Enhancing Teacher Feedback Using Ai-Powered Automated Grading Systems

Abstract

This research looks into the effects of AI-assisted grading systems on feedback in higher education in Pakistan. Using a mixed methods approach, data were collected over an academic semester from 80 faculty members and 800 undergraduate students across 12 public universities in four provinces. The sample included six universities who implemented AI grading systems and six that continued with conventional grading without the AI grading systems. Data were collected from surveys of faculty and students, classroom observations, and interviews regarding feedback quality, time efficiencies, faculty satisfaction, student performance, and engagement, and faculty assessment marking and grading. The primary findings showed that the AI grading systems improved the timeliness, consistency, and personalization of student feedback significantly at the universities using them, per the SPSS data. Faculty reported that the automated grading systems improved their teaching efficiency, reduced grading time, and improved student engagement, but they faced some technological issues and advised caution regarding purely automated grading to grade, for example, a 'subjective assignment'. The qualitative data also noted challenges in using AI grading comparison to rubric criteria. In sum, the findings indicate that there is potential for AI grading to help improve feedback and assessment practices in higher education in Pakistan, although challenges remain.