Abstract:
This study investigates the efficiency of AI grading compared to traditional human
grading methods in educational assessment processes. Utilizing the Baccalytics plat
form as a case study, the research examines the impact of AI-driven grading on grading
time, accuracy, and overall workflow automation. The study employs a controlled ex
perimental design, with human graders constituting the control group and the Bacca
lytics AI test grader representing the experimental group. Results indicate a significant
improvement in grading efficiency with the AI grader, as evidenced by substantially
lower grading times and comparable accuracy to human graders. Furthermore, cor
relations between various variables, such as character count and grading time, shed
light on the factors influencing grading processes. Despite the promising findings,
limitations are acknowledged, including teacher awareness of timing and the need for
ongoing training and support. Recommendations for future research include explor
ing Baccalytics’s scalability, addressing equity gaps in access to personalized learning
experiences, and investigating its impact on student-teacher interactions and feedback
processes. The study contributes to understanding the potential of AI-driven grading
systems in enhancing educational practices and improving learning outcomes.
