Abstract:

A dyadic mentoring process in the form of one-to-one or peer mentoring has been proven to successfully guide higher education students in multiple disciplines. However, the evolution of learning technologies has made it possible to incorporate conversational systems like mentoring bots into such mentoring scenarios. This type of bot can make the one-to-one process scalable by reducing the human mentor’s effort. The research objective is to analyze the possibility to implement a chatbot scaling mentoring recommendation subtask accompanied by students’ acceptance. Therefore, in this study, a mentoring bot has been implemented as an innovative approach to the mentoring process which supports to review students’ own tasks by using the method of self-reflection.
The mentoring bot can be defined as a chatbot that is integrated into the learning management system for a research seminar in computer engineering. In scientific research, reviewing the process and structure of the research in result presentation and reporting helps to improve the quality of the research. Though the students learn in the seminar about the structural process from the lecturer, reviewing all their work before their final performance requires hours which adds a huge load to the mentor. With the purpose of providing the scope of reviewing the assignment to all the students, the mentoring bot is implemented. In contrast to the mentoring process of reviewing to guide and initiate self-reflection among the students, it helps to ensure the accuracy and quality of the research along with identifying knowledge gaps in their application task of the scientific research methodology. Addressing Vygotsky’s scaffolding theory, the conceptual base is developed for this mentoring bot. According to this theory, the bot gives guidance as an expert on the content that the students cannot learn independently. However, an effort has been made to make the students self-regulated learners, who review their own content through self-reflection by combining the scaffolding technique with metacognitive questioning in this mentoring strategy. The students are expected to interact with the chatbot when they feel the need to review their work and, in the process, they would learn to metacognitively reflect on their performed task.
The bot is first integrated in the summer semester of 2021 and a small qualitative evaluation is done to see its total usage among the students. In the following semesters of winter semester 21/22 and summer 2022, an increase has been observed in the total usage from the survey responded to by the students. The findings of the survey show that the students feel supported in their self-regulated learning. 50.5% of the students have reported receiving support for content knowledge understanding. Additionally, 40% of the students are overall satisfied with using the mentoring bot. These results show students’ acceptance and support to reduce the mentor’s load by scaling the review process with the mentoring bot for all the students of the seminar. The implication of mentoring bots for supporting the students metacognitively in the review process for learning scientific research methodology has been presented as an innovative practice in this study along with delimiting the scope of improvements in the process of scalable mentoring.

Keywords:

Mentoring, Chatbot, Scalability, Scientific Research, Review.