The FAQs below address common questions about Hong Kong PISA 2025, covering its purpose, student performance, assessment quality, and student engagement to support the interpretation and use of PISA findings.
Response:
PISA evolves because the knowledge and skills young people need also change. Recent cycles have placed more emphasis on evaluating information, reasoning with evidence, solving complex problems, and learning in digital environments.
Common Confusion:
Some people think PISA is “moving the goalposts” and that changes in the framework make results invalid.
Clarification:
Framework changes are part of PISA’s design, not a flaw. OECD uses technical standards, trend items, scaling procedures and quality checks to support valid comparisons over time where such comparisons are reported. However, trends should still be interpreted carefully, especially when the assessment framework has evolved.
Response:
COVID-19 may be part of the background, but PISA alone cannot prove how much of the performance change was caused by the pandemic.
Common Confusion:
A common assumption is that the decline must be mainly or entirely due to COVID-related school disruption.
Clarification:
This explanation is not sufficient on its own. Many education systems experienced COVID-related disruption, and Hong Kong’s downward trend in most domains began before the pandemic. PISA is a survey, so it can show patterns and associations, but it cannot by itself identify the exact causal share of COVID-19, digital habits, student engagement, curriculum experience or other factors.
Response:
PISA can help identify broad learning needs and guide follow-up support, but it should be used together with local school data. Support should be based on students’ proficiency levels, learning processes, learning motivation and behaviours, and school context, not on rankings.
For PISA 2025, support can focus on at least three areas:
- Strengthening foundations for students below baseline proficiency
- Stretching students towards advanced proficiency
- Developing student agency
Common Confusion:
Some people think PISA can directly identify which individual schools or students are “weak” and prescribe a ready-made solution.
Clarification:
PISA is designed for system-level and group-level analysis, not for ranking individual students or diagnosing individual schools.
Precise support should therefore combine PISA findings with school-based assessment data, classroom observations, student work, attendance and wellbeing data, teachers’ professional judgement, and feedback from students and parents.
The aim is not to find who or what to blame, nor to label schools or students. More importantly, the focus should be on identifying learning needs and providing timely support.
Response:
There is no evidence that sampling, testing period or test administration invalidated Hong Kong’s PISA 2025 results. Hong Kong’s data met OECD quality standards and were considered fit for reporting.
The OECD Hong Kong Country Note states that Hong Kong’s reported data met PISA 2025 quality standards and were fit for reporting.
Reference: OECD, PISA 2025 Results: Hong Kong (China) Country Note, p. 1 / reporting notes section.
Link: https://www.oecd.org/en/publications/pisa-2025-results-volume-i-country-notes_2d4ff9ea-en/hong-kong-china_453f39ca-en.html
Common Confusion:
Some people suggest that changes in sampling, the test administration, or the testing period, or school-level administration may explain the decline.
Clarification:
PISA follows strict OECD technical standards. Schools and students are selected through probability sampling. Schools cannot choose participating students or replace selected students with others. Test administrators follow standardised scripts and timing rules, and administration is subject to quality monitoring.
Hong Kong’s PISA 2025 assessment was conducted from May to July 2025. This timing alone cannot explain the results. Hong Kong’s PISA 2022 main study was also conducted from May to July. Local analysis of PISA 2025 testing dates found no statistically significant association between later testing dates and lower scores after controlling for available student and school variables.
More importantly, any administration matter in PISA 2025 cannot explain the longer-term decline since around 2012. The trend started well before the 2025 administration.
Response:
The use of English as an assessment language does not by itself make the results invalid or incomparable. PISA has procedures for approved assessment languages, translation, adaptation and verification.
Common Confusion:
Some people assume that students taking the assessment in English must have lowered the overall reading score, or that language choice makes cross-cycle comparison impossible.
Clarification:
The available Hong Kong data do not support the claim that English-medium assessment caused the decline. In weighted PISA 2025 data, about 55% of students took the assessment in Chinese and about 45% in English. In unadjusted comparisons, the English assessment group had higher average scores than the Chinese assessment group. After adjusting for socioeconomic background, gender and sampling strata, the differences became smaller. However, the English assessment group still obtained higher average scores than the Chinese assessment group.
Response:
The evidence does not support the claim that Hong Kong students generally did not take PISA seriously. Several indicators were positive:
- Students’ self-reported effort remained broadly stable, at around 7.3 out of 10.
- The rate of non-reached items was low.
- Most students attempted the assessment rather than simply leaving large parts blank.
- Some indicators of late-test fatigue did not worsen enough to explain the overall decline.
One negative indicator was that the share of hasty responses in reading increased from 5.0% in 2018 to 8.8% in 2025, close to the OECD average of about 9%.
Common Confusion:
Some people assume that lower scores mean students “gave up” or did not try.
Clarification:
Student engagement matters, especially in a low-stakes assessment such as PISA. However, the evidence does not show that disengagement explains Hong Kong’s overall trend.
In particular, disengagement cannot explain:
- the longer-term decline since around 2012; or
- the drastic drop of around 24 points in science and 20 points in reading between PISA 2022 and PISA 2025.
The more balanced interpretation is that some students may have been more prone to rapid and inaccurate responding, but this is only one factor to monitor. It should not be used as a single or main explanation for Hong Kong’s performance trend.
Response:
Unfamiliarity with PISA-style items may affect some students, but it is not a sufficient explanation for Hong Kong’s performance.
PISA is intentionally designed to assess whether students can apply knowledge and skills in unfamiliar and real-life situations.
The more important issue is whether students have enough future-ready competencies, including the ability to:
- apply knowledge in unfamiliar contexts;
- evaluate information from multiple sources;
- reason with evidence;
- model complex problems;
- adapt strategies when tasks are difficult;
- explain and justify conclusions.
Common Confusion:
Some people think students did not do well simply because they had not practised PISA-style questions.
Clarification:
This explanation is not convincing on its own. The most unfamiliar domain in PISA 2025 was likely Learning in the Digital World, which used innovative digital tasks involving computational problem-solving, modelling and programming. Yet this was Hong Kong’s strongest area, with a score of 545, among the top-performing systems. This shows that unfamiliar item format alone cannot explain weaker performance in other domains.
The purpose is not to train students for PISA items. Rather, schools should provide more regular opportunities for students to think deeply, solve unfamiliar problems, evaluate evidence, and transfer learning across subjects.
English 
Chinese