COGNITI, THE SOCRATIC AI TUTOR, AND ITS IMPACT ON STUDENT LEARNING IN A CONSERVATORY MUSICOLOGY COURSE

Abstract

The growing prevalence of artificial intelligence (AI) in music education, which has been employed in various domains such as music performance, music composition, music education, and music recommendation, has also raised some questions about academic integrity. In this pilot study, a Socratic AI Tutor, Cogniti, was created to assist students in a first-year undergraduate degree at the Sydney Conservatorium of Music in a unit of study called “This is Music”. Cogniti provided students with critical thinking prompts without any course-related answers. The aim of this study was to explore what happens during the conversation between students and Cogniti, and to discern how the AI tutor facilitated students’ deep learning. A mixed research method combining metadiscourse analysis and thematic analysis was used to analyse conversation transcripts and Cogniti over the five-week study period. Students also provided feedback through a questionnaire. The findings suggest that Cogniti fits well with the Socratic Method in tutoring students' learning. The results of the metadiscourse analysis show the characteristics and differences in the use of metadiscourse markers between Cogniti and the students, and the results reflect the ability to facilitate student learning in terms of language functions. Cogniti also can provide effective guidance to students that is aligned to teaching objectives and course content, and the communication process supports the development of critical thinking and promotes deep learning. Problems included a little misuse of the tool as well as mixed student engagement overall. This study highlights the potential of Cogniti to enhance music education and explores a case for the appropriate use of generative AI tutors in undergraduate music courses.