| dc.description.abstract | Nowadays, people live in an international global village. English becomes a way to communicate with each other, and English reading ability is also a necessary ability. Past research has mentioned that in a study group with study peers, teachers guide and provide a space for a free discussion with a dialogic reading method. Students can discuss storylines with each other and express their own opinions. Students better understand the story content, which has a positive impact on English reading. Therefore, despite the lack of educational resources, we provide care for each student in a limited time. This study develops a chatbot that acts as a facilitator in a two-person book chat group, invites students to share stories, and facilitates students to interact with each other. Compared with the guider chatbot, this study discusses the influence of different chatbots on chatting engagement, chatbot feelings, and English reading interests.
The subjects of this study were 55 fourth grades students from an elementary school in Taiwan. The experimental group（n=30）used a facilitator chatbot, and the control group（n=25）used a guider chatbot. All students participated in an English reading activity once a week. Students freely selected books to read quietly in their seats. Students conducted a book talk activity with chatbots in the previous semester and the next semester. The students in the experimental group worked in pairs to coordinate a chatbot to discuss the content of the same storybook. The students in the control group chatted with the guider chatbot one-on-one. This study analyzes student dialogue, flow experience, chatbot perception, and situational interest in English reading.
The results indicated the students in the experimental group mostly responded to open-ended questions and focused on cognitive participation. In dialogues, students went deep into topics and imitated and agreed with their peers. From the perspective of the study group, the one-way response type was the most common. Students are highly engaged in chatbot chatting activities. Due to the novelty effect, students′ curiosity dropped after chatting with the chatbot for the first time. It also causes student interest in English reading to rise and fall for a short time. From the perspective of prior knowledge, low-achievers interest in English reading increased after using the facilitator chatbot. Compared with the control group, there is no significant difference between the chatting engagement and the chatbot feelings. In the English reading situational interest, the experimental group arouses the Maintained-SI-Value of English reading in the middle stage compared with the control group. There is no significant difference in the later stage. In the dialogue content, we observed that the students in the experimental group more actively greeted, asked questions, asked to respond, and corrected interactive behaviors. In addition, students had hidden relationships with others who went deep into topics, imitated, and agreed with others. Nevertheless, the control group interacted less with others. | en_US |