Vietnam.vn - Nền tảng quảng bá Việt Nam

High School Graduation Exam 2025: Score distribution cannot be good without standardization

The 2025 high school graduation exam ended with many candidates and teachers breathing a sigh of relief, when the score distribution of two subjects that were once considered difficult (math and English) was commented by some experts as "surprisingly beautiful".

Báo Tuổi TrẻBáo Tuổi Trẻ19/07/2025

thi tốt nghiệp - Ảnh 1.

Candidates taking the graduation exam in Ho Chi Minh City - Photo: NHU HUNG

However, praising a beautiful spectrum based solely on the shape of the chart or intuitive feeling lacks scientific basis and can easily lead to misunderstanding.

A "beautiful" score spectrum is only valuable when it honestly reflects students' abilities according to output standards, ensures fairness among candidate groups, and is built on the basis of objective, multiple-choice standard test questions.

This year's math and English score distribution data is actually raising important warnings.

First, let's look at the specs. Maths had an average score of 4.78, with a median of 4.6 - nearly half of students scored below 5 points. Only 12% of candidates scored 7 points or higher, while nearly 8% scored below 2.5 points.

English had an average score of 5.38 and a median of 5.25, but 38% were below average. These figures are not “good” because they show a low average and skewed distribution of results – especially in maths.

According to the principle of assessment in multiple-choice testing, a score spectrum is only valid if it is accompanied by stable indicators of distribution, discrimination and reliability of the test.

The high school graduation exam is now a "two-in-one" exam - both for graduation and for university admission.

And to serve both the graduation and university admission goals, standard multiple choice tests are often designed with about 60-65% of questions at the recognition - comprehension level, the rest are at the application and high application level to classify candidates.

In test design, this is entirely possible if the question bank is standardized and the questions are pilot tested to determine difficulty and discrimination.

However, the reality shows that the current exam has not yet undergone a full standardization process. The transition to the 2018 general education program further exposed the confusion in exam design in math and English.

Due to the lack of standardization, the score distribution becomes less reliable. For math, this year's score distribution is slightly skewed to the right, with the peak falling around 4-4.5 points - meaning that most candidates are concentrated at the low score level.

In a national exam, this means that many students do not meet the minimum requirements for general mathematics proficiency.

Although the English subject has a more "balanced" score distribution, it actually reflects the results of a selected group of candidates - because this is an elective subject, only about 350,000 students took the exam, a sharp decrease compared to previous years. Nearly 40% still scored below average - a number that worries the public.

However, when the test is not standardized, this result mainly reflects the level of "taking the test" but cannot confirm the level of "learning" or the actual quality of teaching. Further analysis by region, learning conditions and school context is needed to have a valid conclusion.

Notably, while very few students achieve 9-10 points in the graduation exam, many schools convert IELTS 6.5 into 9-10 English points - creating a difference that can be unfair if not regulated transparently.

From the above data, the Ministry of Education and Training needs to fully implement the exam standardization process: building a question bank that closely follows output standards, testing questions with representative samples, analyzing difficulty - discrimination and testing overall reliability.

The exam structure must be clearly designed and the matrix and objectives of each part must be transparently announced. This helps students to reduce the pressure of taking the graduation exam, while universities still have a suitable basis for selection.

In addition, it is necessary to adjust the use of international certificates in admissions to avoid creating unfair advantages for a group of students.

Finally, score distribution evaluation should be based on statistical parameters - such as mean, median, standard deviation, score distribution, and test reliability - rather than just looking at graphs or expert opinions.

Back to topic
Le Tuan Phong

Source: https://tuoitre.vn/thi-tot-nghiep-thpt-2025-pho-diem-khong-the-dep-neu-thieu-chuan-hoa-20250719093601269.htm


Comment (0)

No data
No data
PIECES of HUE - Pieces of Hue
Magical scene on the 'upside down bowl' tea hill in Phu Tho
3 islands in the Central region are likened to Maldives, attracting tourists in the summer
Watch the sparkling Quy Nhon coastal city of Gia Lai at night
Image of terraced fields in Phu Tho, gently sloping, bright and beautiful like mirrors before the planting season
Z121 Factory is ready for the International Fireworks Final Night
Famous travel magazine praises Son Doong cave as 'the most magnificent on the planet'
Mysterious cave attracts Western tourists, likened to 'Phong Nha cave' in Thanh Hoa
Discover the poetic beauty of Vinh Hy Bay
How is the most expensive tea in Hanoi, priced at over 10 million VND/kg, processed?

Heritage

Figure

Business

No videos available

News

Political System

Local

Product