Exploring Student and AI Generated Texts: Reflections on Reflection Texts

Abstract

As pointed out by many scholars, Artificial Intelligence (AI) provides both opportunities and challenges in regard to assignments and examination in higher education. The accessibility and use of AI in regard to student assignments, examinations and assessments places demands on teachers’ work in course design and formats of assignments and examination. For teachers, this work is a constant and continuous process, in line with the Scholarship of Teaching and Learning (SoTL) according to Boyer (1991). In order to meet these new demands, teachers need to reflect upon design, as reflective practitioners (Schön, 1987). Reflective design may alleviate the challenges with AI as well as make use of the opportunities with the use of AI. In this paper there are two sets of data. This study aspires to contribute to the current state of AI (ChatGPT) as it is applied in higher education through an empirical study of authentic reflection texts by students in comparison to AI (ChatGPT) generated texts. The first set of data is authentic reflection texts (N=20) written by students. The second set of data is texts generated by AI (ChatGPT). The texts are analysed using reflective thematic analysis (Braun & Clarke, 2019). The themes in the two sets of texts are described, analysed and compared. The two sets of data are then explored, analysed and compared to highlight similarities and differences between the authentic texts and the texts generated by AI. These insights may provide support for teachers in regard to the design of assignments and examinations as well as the practical use of AI (ChatGPT) in higher education.