Exploring EFL Teachers’ Behavioral Intentions to Integrate GenAI Applications: Insights From PLS-SEM and fsQCA

Abstract

The rapid development of generative artificial intelligence (GenAI) applications has opened new possibilities across various fields, including English language education, by enabling personalized and adaptable learning experiences. Responding to the growing trend of integrating GenAI tools into EFL instruction, this study explored Iranian teachers’ behavioral intentions to use GenAI applications, such as ChatGPT, for English teaching in higher education. Anchored in the “UTAUT” framework, the study examined external factors influencing adoption intentions, while the TPACK model assessed internal factors tied to instructors’ AI usage. A structural model featuring 20 hypotheses based on the “UTAUT” and “AI-TPACK” was proposed. Data were gathered from 444 Iranian EFL teachers via an online survey and analyzed using “partial least squares structural equation modeling and fuzzy-set qualitative comparative analysis (fsQCA).” The results highlighted the critical roles of performance expectancy and social influence in shaping adoption intentions. Interestingly, a negative relationship between AI-TPACK and behavioral intentions revealed a paradox: Deeper technological knowledge may hinder, rather than facilitate, AI adoption in teaching. Key drivers of adoption included teachers’ perceptions of GenAI’s potential to enhance instructional performance and support from social networks. Effort expectancy, however, was less significant in this context. The study also identified sociocultural and institutional challenges as crucial barriers, underscoring the need to address these for sustained AI integration. This research enriches the literature by uncovering enablers and barriers to GenAI adoption, offering valuable insights into the sociocultural and institutional dynamics influencing technology integration in diverse educational settings.