Enhancing the EFL Teaching Based on EPCK Model with LLMs: Insights from Lag Sequential Analysis

The generative artificial intelligence (AI) has gained widespread acceptance in education, particularly in teachers professional development as an effective pedagogical strategy. Despite its success, for English as a Foreign Language (EFL) powered learning environment, how teachers find, evaluate, and effectively use information they have not consistently translated into improved academic outcomes. This study aimed to address existing concerns and shed light on the nature of knowledge-building interactions between teachers and students, based on classroom teaching behaviors for EFL. Guided by TPACK theory, particularly based on Pedagogical Content knowledge (PCK), it involved 12 English teachers who participated in exploratory learning environment, in Southern China. It takes 6 high and 6 ordinary performance groups using advanced digital methods including Lag Sequential Analysis, for three grades in the middle school. Taking the teaching performance were collected through video engagement, and analyzed dialogic interactions between teachers and students, it used EPCK (English-PCK) model for composite English subject characteristics. The findings indicated that the using LLMs in EFL teaching positively influenced teachers academic performance. Interaction patterns revealed that teachers predominantly engaged with the LLMs by using EPCK Model on learning sessions, primarily for the purpose of asking questions. Notably, the educational dialogues suggested that the depth of collaborative knowledge construction between teachers and students at a basic level. It highlights the practical pathways and strategic methods for enhancing teachers' professional capabilities, holding significant practical implications for improving K-12 English teaching quality in the intelligent age.