Enhancing Efficiency in MCQ Evaluation: Integrating Voice-to-Text and Excel Automation in Medical Education

Background: Multiple-choice questions (MCQs) are widely used in medical education due to their efficiency in assessing a broad range of knowledge. However, traditional grading of MCQs can be labor-intensive and prone to human error, especially in large volumes. With increasing class sizes, there is a growing need for more efficient and accurate grading systems.

Objective: This study evaluates the effectiveness of integrating voice-to-text technology (VTT) and Excel automation to enhance the efficiency and accuracy of MCQ grading in medical education.

Methods: A total of 3,000 simulated MCQs responses were evaluated using both manual and automated methods. VTT technology was used to transcribe responses, which were then organized into a table using a Large Language Model (LLM) for data parsing and structuring and scored in Excel using automation techniques. Two evaluators assessed the papers, and time taken, error rates, evaluator fatigue, and satisfaction were compared between methods.

Results: The automated method significantly reduced evaluation time to 45.9 minutes (0.76 man-hours) compared to 194.82 minutes (3.25 man-hours) for manual evaluation, with a similar high accuracy rate of 99.96% for both methods. Evaluator fatigue was lower, and satisfaction was higher with the automated method.

Conclusion: The integration of VTT technology and Excel automation significantly improves the efficiency of MCQ grading while maintaining high accuracy. This approach offers a scalable, cost-effective solution for medical education settings, particularly in resource-limited environments. Future research could expand this method to other types of assessments and explore additional automation tools to further enhance educational evaluation processes.