Developing an AI-Assisted Case-Based Learning Model to Improve Students' Political Literacy

Abstract

Political literacy among university students has become a pressing concern in the age of digital democracy, yet the way it is currently taught often falls short. This study set out to develop and test an AI-Assisted Case-Based Learning (AI-CBL) model for the Introduction to Political Science course at STKIP Taman Siswa Bima. We followed the three-phase Plomp development model: preliminary research, prototyping, and assessment. Participants included three expert validators, three course lecturers, and 32 first-year students from the 2024/2025 academic year. We gathered data through validation sheets, practicality questionnaires, and a political literacy test, then analysed them using descriptive statistics, percentage scoring, and the N-Gain formula. Four findings stand out. First, the model proved highly valid, with a mean expert score of 3.84 on a 4-point scale. Second, the feasibility check returned an average index of 91.5%, placing the model in the "highly feasible" range. Third, both lecturers (89.7%) and students (88.4%) rated the model as highly practical. Fourth, students' political literacy improved sharply, with a mean N-Gain of 0.62 and 65.6% of them reaching the high-gain category. Taken together, these results suggest that AI-CBL works as a valid, feasible, practical, and effective approach for strengthening political literacy in higher education. The model also offers a concrete way of bringing generative AI into political education without sidelining the deliberative, case-based reasoning that the field requires.