PISA 2025 Warning: Falling Family Support, Falling Student Scores

A new global assessment links weaker family engagement, digital distraction and declining academic performance, while offering families a practical path forward.

PISA 2025 delivered a warning that families, schools and policymakers cannot afford to ignore. Reading and math performance across developed economies have fallen to the lowest levels ever recorded by the Programme for International Student Assessment, while students in almost every participating country and economy also reported less family support for their education. The findings do not prove that declining parental engagement caused the academic slide, but they show a powerful connection between family interest, student learning and school success.

The Organisation for Economic Co-operation and Development released the results Sept. 8 after more than 760,000 15-year-olds in 91 countries and economies took part in the 2025 assessment. Together, those students represented about 33 million teenagers worldwide.

PISA 2025 shows a troubling decline

The overall picture is sobering.

Average reading scores across OECD countries fell 28 points between 2015 and 2025. The OECD estimates that decline represents about a year and a half of learning. Mathematics scores dropped 22 points during the same period, equal to slightly more than one year of learning.

Science deserves a little more explanation. Science scores remained stable between 2022 and 2025. However, science performance had already been declining for years, and the 2025 OECD average remained at the lowest level recorded in PISA history.

Across OECD countries, about one in five 15-year-olds is now considered a low performer in science, mathematics and reading. In 2022, the figure was closer to one in six.

OECD Secretary-General Mathias Cormann said the results show that reversing the decline has become urgent. He called for education systems to invest in teachers, engage parents and provide targeted help to students and schools that need it most.

Family support for education is falling

One of the report’s most striking findings has little to do with curriculum or school funding.

It involves what happens at home.

Students reported less family support in 2025 than in 2022 in almost every participating country and economy. The decline was especially pronounced among boys and students from disadvantaged backgrounds.

In 2022, 77% of students across OECD countries said their parents or guardians asked them at least once a week what they had done at school. By 2025, that figure had fallen to 71%. Only 46% of students said they regularly discussed problems they were having at school with a parent or guardian.

Andreas Schleicher, the OECD’s director for education and skills, described the changing relationship between some parents and schools this way:

“Parents see themselves more as a client of the school, not as someone who takes personal interest.”

That observation was reported by The Associated Press as part of its coverage of the new PISA results.

The academic difference associated with family interest was substantial.

Students who said a family member asked at least weekly about what they did in school scored about 35 points higher on the PISA science assessment than students who reported less family interest. The comparison included adjustments for socioeconomic differences. The AP characterized that gap as more than a year and a half of learning.

Association does not prove causation

That 35-point difference should not be interpreted to mean that simply asking one question each week will automatically raise a child’s test score.

PISA is primarily observational. Many factors can influence achievement, including income, school quality, attendance, teacher effectiveness, health, neighborhood conditions and access to educational resources.

The OECD acknowledges that the relationship between family support and science performance becomes smaller after researchers account for students’ socioeconomic backgrounds. Still, a positive association remains across many education systems.

What kind of parental involvement helps students?

The broader research record supports the idea that families matter, while also showing that not every form of parental involvement is equally useful.

A 2022 review brought together 23 previous meta-analyses covering 1,177 primary studies conducted over roughly 50 years. Researchers found an overall positive association between parental involvement and academic achievement. The strongest benefits were connected with parents’ academic expectations and aspirations. Direct homework help produced mixed results.

Another major meta-analysis found that strong parental expectations, communication about school activities and helping children develop reading habits were among the forms of involvement most closely associated with achievement.

For families, that suggests several practical steps:

1. Ask about learning, not only grades. “What did you learn today?” can open a better conversation than “What grade did you get?”

2. Encourage regular reading. Books, newspapers, magazines and other thoughtful material all help build comprehension.

3. Set expectations for attendance and effort. Children notice when adults treat education as important.

4. Stay in contact with teachers. Families should not have to wait for a crisis before communicating with school.

5. Create technology boundaries. Homework, reading and family conversations benefit from periods without constant phone notifications.

6. Help children build routines instead of doing their work for them.

The Education Endowment Foundation, which reviews research on education interventions, currently rates parental engagement as a moderate-impact, very-low-cost strategy backed by extensive evidence.

Digital distraction is another warning sign

Family engagement is only part of the story.

PISA 2025 also raises questions about phones, recreational screen time and the rapidly expanding use of artificial intelligence.

Across OECD countries, 28% of students said classmates were distracted by digital devices during most or every science lesson.

The numbers were dramatically lower in several high-performing East Asian education systems.

In Japan and Korea, about 7% of students reported that peers were regularly distracted by digital devices in science class. The Chinese jurisdictions tested by PISA also reported rates below 10%.

The OECD found that leisure use of digital devices at school for more than one hour per day was consistently associated with weaker science performance.

