Evolving Student Assessment: AI-Driven Rubrics for Personalized and Equitable English Language Learning

Abstract

Technological advancements have been reshaping the paradigms of educational scenarios through tools and platforms that enhance the academic experience. The demands of Industry 4.0 and the implications of Moore's Law accelerated these changes paving the way for imbibing innovative approaches in education. This study delves into developing and applying AI-driven rubrics to foster personalized and equitable assessments in English language learning for engineering undergraduates. By leveraging AI tools, the tailored rubrics provide personalized feedback, minimize subjectivity, and improve grading consistency, in alignment with Outcome-Based Education (OBE). This paper studies the impact of AI-generated rubrics on student performance, engagement, and perception, recognizing the advantages and challenges of this approach. The results of the mixed-method study involving 64 first-year BTech students reveal that AI-driven rubrics are perceived as generally transparent, fair, and reflective of student performance. However, students also feel that several areas of the feedback require improvement, like the standard of the feedback, transparency, and AI decision-making. The study ends with suggestions for improving AI-driven rubrics to increase AI capacity to support inclusive, transparent, and individualized evaluation procedures in English language instruction and acquisition.