Leveraging generative AI in pre-service teacher training : insights into assessment as learning

Abstract

Generative Artificial Intelligence (GenAI) tools are increasingly integrated into education, offering opportunities to transform teaching and assessment practices. This study explores the use of ChatGPT by pre-service educators in a tertiary education unit to create instructional materials, critically evaluate these materials through reflection, and identify areas for improvement. Adopting an Assessment as Learning (AaL) approach, the study investigates how engaging with GenAI fosters critical thinking, reflective practices, and the refinement of teaching strategies. Data were collected from 63 pre-service educators, who used ChatGPT to generate lesson plans and associated resources, followed by a thematic analysis of their reflective evaluations. Findings reveal that while AI-generated materials provide valuable initial structures and diverse resources, limitations in linguistic accuracy, differentiation for diverse learners, and time management, require significant human intervention. Despite these challenges, the process encouraged educators to engage deeply with pedagogical principles, enhancing their professional judgment and adaptability. The study underscores the potential of GenAI to support teacher education by fostering reflective practices and higher-order thinking skills. However, it also highlights the need for critical evaluation and customisation of AI-generated outputs. Future research should examine the long-term impact of AI tools in teacher education and their applicability across diverse educational contexts. By addressing both opportunities and limitations, this study contributes to the evolving discourse on integrating GenAI into teaching and assessment.