This study was motivated by the issue of ineffective and subjective assessment of basic manipulative movements by physical education teachers in elementary schools. The purpose of this study was to develop, test the feasibility, and test the effectiveness of the AI MediaPipe-based Manipulative Basic Movement Assessment Application as an objective digital assessment tool. The method used was the Borg and Gall Research and Development (R&D) model with a sequential explanatory design approach. The product was developed through ten stages and tested on 200 respondents (160 students, 40 PE teachers). Data analysis used descriptive and inferential statistics (t-test, Pearson correlation, ANOVA). The results showed that Masivmention was successfully developed as a web application that analyzes three manipulative movements. The product was deemed highly feasible based on validation by subject matter experts (83.3%), media experts (81.3%), and movement instrument experts (84.3%). The product effectively improved the objectivity of assessment with a significant difference between manual scores and application scores (t=6.234; p=0.000) and a strong correlation (r=0.723). In conclusion, Masivmention based on MediaPipe AI is a feasible and effective instrument for assessing manipulative movements in elementary schools.
Copyrights © 2026