Abstract
This study explores the impact of interaction types of generative AI chatbots on learning achievement and user experience. It focuses on Mixed-Initiative Interaction (MII), which is similar to a creative, bidirectional conversation between humans, and Human Initiative Interaction (HII), which supports unidirectional conversations based on prompts from human users. The study conducted an online experiment that participants engaged with chatbots exhibiting these interaction types in a learning context. Results indicate that chatbots supporting Mixed-initiative Interaction enhance learning achievement and increase engagement and motivation. However, they may also increase cognitive load for users. Additionally, acceptance of Mixed-initiative interaction varied according to users' previous experience with ChatGPT. This research highlights the educational value and potential limitations of Mixed-initiative interaction, providing significant insights into the interaction design of generative AI chatbots.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