Schools with cellphone bans or subject-specific rules for digital devices tended to report less classroom distraction.

But there is another important caution.

The OECD did not find clear evidence that cellphone bans by themselves automatically produce higher science scores. Students attending schools with such policies often posted similar or even lower scores in some countries.

The lesson appears to be less about simply banning technology and more about controlling how and when it is used.

Artificial intelligence brings opportunity and risk

AI has already become part of student life.

Across OECD countries, 46% of students reported using AI chatbots to help with learning at least once a week. In Korea, the figure was 51%.

PISA found that students who use AI heavily often perform worse than students who use it less. But students who regularly use AI while also being taught AI literacy tend to perform somewhat better than frequent AI users who do not receive that preparation.

That difference should matter to parents and educators.

AI can help students brainstorm ideas, practice concepts, organize information and receive feedback.

It can also become a shortcut that replaces reading, writing, calculation and critical thinking.

The OECD’s message is not that schools should reject technology. It is that technology should support learning rather than replace effort and understanding.

East Asian education systems remain among the leaders

The highest-performing systems in PISA 2025 remain heavily concentrated in East Asia.

The OECD identified the Chinese jurisdictions of Beijing, Shanghai, Jiangsu and Zhejiang, commonly reported as B-S-J-Z (China), along with Singapore, as the highest performers across science, reading and mathematics. Japan, Korea, Chinese Taipei, Hong Kong and Macao were also among the strongest performers in major subject areas.

That performance has renewed questions about America’s ability to remain competitive in science and technology.

Tracey Burns, chief of global strategy and research at the National Center on Education and the Economy, told the AP:

“To win a technological race, you need the people with the science and the technology and the reading capabilities.”

Still, comparisons involving China require important context.

PISA did not test a representative sample of every student across mainland China. The primary mainland Chinese result comes from Beijing, Shanghai, Jiangsu and Zhejiang. Hong Kong and Macao are reported separately.

What the U.S. PISA results actually show

The American picture is more complicated than a simple international ranking suggests.

U.S. students scored 463 in mathematics, exactly matching the OECD average. American students averaged 490 in reading, compared with an OECD average of 461. The United States also scored above the OECD average in science.

Reading performance fell 14 points from 2022 to 2025.

The math score changed by only two points, which the OECD said was not statistically significant. Science also remained broadly stable.

Even so, U.S. reading and mathematics performance are now among the lowest levels the country has recorded since it began participating in PISA.

About:

• 65% of U.S. students reached at least Level 2 proficiency in mathematics.

• 74% reached that level in reading.

• 78% reached it in science.

U.S. results come with a major caution

This is one part of the story that deserves more attention.

The OECD said U.S. student response rates were below the minimum target. Student exclusions exceeded PISA tolerance levels, and school participation also fell short.

More than 40% of participating U.S. students were not administered the student questionnaire.

Because of those problems, the OECD warned that bias in the U.S. performance estimates cannot be ruled out and said particular caution is needed when interpreting American results.

It also declined to include U.S.-specific questionnaire findings in its country note.

That means the international finding that family support declined should not be presented as proof that American parents specifically became less involved. The U.S. questionnaire data are too incomplete for that conclusion.

Test scores are not the whole story

There are reasonable arguments against putting too much weight on international standardized testing.

PISA does not measure every quality that matters.

It cannot fully capture creativity, character, artistic ability, teamwork, leadership or every form of intelligence. Different cultures and school systems also approach learning differently.

And as the U.S. sampling problems demonstrate, rankings must be read carefully.

But dismissing the results completely would be a mistake.

PISA is designed to measure whether teenagers can apply reading, mathematics and science knowledge to real-world situations. Those skills affect college, careers, workplace productivity and a person’s ability to evaluate information.

That last skill may become even more valuable in the age of AI.

A warning that families can act on

PISA 2025 is troubling, but it is not a verdict on today’s children.

Scores can recover.

Schools can improve.

Family engagement can grow.

Teachers and parents can set healthier boundaries around technology while showing young people how to use AI as a learning tool rather than a substitute for thinking.

Schools also have a responsibility to make family involvement easier. Communication should be welcoming, understandable and accessible to working parents, grandparents, guardians and families whose first language is not English.

Education cannot become a service that parents simply purchase and teachers alone are expected to deliver.

It works best as a partnership.

For families across New York and the Mohawk Valley, one meaningful step costs nothing and can begin tonight.

Put the phone down for a few minutes.

Talk to your child.

And ask one simple question:

What did you learn today?

Meta Title / H1: PISA 2025 Warning: Falling Family Support, Falling Student Scores

Meta Description: PISA 2025 finds record-low scores as family support declines, raising concerns about reading, screens, AI and student learning.

Permalink: /pisa-2025-family-support-student-scores/