Development of Artificial Intelligence (AI)-Based Digital Evaluation Applications to Improve Quality of Feedback for PGSD Students’ Formative

Abstract

The objective of this study was the development and implementation of an artificial intelligence (AI)–based digital assessment application to improve the quality of formative feedback for students of the Primary School Teacher Education (PGSD) Program at Nggusuwaru University. The study employed a Research and Development (R&D) approach using the ADDIE model, which included needs analysis, design, development, limited implementation, and evaluation. Expert validation placed the application in the “valid” category (mean score 3.31), while practicality testing by lecturers and students produced a mean score of 3.20 (“practical”). Pretest–posttest effectiveness testing in two classes (n = 40) showed a significant improvement in learning outcomes (increase of 6–7 points; p < 0.05). The quality of formative feedback was also rated positively by students, with an average response time of 4–5 seconds and perception scores above 3.2. System log analysis during two weeks of implementation recorded an average session duration of 9–10 minutes, page load times of under 3 seconds, an error rate reduction from 4.5% to 3.8%, and uptime over 98%, indicating good system stability. Qualitative findings supported the quantitative results, showing that the application accelerated formative feedback but still required refinement of feedback language and performance optimization under low-network conditions. These findings indicate that the AI-based digital assessment application is feasible for supporting formative evaluation of PGSD students and has the potential for wider implementation in higher education.