Less Usage Means More Effective? The Design & Usage Pattern of An AI Chatbot That Supports Work-Integrated Education for Undergraduates

Abstract

Getting students well-prepared for the workplace is an important goal of many universities. This study investigated the design and student usage pattern of a conversational agent to support undergraduates to complete the subject of Work-Integrated Education (WIE). At the author’s university, the WIE subject requires 300 hours of working experience, plus submitting employment evidence, and a reflective learning statement. Despite having conducted seminars and providing Frequently Asked Questions (FAQs) on the university website, the career advisors are still receiving repetitive questions found in the FAQs. To provide timelier and out-of-class support to students, the investigators built WIEBot, to answer student questions via the Telegram, a social medium. As this is an innovative approach, we decided to conduct research with two objectives. The first objective is to verify the feasibility of building WIEBot using cloud technology and social media. The second objective is to discover the usage patterns of WIEBot. Our findings show that WIEBot met the design goal of providing students with the required information in fewer than 10 conversations in most cases. Based on this experience, we provide some suggestions for further development of similar chatbots for student support. This study contributes to raise the awareness of and provide information on using and developing chatbots in other educational domains. These domains may include direct assistance in learning a subject, or supporting students to perform out-of-classroom activities such as voluntary work. We also suggest to use quantitative models such as the Technology Acceptance Model (TAM) for further research.