<![CDATA[Background: Instructional media are widely implemented in Arabic vocabulary learning, yet their overall effectiveness and the contextual factors influencing their impact require quantitative synthesis. Purpose: This study aimed to estimate the pooled effect size of instructional media on students’ Arabic vocabulary mastery and to examine whether effect size variations were influenced by media type, sample size, and publication year. Method: A quantitative meta-analysis using a random-effects model was conducted on ten one-group pretest–posttest experimental studies at the junior secondary school level. Studies were systematically identified from SINTA-indexed journals and the official website of the Perkumpulan Pengelola Jurnal Kajian Bahasa Arab (PPJKBA) following PRISMA 2020 guidelines and predefined inclusion criteria. Effect sizes were calculated using Cohen’s d and corrected to Hedges’ g, with standardized mean differences computed using the pooled standard deviation of pretest and posttest scores as the standardizer. Statistical analyses were performed using JASP. Results and Discussion: The meta-analysis revealed a large and statistically significant pooled effect size (SMD = 2.549; 95% CI [1.388, 3.710]; p < 0.001), indicating substantial improvements in Arabic vocabulary mastery. However, heterogeneity was very high (I² = 94.43%), and the 95% prediction interval (−0.99 to 6.08) included zero, suggesting considerable variation in true effects across educational contexts. Moderator analysis showed no significant differences according to media type, sample size, or publication year. Potential publication bias was indicated by significant Begg’s test (Kendall’s τ = 0.854, p < 0.001) and Egger’s regression test (t = 8.448, p < 0.001). Conclusions and Implications: Instructional media have a strong positive effect on Arabic vocabulary learning outcomes and support the integration of diverse instructional media in vocabulary instruction, although the pooled effect size may be overestimated due to the exclusive use of one-group pretest–posttest designs. Future research is encouraged to explore additional contextual and pedagogical moderators.]]>
