Abstract:

As generative AI becomes part of everyday study practices, higher education faces a practical design question: how can AI tools be moved beyond ad hoc question answering toward more systematic support for learning processes? This paper explores the potential of module-specific chatbots to support learning in higher education and examines how students use such tools in two business-related modules, Management Accounting and Investment & Finance.

Empirically, the study draws on an exploratory, descriptive survey investigating students’ awareness, adoption, use, and perceived usefulness of dedicated module bots. Initial data from N = 124 students indicate a highly AI-familiar student population: 98% reported using chatbots for learning in general, 95% were aware of the dedicated module bot, and among those answering the satisfaction item, 86% reported being satisfied or very satisfied. Reported use cases centered on clarifying domain-specific questions and included checking solutions, generating summaries or additional exercises, self-quizzing, and support for course-related tasks. Students also expressed interest in closer alignment with course materials, broader source integration, and progress-related features.

The survey did not directly measure self-regulated learning (SRL) through established SRL scales or behavioral indicators. As a result, SRL-related conclusions remain necessarily cautious. Rather than claiming demonstrated effects on SRL, the paper interprets students’ reported use cases through an SRL-informed lens, mapping them onto activities such as help-seeking, understanding checks, feedback-related support, and practice-oriented learning. This allows the paper to connect descriptive usage patterns with a theory-informed discussion of how chatbot design might better support learning processes.

Building on these findings, the paper derives preliminary design implications for the next generation of module-specific educational chatbots. The results suggest that such tools should move beyond content tailoring alone toward features that more deliberately scaffold SRL phases, including planning prompts, structured self-assessment, adaptive practice generation, and progress-aware follow-up. A follow-up survey using more direct measures of learning processes and support strategies is planned to examine these assumptions more systematically and strengthen the empirical basis for assessing the chatbot’s role in scaffolding SRL.

Keywords:

Module-specific educational chatbots, Self-regulated learning (SRL), Generative AI in higher education, Student adoption and perceptions, Learning design perspective for AI-supported learning.