Claim Missing Document
Check
Articles

Found 2 Documents
Search

Pemberdayaan Guru SMP Melalui Pengoptimalan Aplikasi Artificial Intelligence Berbasis Personalized Learning Imam Rofiki; Puguh Darmawan; Tuti Mutia; Mutiara Sani; Laila Nurfitriya; Syekha Vivi Alaiya; Sri Wahyuni
Jurnal Pengabdian Pendidikan Masyarakat (JPPM) Vol 7 No 2 (2026): Jurnal Pengabdian Pendidikan Masyarakat (JPPM) Vol.7 No 2 (Oktober 2026)
Publisher : LPPM UNIVERSITAS MUHAMMADIYAH MUARA BUNGO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52060/jppm.v7i2.4093

Abstract

A recurring issue in education is students’ difficulty in understanding complex concepts due to the limited availability of contextual learning media, visual representations, and technology integration. Limited teacher access to digital literacy further exacerbates this challenge. Therefore, innovation is needed to facilitate teachers in planning interactive, in-depth, and personalized learning. This community service activity aims to enhance teachers’ ability to integrate artificial intelligence, deep learning, and personalized learning through training and the PEDAGOGI website. The program involved 46 teachers from SMP Negeri 2 Gondang. This community service activity is carried out through three stages, namely preparation, implementation, and evaluation. The training introduced AI, deep learning, and personalized learning, followed by practical sessions on utilizing AI to develop deep learning–based lesson plans, interactive learning media, and produce personalized learning–based assessments. The results indicate a significant improvement in teachers’ competencies, reflected in their ability to produce 182 technology-based instructional materials. Program evaluation showed that 94.6% of teachers were satisfied with the implementation and provided positive feedback regarding the usefulness, clarity, and relevance of the materials. Improvements in the partners’ empowerment level were evident in two aspects. In the socio-educational aspect, teachers demonstrated increased confidence in using AI, improved conceptual understanding (shifting from predominantly levels 2–3 to levels 3–4), and increased frequency of AI usage in classroom planning. In the production aspect, the number of instructional products increased by 78% compared to the initial baseline, indicating enhanced teacher autonomy in producing digital learning materials.  
MathVerse M6: Exploring Students’ Mathematical Critical-Thinking Performance in AR–Metaverse-Supported Learning Ardi Dwi Susandi; Puguh Darmawan; Siti Fatimah; Thesa Kandaga; Imam Rofiki; Entit Puspita; Hasan Basri
Online Learning In Educational Research (OLER) Vol. 6 No. 4 (2026): Online Learning in Educational Research
Publisher : CV FOUNDAE

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58524/oler.v6i4.1321

Abstract

Mathematical critical thinking involves distinct reasoning processes that are rarely evaluated separately in immersive technology-supported mathematics. This study aims to develop and formatively assess MathVerse M6 and describe overall and indicator-level change and exploratory phase-linked associations. An educational design research approach included 16 mathematics teachers in preliminary needs analysis and 118 Grade 9 students in a small trial (n = 28) and a large one-group pre–post-test trial (n = 90) across four schools in Cirebon, Indonesia. Three experts rated five artefacts, and classroom implementation and student responses were observed. Mean expert ratings were 3.48–3.82/4. The large-scale mean mathematics score increased from 48 to 82; mastery (score ≥75) increased from 0/90 to 75/90 (83.3%). Inference (+38 percentage points) and analysis (+35) showed the largest descriptive indicator changes; self-regulation (+18) showed the smallest. Previously reported phase and artefact correlations used only six and five aggregate pairs and remain exploratory. MathVerse M6 was judged suitable for trial and showed promising within-group changes, but the absence of a control group, clustered classes, single-item indicators, and inconsistent effect-size reporting preclude causal conclusions.