This meta-analysis study evaluates the effectiveness of research and development (R&D) of Android-based educational games in enhancing students' mathematical reasoning and its moderating factors. From 13 primary studies that met the inclusion criteria, analysis using R Studio (Meta Package) and the Random Effects Model (REM) showed an overall effect size was categorized “moderate” ( \(0.988, 95\% CI \left[0.557, 1.419\right], p<0.01)\) , indicating that while R&D of Android-based educational games has a significant impact on improving students' mathematical reasoning, it is not entirely dominant or exceptionally large. Moderator analysis revealed that the effectiveness is influenced by the development model, educational level, region/continent, and class size, but not by publication type, independent variables, software used, subject area, or implementation duration. These findings support the R&D of android-based educational games in mathematics learning, especially if implemented using the 4D model, applied at the University level (by considering the number of students in the class) in countries in the Asian continent. In addition, these findings indicate that the effectiveness of R&D of android-based educational games does not depend on the type of publication, independent variables, software used, subject/branch of material, and duration of implementation, so this approach is flexible and effective in various situations and conditions.
