Improvised progressive model based on automatic calibration of difficulty level: A practical solution of competitive-based examination

Abstract

Online learning has grown due to the advancement of technology and flexibility. Online examinations measure students' knowledge and skills. Traditional question papers include inconsistent difficulty levels, arbitrary question allocations, and poor grading. The suggested model calibrates question paper difficulty based on student performance to improve understanding assessment. The suggested student assessment system paradigm involves determining difficulty, creating the exam, and assessing the student. Based on the previously established relationship between question difficulty and right responses, questions are computed and then divided into difficulty categories. This model improves testing by adapting to the student's ability in real-time. This method ensures that all students are graded uniformly and fairly using pre-determined questions and criteria. The methodology can also cut exam creation and administration time, freeing up teachers and administrators to focus on other assessment tasks. It considers more evidence, learner-centered assessment can help employers evaluate candidates more accurately and meaningfully. It might boost academic productivity by letting assessors quickly write high-quality papers and save up time for deeper investigation and experimentation. This may accelerate scientific progress. Automatic paper generation raises ethical questions about research validity and reliability.