ABSTRACT VIEW
TOWARDS A MORE EFFICIENT IMPLEMENTATION OF TEACH-BACK IN THE TEACHING OF ARTIFICIAL INTELLIGENCE
G. Kovács, S.S. Al-Azzawi, H. Mokayed, M. Liwicki
Luleå University of Technology (SWEDEN)
Following its initial use in health care, where teach-back was applied to probe patient understanding of their treatment or medication management, teach-back also found its way to higher education. Here, the meaning of teach-back was extended to also include the case where students are teaching back to their classmates the concepts and topics covered during the course. The resulting method was first applied in the humanities, for self-assessment and to enhance student engagement. In our earlier work, we implemented teach-back in STEM (Science, Engineering and Mathematics), as part of an introductory course about Artificial Intelligence (AI), and coupled its use with group work. After this initial implementation, the results were encouraging both in terms of passing rates and student feedback. Based on this experience, we extended the teach-back method by making it mandatory for one session, and by providing students with more scaffolding in the form of detailed suggestions about how they can engage in the process. Furthermore, we have refined the questionnaire for evaluating student opinion about teach-back and collected more qualitative data about how students perceive its implementation. Lastly, we also applied the method in a self-paced, distance learning version of the course with a larger student cohort.

In this paper, we will describe how we developed and implemented the teach-back method in our courses and evaluate the outcomes of this practice. This evaluation includes examining the summary reports written by students about their learning process, the results of oral examinations, and the results of the course evaluation questionnaire. Course development work at our university includes an evaluation questionnaire that course-responsibles are required to share with the students after each iteration of their courses. Here, we added to the standard evaluation form eight questions specific to the implementation of teach-back. With these questions, we intended to measure student opinions about our method and implementation in terms of student engagement, understanding, retention, confidence, and their opinion on different options for engaging in teach-back. The course evaluation and course outcomes show teach-back having a positive impact on students, with most students agreeing to various degrees that teach-back is an engaging activity which increased their understanding and retention of the material. Moreover, most students who filled out the questionnaire self-reported that they continued to participate in teach-back beyond the one mandatory session. Based on our experience, however, we also found possible improvements to be implemented, which we will outline in the closing section of our paper.

Keywords: Student Support and Motivation, Collaborative and Problem-based Learning, Active and Experiential Learning, Student engagement, Teaching AI, Peer Discussion.