Designing a mobile chatbot-based learning journaling system for intrinsic motivation and engagement

Abstract Journaling enables students to reflect on their learning processes and thereby strengthen their self-regulation, a key competency for meeting academic goals. Previous work has shown that students benefit from digital support when creating learning journals, such as through mobile learning journaling systems. Yet, a major issue with such systems is the rapid decline in user motivation and engagement, often occurring after only a brief period of use. To address this challenge, we present a mobile chatbot-based learning journaling system that guides students through structured reflective journal entries and supports writing through an LLM-based journaling assistant. The novel system combines (1) an example-based built-in course that teaches reflective journaling through modeled responses with (2) an interactive journaling assistant that scaffolds students’ entries by generating follow-up questions and rewrite suggestions. In a randomized field experiment with 179 students using the system over 22 days, we examined the impact of both design principles on intrinsic motivation and behavioral engagement. While our results indicate that the built-in course can increase intrinsic motivation, we find no evidence that the LLM-based journaling assistant improves intrinsic motivation. Regarding engagement, both design principles provide benefits in different ways: the course shows a rather constant positive influence, whereas the LLM-based assistant appears to form a feedback loop with continued use, increasing engagement over time.