GCSE computer science: what actually trips students up.

This year's GCSE computer science results, as summarised by Teach Computing, told a familiar story: attainment edged up, girls outperformed boys by a clear margin, and entries fell for the second year running. Behind those numbers is a subject that students find harder than it looks — and, in our experience, harder for reasons that have little to do with programming languages.

Gap one: reading code is a different skill from writing it

A student can write a working loop and still freeze when asked what a given loop will output. Exam papers lean heavily on tracing — following variables line by line — and it is rarely taught explicitly. We spend whole sessions on it: pencil, paper, a table of variable values, no computer. It feels slow. It moves grades.

Gap two: decomposition

"Write a program that…" is where most students stall, not because they can't code but because the task is too big to start. Breaking a problem into steps — what do I need to store, what happens first, what repeats — is the actual skill, and it transfers to maths, science and every project they will ever do. We teach it before any code is written, usually as a list of plain-English sentences.

Gap three: vocabulary

Computer science is drowning in terms that mean something precise — iteration, abstraction, validation, concatenation — and students who use them loosely lose marks on questions they understand. A weekly five-minute vocabulary check is cheap and disproportionately effective.

What to do about it

• Trace code on paper every week, from the first term.
• Plan in sentences before typing. Keep the plan next to the code.
• Say the words out loud — to a tutor, a parent, a peer. Explaining is testing.
• For A Level: start the project early, and treat the write-up as half the work, because it is.

The falling entry numbers matter, incidentally. Teach Computing points to the wider conversation about AI and programming as one factor. Our view is simple: AI has made the mechanics of writing code cheaper and the thinking behind it more valuable. Which is precisely what a good computer science education teaches.