Next Level

Why handwriting is the whole point

Every design decision in Next Level follows from one belief: in mathematics, the work is the thinking. An app that only checks the final number teaches children that math is about final numbers. It isn't.

What tapping can't teach

Multiple choice and answer boxes measure one bit per problem: right or wrong. They cannot see whether the child set the problem up correctly, organised the steps, estimated before computing, or checked the result. Those are the skills that separate "can sometimes compute" from "can do mathematics", and they are exactly the skills international assessments like PISA isolate when they ask students to formulate a problem, not just solve it.

So in Next Level there are no answer boxes and no options to tap for credit. Every problem is answered in handwriting, working included. An AI vision model reads the page and gives feedback on three things: the answer, the setup (rated 1 to 3 stars for clarity and structure), and the working. The game's incentives point the same direction: shown work earns roughly three times the points of a bare answer, so the game cannot be won by guessing.

The learning science we lean on

We built deliberately on the least controversial findings in learning research, and we will be honest about the strength of each.

  • Retrieval practice. Producing an answer from memory beats re-reading or re-watching. One of the most robust effects in cognitive psychology (Roediger and Karpicke 2006; Rowland 2014). Every item in the app is a retrieval event; there is nothing to passively consume for credit.
  • Spaced practice. Skills decay when practice is massed and stick when it is distributed (Cepeda et al. 2006). Mastered units resurface on expanding intervals: 1, 3, 7, 16, 35 days.
  • Worked examples with fading. Novices learn methods best from worked examples whose support is gradually removed (Sweller et al. 1998; Barbieri et al. 2023, a meta-analysis in mathematics specifically). Lessons run worked example, then scaffolded problems, then independent practice.
  • Productive struggle. Attempting a novel problem cold, then getting the lesson, improves conceptual understanding (Kapur 2014). Every lesson opens in try-first mode.
  • Immediate, task-focused feedback. Feedback helps when it is fast and about the task, not the person (Hattie and Timperley 2007; Shute 2008). Feedback arrives in seconds and talks about the work: "write each step under the last", not "good job".
  • Mastery gating. Mathematics is hierarchical; moving on with gaps compounds them. Each unit ends in a mastery check and the next opens at 80%, not when the calendar says so.

And handwriting itself? Here is the honest version. There is suggestive evidence that writing by hand engages broader sensorimotor networks and supports deeper encoding than typing (Mueller and Oppenheimer 2014; Van der Weel and Van der Meer 2024). But most of that research is about note-taking and letter learning, not about doing mathematics by hand, and some results have faced replication challenges. We treat pen-first as a well-motivated design bet, not a proven law. What is not a bet: handwriting makes the child's thinking visible, and visible thinking can be taught. Tapped answers cannot.

Why now, in Iceland

In PISA 2018, Icelandic 15-year-olds scored 495 in mathematics, above the OECD average. In PISA 2022 the score was 459, below the OECD average of 472, and roughly one in three students was below the baseline level the OECD considers necessary for a numerate adult life. (Figures from the OECD's PISA 2022 results, verified against the primary report.)

At the same time, the feedback a family actually receives is thin: a handful of letter grades, each spanning a wide range of actual skill. A child can carry a fine-looking grade while quietly missing prerequisites, and a parent has almost no way to see it.

Next Level is one parent's answer to that gap: daily, handwritten, mastery-gated practice mapped to the Icelandic national curriculum and to the mathematical domains PISA measures, with the child's real work kept and visible. It practices the kind of mathematical reasoning and written problem-solving that PISA measures. It is not a "PISA prep" product, and we promise no score gains; nobody honestly can.

What we deliberately do not claim

  • We do not claim handwriting is proven superior for math. It is our design bet, stated above.
  • We do not promise grade or PISA-point improvements.
  • The app is not affiliated with or endorsed by the OECD, PISA, or the Icelandic Ministry of Education.

References

  • Barbieri, C. A. et al. (2023). A meta-analysis of the worked examples effect on mathematics performance. Educational Psychology Review, 35.
  • Cepeda, N. J. et al. (2006). Distributed practice in verbal recall tasks. Psychological Bulletin, 132(3).
  • Hattie, J. and Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1).
  • Kapur, M. (2014). Productive failure in learning math. Cognitive Science, 38(5).
  • Mueller, P. A. and Oppenheimer, D. M. (2014). The pen is mightier than the keyboard. Psychological Science, 25(6).
  • OECD (2023). PISA 2022 Results (Volume I) and the Iceland country note.
  • Roediger, H. L. and Karpicke, J. D. (2006). Test-enhanced learning. Psychological Science, 17(3).
  • Rowland, C. A. (2014). The effect of testing versus restudy on retention. Psychological Bulletin, 140(6).
  • Shute, V. J. (2008). Focus on formative feedback. Review of Educational Research, 78(1).
  • Sweller, J. et al. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10(3).
  • Van der Weel, F. R. and Van der Meer, A. L. H. (2024). Handwriting but not typewriting leads to widespread brain connectivity. Frontiers in Psychology, 14.