The Efficacy of Generative Artificial Intelligence in Developing Science Education Preservice Teachers’ Writing Skills: An Experimental Approach

Abstract

Given the continued debates sparked by recent technology innovation around generative artificial intelligence and the varied findings on its use for productive learning, this paper addresses the efficacy of ChatGPT from a preservice teachers’ perspective and science education viewpoint. This study adopted a mixed-method research design with an experimental approach for methodology. For data collection, post-tests and interviews were used to elicit information from the participants on the efficacy of ChatGPT in improving their writing skills and their opinions on its efficacy for learning. The paper revealed that ChatGPT might not significantly improve preservice teachers' writing skills. While this may be contextual, our findings provide valuable insights into pedagogical strategies for resisting techno-deterministic framings of present-day learning. This study contributes to the corpus of studies in the scientific literature on the efficacy of ChatGPT in improving student achievement. Besides, it unpacks pedagogical strategies that may be applied to improving students' writing skills.