Theoretical background of the game design element "chatbot" in serious games for medical education.

Abstract

Background: The use of virtual patients enables learning medical history taking in a safe environment without endangering patients' safety. The use of a chatbot embedded in serious games provides one way to interact with virtual patients. In this sense, the chatbot can be understood as a game design element, whose implementation should be theory driven and evidence based. Since not all game design elements are already connected to theories, this study aimed to evaluate whether the game design element chatbot addresses the need for autonomy rooted in the self-determination theory.

Method: A cross-sectional study was conducted to compare two distinct chat systems integrated in serious games with one system being an open chatbot and the other system being a constrained chat system. Two randomized groups of medical students at a German medical school played one of two serious games each representing an emergency ward. The data collected included both objective data in terms of students' question entries and subjective data on perceived autonomy.

Results: Students using the open chatbot generally asked significantly more questions and diagnosed significantly more patient cases correctly compared to students using a constrained chat system. However, they also asked more questions not directly related to the specific patient case. Subjective autonomy did not significantly differ between both chat systems.

Conclusion: The results suggest that an open chatbot encourages students' free exploration. Increased exploration aligns with the need for autonomy, as students experience freedom of choice during the activity in terms of posing their own questions. Nevertheless, the students did not necessarily interpret the opportunity to explore freely as autonomy since their subjectively experienced autonomy did not differ between both systems.