<![CDATA[This study analyzes the efficiency of the learning media development process carried out by prospective mathematics education teachers using the Data Envelopment Analysis (DEA) approach. The sample consists of 30 graduates of the mathematics education study program from 2022–2024, whose final projects on the development of learning media or teaching materials are treated as Decision Making Units (DMUs). The input variables include the grade in the learning media/multimedia course and the duration of the research, while the output variables consist of the percentage scores of content expert validity, media expert validity, teacher practicality, and student practicality. The analysis was conducted using an output-oriented DEA model under the assumptions of Constant Returns to Scale (CRS) and Variable Returns to Scale (VRS). The results show that the CRS model identifies 11 DMUs (36.67%) as efficient, whereas the VRS model identifies 16 DMUs (53.33%) as efficient, and the average efficiency scores of both models lie in a high range (≥0.93). The slack calculations indicate that several DMUs still need to improve their output quality, particularly in the components of content validity and teacher practicality. These findings highlight the need to strengthen supervision and systematic mentoring to enhance the content quality and practicality of learning media developed by prospective teachers. This study is limited by the relatively small number of DMUs and its focus on a single study program, so the generalization of the results should be made cautiously, and further research with a larger sample and more diverse DEA model specifications is recommended.]]>
