GCSE Computer Science: What 'Decomposition' Looks Like in an Exam Answer — and Why Students Lose Marks

When a GCSE Computer Science question asks a student to "decompose" a problem, examiners are looking for something quite specific: evidence that the student can take a large, messy task and split it into smaller, named parts that could each be tackled separately. Many students lose marks because they instead write a list of steps — essentially an algorithm — and hope that counts.

It doesn't. Steps describe how something is done in order. Decomposition shows what the problem is made of. That distinction is the difference between a weak answer and a mark-worthy one.

What examiners actually want to see

Across the main exam boards (AQA, OCR, Edexcel, Eduqas and WJEC), decomposition questions tend to reward answers that:

• Identify the overall problem clearly.
• Break it into sub-problems that are smaller, self-contained tasks.
• Where appropriate, break those sub-problems down further.
• Make clear how the parts fit together to solve the original problem.

A good answer reads less like a recipe and more like a plan for who-does-what. Each sub-problem should be something a programmer could, in principle, go away and work on independently.

A weak answer vs a mark-worthy answer

Consider this common question type:

"A school wants a system that allows students to borrow books from the library. Describe how decomposition could be used to help design this system."

A weak answer

First the student logs in. Then they search for a book. Then the system checks if it is available. Then it records the loan. Then it prints a receipt.

This is a sequence of steps. It might describe the system working, but it doesn't show the student breaking the problem into parts. In most mark schemes this would score poorly — often one mark at best, if any.

A stronger answer

The problem can be broken down into several sub-problems:

• User management: registering students, storing their details, handling logins.
• Book catalogue: storing book information, allowing searches by title or author.
• Loans: recording which student has borrowed which book, and when it is due back.
• Returns and overdue items: updating records when a book comes back and identifying late returns.
• Reporting: producing lists of current loans or overdue books for staff.

Each sub-problem can be designed, programmed and tested on its own, and then combined to form the full system.

Notice what this answer does:

1. It names distinct areas of the problem.
2. Each area is a recognisable chunk of work, not a single action.
3. It explicitly says the parts can be developed independently and then combined — which is the whole point of decomposition.

That final sentence matters more than students realise. Examiners often reward candidates who show they understand why decomposition is useful, not just that they can do it.

The common trap: confusing decomposition with algorithm design

This is where most marks are lost. A question about decomposition is not asking for an algorithm. If the student finds themselves writing "first… then… next… finally…", they are almost certainly answering the wrong question.

A useful check for your child: after writing their answer, can they point to three or four nouns describing parts of the system or problem? If the answer is mostly verbs in order (log in, search, check, record), it is a sequence, not a decomposition.

A second example: planning an algorithm

Decomposition appears in algorithm questions too, not only system design.

"A program is needed to work out a student's final grade from their coursework mark and exam mark, apply a late penalty if coursework was handed in late, and display a message. Explain how decomposition could help plan this algorithm."

A mark-worthy response might identify these sub-problems:

• Input: collecting the coursework mark, exam mark, and whether the work was late.
• Validation: checking the marks are within the allowed range.
• Penalty calculation: reducing the coursework mark if it was late.
• Grade calculation: combining the two marks using the correct weighting.
• Output: converting the final mark into a grade and displaying the appropriate message.

Each sub-problem could become its own subroutine (a function or procedure) in the final program. Mentioning this link to subroutines is often worth an additional mark, because it shows the student understands why we decompose: to make code easier to write, test and reuse.

What a good written response includes

When your child is practising these questions, encourage them to include:

• A one-line statement of the overall problem.
• A clearly labelled list of sub-problems, each described in a sentence.
• Where relevant, a brief note that each sub-problem could be coded as a separate subroutine.
• A short justification: easier to test, easier to divide between people, easier to maintain, easier to reuse.

Three or four well-chosen sub-problems usually score better than ten vague ones. Examiners want to see judgement, not a brainstorm.

Common mistakes to watch for

• Writing an algorithm instead of a decomposition. The single biggest cause of lost marks.
• Sub-problems that are too small, such as "ask for name", "ask for age", "ask for class". These are steps within a sub-problem, not sub-problems themselves.
• Sub-problems that overlap heavily, suggesting the student hasn't really separated the parts.
• Forgetting to explain the benefit. Many mark schemes award a mark for stating why decomposition helps.
• Treating decomposition as a one-off definition rather than something to actually do in the answer.

How to help your child practise

Pick a familiar system — a cinema booking site, a fitness tracker, a messaging app — and ask your child to list four or five sub-problems it would need to solve. Then ask them to pick one sub-problem and break that down further. Doing this out loud, without writing code, builds the habit of thinking in parts rather than steps.

If your child consistently writes algorithms when asked to decompose, or struggles to see where one sub-problem ends and another begins, that is exactly the sort of specific weakness one-to-one tuition can address. At Luminary Tutoring we work through past-paper questions of this type with students, pointing out where an answer would gain or lose marks and helping them develop a reliable approach before they reach the exam hall.