EFL Teachers' Perceptions of the Use of Generative AI for Speaking Assessment in Higher Education

Abstract

Speaking assessment is considered a fundamental component of English language education. However, traditional human-based speaking assessment is associated with several challenges, including subjectivity, inconsistency, and the difficulty of providing immediate and detailed feedback. In this regard, Generative Artificial Intelligence has emerged as a promising solution to address these limitations. This study investigates oral expression university teachers' perceptions of the use of Generative Artificial Intelligence for speaking assessment. Specifically, it seeks to answer the following questions: (1) What are the challenges and limitations of current human-based speaking assessment? (2) What are oral expression teachers' perceptions of the use of Generative Artificial Intelligence tools for speaking assessment in higher education? (3) What are the perceived concerns and requirements for effectively implementing Generative Artificial Intelligence in speaking assessment? To achieve the aim of the study, a quantitative questionnaire is designed and administered to 81 oral expression teachers from different Algerian universities. The collected data are analyzed using Statistical Package for the Social Sciences version 24. The findings reveal that teachers generally perceive Generative Artificial Intelligence as a viable solution to the limitations of traditional human-based speaking assessment and hold positive perceptions toward its integration. However, they emphasize that its implementation should remain under human supervision and that developing teachers' digital literacy is a necessary requirement for its effective use. The study recommends conducting experimental research to provide empirical evidence on the effectiveness of Generative Artificial Intelligence in improving speaking assessment practices.