Electronic Theses and Dissertation

Media digital berupa komik digital berbasis deep learning dapat menjadi alternatif untuk meningkatkan motivasi belajar dan pemahaman konsep IPA pada materi Gerak. Penelitian ini bertujuan mengembangkan media digital berbasis deep learning serta menguji validitas, kelayakan, dan efektivitasnya. Penelitian menggunakan metode Research and Development (R&D) dengan model ADDIE. Data kualitatif diperoleh dari saran validator, sedangkan data kuantitatif berasal dari angket validasi, angket kepraktisan, serta hasil pretest dan posttest. Analisis data menggunakan indeks Aiken’s V, uji paired sample t-test, dan N-Gain. Hasil menunjukkan media memiliki validitas tinggi dengan nilai Aiken’s V sebesar 0,81 dari ahli materi dan 0,80 dari ahli media. Motivasi belajar siswa meningkat dari rata-rata 2,8 menjadi 4,3 (p < 0,05; N-Gain = 0,7, kategori tinggi). Pemahaman konsep IPA juga meningkat dari rata-rata 35,28 menjadi 78,56 (p < 0,05; N-Gain = 0,7, kategori tinggi). Dengan demikian, komik digital berbasis deep learning dinyatakan valid, layak, dan efektif untuk meningkatkan motivasi belajar serta pemahaman konsep IPA siswa pada materi Gerak.

Digital media in the form of a deep learning-based digital media can serve as an alternative to enhance students' learning motivation and understanding of science concepts on the topic of Motion. This study aimed to develop a deep learning-based digital comic and evaluate its validity, feasibility, and effectiveness. The study employed the Research and Development (R&D) method using the ADDIE model. Qualitative data were obtained from validators' suggestions, while quantitative data were collected through validation questionnaires, practicality questionnaires, and students' pretest and posttest results. The data were analyzed using Aiken's V index, the paired-samples t-test, and the N-Gain test. The results indicated that the developed media had high validity, with Aiken's V values of 0.81 from the subject matter expert and 0.80 from the media expert. Students' learning motivation increased from a mean score of 2.8 to 4.3 (p < 0.05; N-Gain = 0.7, high category). Their understanding of science concepts also improved from a mean score of 35.28 to 78.56 (p < 0.05; N-Gain = 0.7, high category). Therefore, the deep learning-based digital comic was found to be valid, feasible, and effective in improving students' learning motivation and understanding of science concepts on the topic of Motion.