Teach-Primary-Issue-20.1

Designing tasks that build computational thinkers will support your curriculum far beyond computing, says Karl McGrath C omputing gives children powerful ways to think; after all, it is a subject that teaches them how to think. When we teach programming and coding well, we help pupils understand how systems behave, why outcomes happen, and how logical decisions shape a sequence. Coding becomes a form of reasoning rather than a set of digital skills. This viewpoint guides my planning, and I try to design tasks that allow children to grapple with concepts such as decomposition, abstraction, and algorithmic thinking, because these habits mirror the disciplinary reasoning found in maths, science, and computing. If children can think well, they can code well. Coding as reasoning We’ve all watched children produce something colourful on Scratch or MakeCode and been unsure about what they have actually learned. Busy screens can mask shallow thinking, but good computing lessons take a different route; they start with the concept, not the tool. They offer children well-structured tasks that slow the process down, create space to examine behaviour, and draw attention to what matters in the code. In my own classroom, I use a combination of paper-based reasoning tasks, prediction prompts, unplugged models, and structured programming episodes. These sit comfortably within the PRIMM approach (Predict, Run, Investigate, Modify, Make). The aim is not to rush children towards an app, a screen, or a finished product. It is to help them articulate what they believe the system is doing, why it behaves the way it does, and how a small change can ripple through a sequence. This approach was spearheaded by Sue Sentance and is supported by research from Computing At School (CAS) and the Education Endowment Foundation (EEF), both of which remind us that problem-solving develops when children explain their thinking and test their ideas through structured practice. Sue’s central argument is that code, as children encounter it, is a language with its own rules and syntax, so it needs to be taught in a similar way to reading and writing. We learn by predicting, inferring, and interpreting before we can create or innovate our own work. The scientific method When we teach coding as reasoning, we lean naturally into the practices that underpin mathematical and scientific enquiry. The scientific method offers a clear structure. It asks children to predict, test, and then evaluate. Coding tasks can mirror this without much additional work or resources. A good example comes from an unplugged comparison task, in which children examine three pieces of MicroPython code printed inside the outline of a head and decide which is correct. The task is deliberately abstract. It asks children to focus on structure rather than presentation. They look for indentation, consistent conditions, and the correct flow in the loop. The head outlines create a sense of Coding for UNDERSTANDING “Coding becomes a form of reasoning rather than a set of digital skills” www.teachwire.net | 41 S T EM S P E C I A L