Task-specific Generative Artificial Intelligence Prompting Practices: Pre-service Language Teachers’ Perceptions and Practices

There has been keen interest in the application of generative artificial intelligence (GenAI) tools in language education, as well as understanding the knowledge and skills language teachers need to use these tools effectively and responsibly in their professional practices. Scholars have speculated that to use these tools effectively, teachers need certain task-specific competence. Yet, there have not been any empirical studies to date, that have explored how language teachers interact with GenAI tools to complete specific professional tasks, such as lesson planning. Therefore, this study aimed to explore the ways language teachers interact with GenAI when engaged in lesson planning and identify the knowledge and skills needed to effectively prompt GenAI tools to get the desired outcomes. The study involved collecting lesson plans, GenAI logs and written reflections from Hong Kong pre-service language teachers who were tasked with improving a lesson plan with GenAI assistance. Data analysis revealed that the teachers engaged in a three-stage interactional decision-making process when using GenAI tools: (1) identifying the task to be assisted by GenAI; (2) engaging iterative interactions with GenAI; and (3) incorporating GenAI-generated content to the intended task. In addition, by analyzing the content of the participants’ prompts, three aspects of effective task-specific prompting practices were identified. These are (1) pedagogical content knowledge, (2) GenAI knowledge, and (3) prompting skills. The findings provide a framework that can assist teacher educators and professional development providers in supporting language teachers to develop the competence needed to effectively utilise GenAI in their professional practice.