Home // DIGITAL 2023, Advances on Societal Digital Transformation // View article
Authors:
Sibylle Kunz
Claudia Hess
Keywords: ChatGPT; learners’ task; computer science lessons; morphological analysis; prompt engineering.
Abstract:
Challenging learners at school with interesting activities that foster competencies that are later required at university is an important part of a teacher’s job. But creating these tasks is time-consuming, and often these activities are traditionally more adapted to the interests of male than female learners, especially in Science, Technology, Engineering and Mathematics (STEM) courses. This, among other factors, contributes to the phenomenon of “the leaky STEM-pipeline” – losing the girls over the course of time before they chose their university career or vocational education. This paper uses an iterative approach to describe how a Large Language Model (LLM), such as ChatGPT can be used to generate interesting and female-oriented learner tasks based on a formerly developed framework of categories in a morphological analysis that helps to design computer science exercises in a structured and consistent way. The LLM is provided with the relevant features of the framework and iteratively asked to generate exercises with respect to girls’ interests and corresponding grading criteria, wrapping the task in a motivating story. The results show how a generative artificial intelligence can be used as a productive approach to the creation of teaching material targeted at a specific audience.
Pages: 13 to 19
Copyright: Copyright (c) IARIA, 2023
Publication date: September 25, 2023
Published in: conference
ISBN: 978-1-68558-115-2
Location: Porto, Portugal
Dates: from September 25, 2023 to September 29, 2023