Programme

P.98 -- Posters, VCTP-50

Date: Thursday, 7 August 2025

Time: 08:30 - 10:00

Statistical Analysis of Score Distribution Transitions in Non-Centralized Assessments: A Case Study of Final Exam

Chu Thuy Anh(1), Tran Cong Phong(2)

1 Institute of Physics, 2 Antomic Molecular and Optical Research Group, Institute for Advanced Study in Technology, Ton Duc Thang University

Data from the Hanoi Department of Education and Training reveal concerning trends in the results of the 12th-grade student quality assessment conducted in March 2025. Notably, 6.78% of students scored below 3 out of 10, with 0.91% receiving critically low scores of 1 or less—placing them at significant risk of being ineligible for graduation without substantial improvement in the official final exam. Furthermore, 32% of students scored below average (under 5 points), highlighting widespread academic underperformance. Among all subjects, Mathematics recorded the highest rate of below-average scores (51.69%), followed closely by Geography (51.42%) and Biology (50.41%). These results contrast sharply with data from the 2020 national high school exam, where Mathematics and Physics had the highest mean scores, while English had the lowest ones. Interestingly, Biology scores exhibit a near-normal distribution, suggesting underlying stability in assessment or instruction in that subject. This study applies statistical modeling techniques to investigate the transitions in score distributions across subjects, aiming to uncover the mechanisms behind these shifts. By analyzing structural and systemic factors influencing score variance, we propose an adjusted model to support fairer and more effective assessment design for future national examinations.

Presenter: Chu Thuy Anh