Epistemic Network Analysis of Pre-Service Teachers’ Generative AI interactions for Lesson Design: The Role of AI Competency and Digital Literacy

Abstract

Artificial Intelligence, particularly Generative AI is increasingly used by teachers for instructional design, making the prompt design an essential competency. Effective prompt design requires competencies beyond digital literacy, including AIrelated knowledge and competencies. However, existing assessments rely mostly on self-reported skills, which may fail to capture how these competencies are being reflected in lesson design tasks. This study applied Epistemic Network Analysis to examine how Pre-service teachers’ prompt design competencies differ across levels of AI competency and Digital literacy. 43 participants completed an activity to design prompts to explain the concept of momentum to Grade 9 students. Based on median split approach on AI competency and Digital literacy, participants were grouped as High-High, High-Low, Low-High and Low-Low. Prompts were coded using the COSTAR rubric, across six dimensions - Context, Objective, Style, Tone, Audience, and Response. ENA modeled co-occurrence relationships among these elements across groups. Results showed different epistemic structures across competency levels, with high digital literacy groups demonstrating strong centralization around the objective, while high AI competency groups showing more centralization towards Audience and Tone dimensions. These results shows how AI competency and digital literacy shapes reasoning for prompt design, emphasizing the need for training the teachers to develop AI competency.