Parents often ask us a fair question: will learning to code actually help my child at school, or is it just another screen activity? For years the honest answer was "probably, but the evidence is thin." In 2026 that started to change. Several new studies, published in peer-reviewed journals from Springer Nature, Frontiers and Taylor & Francis, looked directly at what happens to children's thinking and schoolwork when they learn to code.
The short answer
Yes, well-designed coding instruction is linked to better math and literacy outcomes in young children, and to stronger problem-solving in older ones. The strongest new evidence comes from a large randomized trial in kindergarten to Grade 2 classrooms. The effect is not magic, though. It depends on how coding is taught.
The biggest 2026 study: 1,705 children in K–2
In May 2026, researchers Ghaida Alrawashdeh and Marina Bers published a randomized controlled trial in the Early Childhood Education Journal. Marina Bers leads the DevTech Research Group at Boston College and created ScratchJr, one of the most widely used coding apps for young children.
The study followed 1,705 students in kindergarten to Grade 2 across two U.S. states. Some classrooms used a coding curriculum and others continued with regular instruction. Two findings stand out for parents:
- Coding skills grew. Children in the coding classrooms made significantly larger gains in coding and computational thinking than children in comparison classrooms.
- The gains carried over. Growth in coding skills significantly predicted improvements in both math and literacy.
The authors note that earlier evidence on this kind of "transfer" was limited and mixed, and that the benefit was strongest when the curriculum was inclusive and deliberately connected to other learning. In other words, coding helped with reading and math when it was taught as a way of thinking, not as an isolated tech skill.
Ages 8–12: problem-solving jumped
A February 2026 experiment in Frontiers in Psychology followed 200 primary school students aged 8 to 12. Half took a 24-week coding course: 16 weeks of Scratch, then 8 weeks of Python. The other half followed the regular curriculum.
- Computational thinking scores in the coding group rose from 20.4 to 30.6, a large effect compared with the control group.
- Problem-solving scores rose from 17.8 to 23.5.
- The biggest improvement was in decomposition, breaking a big problem into smaller parts, followed by abstraction and pattern recognition.
- Boys and girls improved by similar amounts.
- Children who spent more hours coding tended to improve more.
One detail is worth knowing. The problem-solving gains came mostly from logic puzzles and pattern tasks. Arithmetic reasoning did not change significantly. Coding seems to strengthen how children approach problems, which supports math, rather than replacing math practice itself.
Not every program gets the same results
A third 2026 study, from researchers at Durham University in the U.K. and published in Computers in the Schools, looked at volunteer-run after-school Code Clubs. It found small positive effects on coding skills and on non-cognitive skills such as confidence, smaller than in the classroom studies above.
That contrast is useful. Structured, regular instruction with a clear progression produced bigger gains than informal, drop-in activity. For parents, the lesson is to look at how a program is built, not just whether it says "coding" on the label.
Why coding might help with reading and math
Researchers point to a few overlapping skills:
- Sequencing. Putting instructions in the right order is the same skill as retelling a story or following the steps of a math problem.
- Decomposition. Splitting a big task into small steps helps with word problems and with planning a piece of writing.
- Debugging. Finding and fixing a mistake teaches children to check their work rather than give up.
- Precise language. Code only works when instructions are exact, which builds careful reading habits.
Education Week reported in June 2026 that even as the large nonprofit Code.org rebranded as CodeAI to focus on artificial intelligence, it described computational thinking as a foundational skill that helps students build and question technology instead of only using it.
What this means when choosing a program
Based on the 2026 research, look for a program that:
- Meets every week over a full term, not just once.
- Follows a clear progression from blocks to text code as children grow.
- Connects coding to stories, math and real projects.
- Asks children to explain and present their work, which builds language skills.
That is how we built CODEship Academy's four levels, from Explorers in K–Grade 1 to Engineers in Grades 6–8. If you would like to see it for yourself, every family can book one free trial class this November, in person in Oshawa or live online.