Table of Contents
Computational Problem Solving PISA 2025: Build It at Home
Computational problem solving PISA 2025 is a new domain: US 513 vs OECD 500. What the tasks look like, the six levels, and how to practice off-screen.
Computational problem solving PISA 2025 is the one genuinely new thing in this year’s OECD results, and it is the part parents should care about most. The domain tested whether 15-year-olds can solve problems using modeling and programming tools, run systematic experiments, analyze the data they generate, and debug what they build. US students averaged 513 against an OECD average of 500, per the OECD country note. Macao (China) led at 572, with Singapore at 563 and B-S-J-Z (China) at 560.
Here is the useful part: the skill being measured is not coding. Only part of it happens on a keyboard. The rest is a habit of mind you can build with paper, dice, and a timer.
Key Takeaways
- PISA defines the domain as “an iterative and self-regulated process of knowledge building and problem solving using computational tools and practices.”
- Students got two extended units with access to tutorials, feedback, and examples, so the test rewards learning during the task, not prior knowledge.
- OECD average is 500. US 513. Top systems: Macao (China) 572, Singapore 563, B-S-J-Z (China) 560, Japan 557, Chinese Taipei 551.
- About 26% of OECD students reach Level 5 or 6. Level 3, where students “combine a wider range of commands to build functional programs,” is the practical benchmark.
- Self-regulated learning is baked into the construct: whether a student can stay engaged “even when it becomes challenging or frustrating.”
What computational problem solving PISA 2025 actually tested
Computational problem solving is the assessment of whether a student can use computational tools and practices, such as modeling or block-based programming, to build something that works and then fix it when it does not. The OECD describes performance on the scale as the ability to solve problems “using computational tools, such as modelling or programming tools, and through computational practices, such as conducting experiments, analysing data, and building and debugging computational artefacts.”
The test design is the interesting choice. From Volume I, released September 8, 2026: “Students complete two extended test units, during which they solve increasingly complex tasks and have access to learning resources such as tutorials, feedback and examples.” That means a student who has never used the tool before can still do well, if they read the tutorial, try something, look at the result, and adjust.
The OECD is explicit that this makes self-regulation part of the measurement: “A student’s ability to succeed in the digital world is greatly influenced by their self-regulated learning competencies, i.e. how they manage their own learning, and, crucially, whether they can maintain their motivation and stay engaged with a task, even when it becomes challenging or frustrating.”
This domain sits inside a larger PISA innovation called Learning in the Digital World. The full analysis of how students self-regulated during those tasks comes in Volume III, Effective Learners in the Digital World, which the OECD lists as forthcoming.
The six levels, translated for parents
The OECD’s proficiency descriptors read like a developmental map. I have translated each level into what it looks like at a kitchen table.
| Level | What PISA says students can do | What it looks like at home |
|---|---|---|
| Below 1 | Explore the software interface “in a rudimentary way, rather than working towards a specific goal” | Clicking around with no plan |
| 1 | Minor modifications to simple programs; adjust a single input in a basic model | Changing one number to see what happens |
| 2 | Simple operations such as instructing a program to repeat an action; begin testing ideas | ”Do this five times” and checking the result |
| 3 | Combine a wider range of commands to build functional programs; run basic controlled experiments | Building something that works; changing one variable at a time |
| 4 | Moderately complex programs and models; exploit repeated patterns, with minor bugs | Noticing “I wrote the same thing four times, I should loop it” |
| 5 | Relatively sophisticated programming and modeling | Planning the structure before building |
| 6 | Optimal, problem-free programs without unnecessary redundancy | Finding the simplest version that still works |
The distribution across OECD countries: 26% at Level 5 or 6, 21% at Level 4, 17% at Level 3, 13% at Level 2, 17% at Level 1, and 5% below Level 1. The OECD notes there is no formal baseline threshold on this scale, “although Level 3 can be regarded as a useful benchmark, as it marks the point at which students begin to demonstrate sufficiently solid knowledge and skills to progress more rapidly.”
One more result worth knowing: some systems did much better here than their science score predicted. Hong Kong (China), Malaysia, Brunei Darussalam, the Philippines, Macao (China), and Iceland all performed “over 30 points above expectations based on their science performance.” Malaysia is the standout, scoring 419 in science but 484 in computational problem solving.
How to Teach Your Kid About Computational Problem Solving
The PISA 2025 framework names four practices: systematic experimentation, data analysis, building, and debugging. None of them requires a computer. Each activity below targets one directly.
Ages 5–8: The Robot Game
You are the robot. Your child gives you instructions to make a peanut butter sandwich, and you follow them with malicious literalism. “Put the peanut butter on the bread” means you place the closed jar on the loaf. They will laugh, then get frustrated, then start being precise. That frustration is the lesson: a computer does what you said, not what you meant.
Add the loop concept once the sandwich works: “Can you say that in fewer instructions?” If they say “spread it four times,” you have just taught a repeat command, which is PISA Level 2.
Ages 9–12: One Variable at a Time
Build a paper airplane, then ask one question: does adding a paperclip make it fly farther? The rule is that only the paperclip changes. Same plane design, same throwing spot, same person throwing. Five throws each way, write down every distance, compare the averages.
This is exactly what PISA calls conducting “basic, controlled experiments in order to infer simple relationships between different factors in a model,” which sits at Level 3. The written record matters more than the result. Kids who keep data learn to trust data.
Ages 13+: Debug Something You Did Not Write
Give your teen a working system with one deliberate fault. Easiest version: write a set of instructions for a task they know (a recipe, a route, a game setup), insert one error, and hand it over cold. Their job is to find the fault by testing, not by rereading.
The step most kids skip is forming a hypothesis before testing. Ask them to say out loud what they think is broken and what test would prove it. That is the difference between Level 3 and Level 5 on the PISA scale: sophistication comes from planning, not from typing faster.
If they want the on-screen version, Scratch, Tinkercad, and spreadsheet modeling all exercise the same loop. Our comparison of computational thinking versus coding for kids covers what each actually builds.
The question to ask: “What would you change first, and how would you know if it worked?”
Why the tutorial design matters more than the score
In the United States, where the National Center for Education Statistics runs the sample, most standardized tests reward what a student already knows. This one deliberately hands them resources mid-task. That design choice tells you what the OECD thinks the future skill is: not recall, but the ability to pick up an unfamiliar tool quickly and make it do something.
I spent fifteen years building consumer hardware, and that is a fair description of the job. Nobody arrives knowing the tool chain. The engineers who move fastest are the ones who read the documentation, try the smallest possible version, look at what broke, and adjust. The kids at Level 6 are doing the junior version of that.
The OECD’s September 8 release makes the same point about digital tools generally. It also means a low score can mean two different things: a student who cannot build, or a student who would not read the tutorial. Those need different responses at home. The first needs practice. The second needs a reason to persist, which is the PISA engagement story in miniature.
What to actually do at home
Give the tutorial habit a name
When your child hits something unfamiliar, the instinct is to ask you or to quit. Name the third option: “read the first two steps, try one thing.” Say it often enough that it becomes what they do before they ask.
Insist on one change at a time
Kids debug by changing five things at once and hoping. Make “change one thing, then test” the house rule for anything that is not working, from a Minecraft redstone circuit to a bike brake. It is the single most transferable habit on this list.
Keep a results notebook
A cheap notebook where experiments get written down (what I changed, what happened) turns guessing into data analysis. PISA rewards analyzing the data you generated. Most kids never generate any.
Let the frustration sit for ninety seconds
The construct explicitly includes staying engaged “even when it becomes challenging or frustrating.” If you rescue at the first sign of friction, you are removing the thing being measured. Count to ninety before helping.
What not to do
Do not buy a coding course on the strength of a PISA domain name. The assessment rewards experimentation, debugging, and persistence, and a curriculum that walks a child through pre-built projects step by step can produce completion without any of the three. If a program never lets your kid be stuck, it is not building this skill.
What to Watch For Over the Next 3 Months
- Week 4: Your child changes one variable at a time on something that is broken, without being told.
- Month 2 red flags: Asking for help before trying anything, or abandoning a task within a minute of the first error. Both are the self-regulation gap PISA is measuring.
- Month 3 self-check: Ask your child to explain a bug they found and fixed this month. If they can describe the test they ran, the habit has taken.
Frequently Asked Questions
Is computational problem solving the same as coding?
No. Programming is one of the tools PISA students could use, alongside computer modeling, but the domain also covers running controlled experiments, analyzing results, and debugging. A kid can be good at this and know no programming language, and a kid can know Python and still score at Level 1.
How did the United States do?
The US averaged 513 against an OECD average of 500, above average on an absolute scale. Top OECD countries were Australia, Estonia, Japan, Korea, and New Zealand, all above 530. Macao (China) led overall at 572.
What level should I want my child to reach?
The OECD says there is no formal baseline, but flags Level 3 as the practical benchmark: the point where students combine a range of commands to build working programs and can run basic controlled experiments. About 64% of OECD students reach Level 3 or higher.
Can this be practiced without a screen?
Yes, and I would start there. Systematic experimentation, one-variable-at-a-time testing, keeping a results record, and hypothesis-driven debugging all work with paper airplanes, dice, and recipes. The screen version adds syntax, not thinking.
Why did some countries score higher here than in science?
The OECD reports that Hong Kong (China), Malaysia, Brunei Darussalam, the Philippines, Macao (China), and Iceland performed over 30 points above what their science scores predicted. The report does not explain why, and I would not guess. It is a genuinely open question worth watching in Volume III.
About the author
Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.
Sources
- OECD. (2026). “PISA 2025 Results (Volume I): Future-Ready Students,” Chapter 2, computational problem solving section and Table I.2.4. OECD Publishing. https://www.oecd.org/en/publications/pisa-2025-results-volume-i_73451bc5-en.html
- OECD. (2026). “PISA 2025 Assessment and Analytical Framework.” OECD Publishing. https://doi.org/10.1787/86c36975-en
- OECD. (2026). “PISA 2025 Results (Volume I): Country Note, United States.” OECD Publishing. https://www.oecd.org/en/publications/pisa-2025-results-volume-i-country-notes_2d4ff9ea-en/united-states_0c8cbc7c-en.html
- National Center for Education Statistics. (2026). “Program for International Student Assessment (PISA).” NCES. https://nces.ed.gov/surveys/pisa/
- OECD. (2026). “PISA 2025: Students’ reading and mathematics performance declined sharply across the OECD.” OECD press release, September 8, 2026. https://www.oecd.org/en/about/news/press-releases/2026/09/pisa-2025-students-reading-and-mathematics-performance-declined-sharply-across-the-oecd.html
- Barnum, M., & Meltzer, E. (2026, September 8). “PISA results 2025 show U.S. gaining as other countries decline.” Chalkbeat. https://www.chalkbeat.org/2026/09/08/pisa-results-2025-show-us-gaining-as-other-countries-decline/
