Abstract:

Formative assessment is widely recognized as a critical mechanism for supporting learning, promoting self-regulation, and enhancing student engagement in higher education. By providing ongoing feedback on students’ progress and understanding, formative assessment helps learners identify areas for improvement, adjust learning strategies, and develop greater responsibility for their own academic development. Despite its pedagogical importance, implementing formative assessment effectively at scale remains a significant challenge in large undergraduate courses. This study investigates the influence of artificial intelligence–supported formative assessment systems on learner self-regulation within large undergraduate courses.

In this paper, artificial intelligence (AI) is defined as computational systems capable of analyzing learner performance data and adapting feedback based on identified learning patterns. AI-supported assessment systems can generate personalized feedback messages, visualize learning progress, and recommend targeted practice activities that correspond to individual learner needs. Such systems have the potential to complement instructor feedback by providing immediate responses and continuous guidance throughout the learning process. However, their pedagogical implications, particularly regarding the development of students’ self-regulated learning skills, require further investigation.

The research employs a quasi-experimental design involving 624 undergraduate students enrolled in introductory social sciences courses at a large metropolitan university. Participants were divided into two groups: one group engaged with traditional instructor-led formative feedback practices, while the second group used an AI-supported assessment system integrated within the course’s digital learning platform. The AI system delivered personalized feedback, progress visualizations, and adaptive practice recommendations designed to help students identify learning gaps and improve their study strategies.

The results demonstrate statistically significant improvements in several self-regulatory behaviors among students who used the AI-supported formative assessment system. In particular, students in this group showed stronger development in monitoring their learning progress and adapting their study strategies in response to feedback. Qualitative reflections collected from participants further suggest that timely and individualized feedback enhanced learners’ sense of control over their learning processes and encouraged greater responsibility for managing their academic progress.

At the same time, the study identified several potential risks associated with AI-supported assessment systems. Some participants reported a tendency to rely heavily on automated recommendations, occasionally reducing opportunities for independent critical reflection on their learning strategies. These findings highlight the importance of maintaining a balance between automated guidance and opportunities for student agency.

The paper concludes by outlining design principles for AI-supported formative assessment systems that combine technological efficiency with pedagogical transparency. By integrating automated feedback with reflective learning activities and instructor guidance, such systems can support the development of autonomous and self-regulated learners in technology-rich undergraduate learning environments.

Keywords:

Formative assessment, artificial intelligence in education, self-regulated learning, learning analytics, undergraduate students.