in Arabic language learning. Theoretically, these results underscore the importance of examining learning experiences as a factor more closely linked to the development of self-efficacy. Therefore, teachers should not assume that students’ self-efficacy is determined by gender or educational level.]]>
