Answers are grounded in this paper and cite their sources.
This community service program assisted teachers in using generative artificial intelligence (AI) to develop and review learning assessments at MA Muhammadiyah 1 Malang. Twenty-two teachers from different subjects participated in a face-to-face workshop and online mentoring from 24 April to 1 May 2026. Participants practised developing assessment items with ChatGPT, designing visual presentations using Canva AI, checking AI-generated outputs for alignment with learning objectives and cognitive demands, and revising their assessment products. Evaluation included 20-item pre- and post-tests, a product assessment rubric, and a 20-item participant response questionnaire. The mean test score increased descriptively from 15.95 to 18.41 out of 20. The mean product score was 86.08, with 11 products rated Good and 11 rated Very Good, while the questionnaire score reached 79.27% of the maximum possible score. Based on these evaluations, the mentoring program achieved its intended objectives during the training period. However, some assessment products still require greater cognitive depth in higher-order thinking tasks. These findings reflect outcomes observed during the program and do not establish the empirical validity of the assessments in classroom use or the long-term sustainability of teachers' skills.
No takes yet. Share an insight, caveat, or question.
Hafiizh et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: