Table of Contents
Computational Thinking Activities Kids Can Do in 12 Weeks
PISA 2025 added a computational problem-solving test. These computational thinking activities kids can do at the kitchen table build the same skill in 12 weeks.
In September 2026, the OECD published the first PISA results for a new domain: computational problem solving. Not coding. Problem solving with computational tools, measured by whether a 15-year-old can build a working model, run an experiment on it, read the data, and debug what they built. U.S. students averaged 513 against an OECD average of 500. Macao scored 572. Singapore had more than half its students at the top two levels, against about a quarter across the OECD. The gap is not about who owns a laptop. The computational thinking activities kids need to close it cost almost nothing, and the research says the paper-and-pencil ones work too. Here is a 12-week sequence you can run at your kitchen table, in order, with the evidence for each stage.
Key Takeaways
- PISA 2025’s computational problem-solving domain tested modelling, programming tools, running experiments, analysing data, and debugging, not syntax. Top scores: Macao 572, Singapore 563, B-S-J-Z China 560, Japan 557; U.S. 513, OECD average 500.
- A 2023 systematic review and meta-analysis in the International Journal of STEM Education covering 49 K–12 studies found unplugged activities produced a large effect on computational thinking (Hedges’s g = 1.028, 95% CI [0.641, 1.415]).
- A 2025 meta-analysis found on-screen activities slightly outperform unplugged ones (g = 0.606 vs 0.501). Both work. The cheap one works nearly as well, which matters if your budget is zero.
- The sequence that matches the evidence: four weeks unplugged, four weeks Scratch, four weeks Python, with debugging built into every week.
- The single highest-value habit: make your kid explain the bug before fixing it. Debugging out loud is the closest home analogue to what PISA actually scored.
What PISA 2025 actually tested
Computational problem solving, as PISA defines it, is the capacity to make progress toward a goal and build a functional solution using computational tools such as modelling or programming environments, plus practices like running experiments, analysing data, and building and debugging computational artefacts. Read that definition twice. There is no mention of a language. A student who can plan, test, read a result, and fix what’s broken scores well whether they’ve written a line of Python or not.
The OECD’s Volume I report, released September 8, 2026, put Macao (China) first at 572, Singapore at 563, the Chinese provinces of Beijing-Shanghai-Jiangsu-Zhejiang at 560, and Japan at 557. The OECD average was 500 and the United States came in at 513. That’s one of the few PISA 2025 numbers where U.S. students beat the average, and Chalkbeat noted that American 15-year-olds were also less likely than peers elsewhere to answer hastily, which suggests real engagement rather than guessing.
One caveat on the U.S. figure: over 40% of American test-takers didn’t finish the background questionnaire this round, and the U.S. results carry an asterisk for incomplete school response. Treat 513 as a reasonable estimate, not a precise ranking.
Why the best computational thinking activities kids do start on paper
This is where most parents go wrong. The instinct is to buy something, a robot kit, a tablet app, a subscription. The evidence says start with paper.
A 2023 systematic review and meta-analysis in the International Journal of STEM Education examined 49 K–12 studies published between 2006 and 2022 on unplugged activities and computational thinking. The meta-analytic portion, 13 studies with 16 effect sizes, found a large overall effect (Hedges’s g = 1.028, 95% CI [0.641, 1.415]). Board and card games were the most common activity type. Del Olmo-Muñoz, Cózar-Gutiérrez, and González-Calero, writing in Computers & Education in 2020, found unplugged activities improved computational thinking in early primary students and also raised motivation.
The honest counterweight: a 2025 meta-analysis in Educational Technology Research and Development comparing the two approaches head-to-head found plugged (on-screen) activities slightly more effective (g = 0.606) than unplugged (g = 0.501). So screens aren’t worse. Unplugged is close, free, and works for kids who can’t yet type. Sequencing from unplugged to plugged is a practical choice, not a research-mandated one.
The mechanism, in plain terms: computational thinking is the habit of turning a vague problem into steps a dumb machine could follow. The dumb machine can be a sibling. When your 7-year-old writes instructions for making a sandwich and you follow them literally, putting the peanut butter on the outside because they never said “open the bread,” they have learned the thing PISA measures. No electricity required.
How to Teach Your Kid About Computational Problem Solving
Ages 5–8: The literal robot
You are the robot. Your child gives instructions, you execute them exactly as stated, and nothing else. Getting a cup of water. Drawing a square. Putting on a shoe. Every failure is information, and it’s funny, which is why it sticks. Fifteen minutes, twice a week. The skill being built is decomposition: breaking a goal into steps small enough to be unambiguous.
Ages 9–12: Sort a deck, then time it
Give them a shuffled deck of 20 cards and ask them to sort it, writing down the method afterward. Then do it again with a different method and time both. They will discover, on their own, that some methods are faster. That’s algorithm analysis, and it’s the same idea as Big-O notation, minus the notation. Follow with a Scratch project that has a bug you introduced on purpose.
Ages 13+: Build something that breaks
The project matters less than the failure. A Python script that renames a folder of photos, a Scratch game with a scoring bug, a spreadsheet model of their allowance. The requirement: they must keep a bug log, one line per bug, with the symptom, the guess, and the fix. Three weeks of that log is worth more than any course.
The question to ask: “Before you change anything, tell me what you think is wrong and how you’d know if you were right.” A kid who can answer that is debugging. A kid who can only change things at random is guessing, and guessing is what PISA’s lowest levels look like.
The 12-week kitchen-table sequence
Twenty to forty minutes, two or three times a week. Nothing here requires a purchase beyond a deck of cards and a free Scratch account. The last four weeks need a computer with Python installed, which is free.
| Week | Activity | Tool | Skill it builds |
|---|---|---|---|
| 1 | Literal-robot sandwich / water instructions | Paper, a willing adult | Decomposition, precision of language |
| 2 | Draw-by-instruction: one kid describes a shape, the other draws it blind | Paper, pencils | Abstraction, communication of state |
| 3 | Card sorting: invent two methods, time both | Deck of cards, phone timer | Algorithms, efficiency comparison |
| 4 | Treasure-hunt maze with numbered instruction cards, including a deliberate error | Index cards, floor tape | Sequencing, first taste of debugging |
| 5 | First Scratch sprite: make it move in a square | Scratch (free) | Loops, coordinates |
| 6 | Add a score and a win condition | Scratch | Variables, conditionals |
| 7 | Introduce a bug on purpose, then swap projects with a sibling or friend to fix each other’s | Scratch | Reading someone else’s logic |
| 8 | Extend the game with one feature they choose | Scratch | Planning, incremental building |
| 9 | Python: print, input, variables. Write a program that asks for a name and a number and prints a sentence | Python (free) | Syntax as a new notation for old ideas |
| 10 | Loops and lists: a program that prints a multiplication table, then one that finds the biggest number in a list | Python | Iteration, data structures |
| 11 | A real utility: rename files, total a shopping list, or convert units | Python | Modelling a real problem |
| 12 | The bug log review: pick the three hardest bugs from 11 weeks and explain each one to you | Notebook | Metacognition, the PISA skill |
The transitions matter more than the weeks. Week 5 is the first screen, and it should feel like a reward for four weeks of paper. Week 9 is the jump to text, which is where most kids quit; our guide on when to switch from Scratch to Python covers the readiness signals. Week 12 is the whole point.
What to do at home while the 12 weeks run
Keep the bug log visible
A cheap notebook on the kitchen counter. One line per bug: what happened, what you thought was wrong, what was actually wrong. Kids who write these down start recognizing bug patterns by week 6. It’s also the only artifact that shows progress, which matters when motivation dips around week 7.
Don’t fix it for them
The hardest parenting move in this entire sequence. When a 10-year-old’s Scratch sprite won’t move and you can see the problem in two seconds, say nothing. Ask, “What did you expect to happen, and what happened instead?” That question is the whole discipline.
Do it with them, badly
Write your own Python script and let them see you fail at it. The research on co-learning is decent and the modelling effect is obvious: a kid who watches a parent debug learns that being stuck is normal. Our piece on what research says about learning STEM alongside your kids covers why using real tools beats toy versions.
Twenty minutes beats two hours
Three 25-minute sessions a week finishes the sequence. One three-hour Saturday marathon does not, because debugging requires a fresh brain and a frustrated kid learns that programming feels bad.
What not to do
Don’t buy the robot kit in week 1. A kid who hasn’t done the literal-robot game will treat the robot as a toy that moves, not a machine that executes instructions. If you want hardware, add it at week 8, when they already know what a loop is. And skip the “learn to code in 30 days” courses: PISA measured debugging and modelling, not syntax speed.
What to Watch For Over the Next 3 Months
- Week 4: Your kid gives instructions that anticipate a literal interpretation (“open the bread bag first”). That’s the whole unplugged phase paying off.
- Month 2 red flags: They change code at random and hope. They can’t say what they expected to happen. They ask you to fix it before trying anything. All three mean go back to week 3 and do more card sorting; the abstraction didn’t land.
- Month 3 self-check: Hand them a working program with one bug you added and time how long until they find it. Under ten minutes at age 11 is strong. Then ask them to explain the fix. If the explanation is clear, they have the skill PISA scored.
Frequently Asked Questions
Does my kid need to know math first?
No. The unplugged weeks need counting and reading. The Scratch weeks need coordinates, which Scratch teaches visually. Python’s loops and lists need no algebra. If anything, computational thinking helps math rather than depending on it; see our piece on computational thinking vs. coding.
Is Scratch too babyish for a 12-year-old?
It depends on the project, not the tool. A 12-year-old building a platformer with collision detection and a scoring system is doing real work. If they resist the look of it, go straight to Python at week 5 and accept slower progress, or try micro:bit, which is text-optional and has physical output.
What if we skip the unplugged weeks?
You’ll probably be fine, and the 2025 head-to-head meta-analysis suggests on-screen activities are marginally more effective anyway. But the unplugged weeks are where the precision-of-language habit forms, and they cost nothing. Skipping them mostly makes sense for a kid already comfortable with a screen-based tool.
How do I know if it’s working if there’s no test?
The bug log and the week-12 explanation. Both are better signals than any online quiz, because they measure whether your child can reason about a system they built, which is exactly what the PISA scale describes.
Can I do this with two kids of different ages?
Yes, and it works better. Give the older kid the job of executing the younger one’s instructions in weeks 1–4, then have them swap Scratch projects to debug each other’s in week 7. Explaining a bug to a sibling is the strongest version of the week-12 exercise.
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). “Student performance in PISA 2025.” PISA 2025 Results (Volume I). September 8, 2026. https://www.oecd.org/en/publications/pisa-2025-results-volume-i_73451bc5-en/full-report/student-performance-in-pisa-2025_23e075c2.html
- OECD. (2026). “PISA 2025 Results (Volume I).” https://www.oecd.org/en/publications/pisa-2025-results-volume-i_73451bc5-en.html
- Chalkbeat. (2026). “PISA results 2025 show U.S. gaining as other countries decline.” September 8, 2026. https://www.chalkbeat.org/2026/09/08/pisa-results-2025-show-us-gaining-as-other-countries-decline/
- Lu, C., Macdonald, R., Odell, B., Kokhan, V., Demmans Epp, C., & Cutumisu, M. (2023). “Fostering computational thinking through unplugged activities: A systematic literature review and meta-analysis.” International Journal of STEM Education, 10, 47. https://stemeducationjournal.springeropen.com/articles/10.1186/s40594-023-00434-7
- Del Olmo-Muñoz, J., Cózar-Gutiérrez, R., & González-Calero, J. A. (2020). “Computational thinking through unplugged activities in early years of Primary Education.” Computers & Education, 150, 103832. https://www.sciencedirect.com/science/article/abs/pii/S0360131520300348
- “Comparing the effects of unplugged activities and plugged activities on the development of students’ computational thinking: a meta-analysis.” (2025). Educational Technology Research and Development. https://link.springer.com/article/10.1007/s11423-025-10522-w
- OECD. (2026). “PISA 2025 Assessment and Analytical Framework.” https://www.oecd.org/content/dam/oecd/en/publications/reports/2026/05/pisa-2025-assessment-and-analytical-framework_3abce6d6/86c36975-en.pdf
- MIT Media Lab. Scratch. https://scratch.mit.edu
