Pre-service Indonesian language teachers frequently struggle to translate Learning Outcomes (Capaian Pembelajaran) into concrete Learning Objectives and Trajectories (Tujuan and Alur Tujuan Pembelajaran), while South Sumatran local wisdom remains largely absent from their lesson designs and unsupervised generative-AI use further erodes critical digital literacy. This study reports a preliminary needs assessment and formative pilot evaluation of PAMONG-AI, a Socratic, local-wisdom-based pedagogical chatbot developed to accompany the Lesson Planning for Indonesian Language and Literature course at Universitas Sriwijaya. A sequential mixed-methods design combined a 20-item baseline competency test (N = 73) with a 51-item post-trial questionnaire (N = 80) covering technology acceptance, Socratic-interaction quality, local-wisdom integration, curriculum self-efficacy, usability, and dissemination readiness, supplemented by open-ended responses analyzed thematically. Results showed low baseline mastery (M = 33.8%, SD = 10.98), with the weakest areas being Deep Learning pedagogical principles (5%), learning-model syntax (9%), Modul Ajar structure (14%), and Kurikulum Merdeka/Kurikulum Berbasis Cinta distinctions (18%), precisely the gaps the chatbot was designed to close. After piloting the prototype, students reported favorable technology acceptance (M = 3.62/5), valued its Socratic questioning (M = 3.54/4) and local-wisdom integration (M = 3.54/4), and indicated improved curriculum self-efficacy (M = 3.53/4), while usability scored moderately (adapted SUS = 56.8/100), signalling concrete areas for refinement before expert validation. These findings validate the chatbot’s design rationale and provide an empirical foundation for its next development iteration.
Copyrights © 2026