AQA A-Level Computer Science: Preparing for an On-Screen Paper, a Written Paper and the NEA

AQA A-Level Computer Science 7517 is a linear, untiered qualification with three components. Paper 1 is an on-screen examination lasting 2 hours 30 minutes (100 raw marks, 40% of the A-level) testing programming ability and theoretical knowledge from content sections 10-13, together with systematic problem-solving skills from section 22. Paper 2 is a written examination, also 2 hours 30 minutes and 100 raw marks for 40%, testing content sections 14-21. The non-exam assessment – the computing practical project – is worth 75 raw marks and 20%, internally assessed and externally moderated by AQA. These notes complement the site’s guides to Fundamentals of Programming and Data Structures and Arrays.

Paper 1 is sat at a computer – revise at a computer

The single most consequential fact about this specification is that Paper 1 is an on-screen exam in which you write and run code. Revision done entirely on paper does not rehearse the thing being assessed. Exam-preparation priority: practise in the exact environment and language your centre uses, without an internet search available, and get fluent at the mechanical parts – creating a file, running it, reading the error message, fixing it. Candidates who are strong programmers but slow with their toolchain lose real marks to friction that has nothing to do with computer science. AQA also issues Preliminary Material, a Skeleton Program (in each of the programming languages) and, where appropriate, test data for use in the exam, so time spent reading, running and adapting that program before the day is part of Paper 1 preparation.

The NEA is 20% and it is the component most under your control

Unlike the two papers, the practical project is produced over months with feedback available. It is the only component where the mark is not decided in 150 minutes. Exam-preparation priority: treat the project’s documentation – analysis, design, testing evidence, evaluation – as carrying marks in its own right rather than as paperwork around the code. A working program with thin evidence of systematic development typically scores below a more modest program that is thoroughly analysed, designed, tested and evaluated.

The two papers test disjoint content sections – so revision cannot be generic

Paper 1 covers sections 10-13 and 22; Paper 2 covers sections 14-21. They do not overlap in the way many A-level science papers do. Exam-preparation priority: revise by paper, using the specification’s own section numbering as the checklist, rather than working through a textbook in order. It is entirely possible to feel well prepared overall and to have concentrated almost all of that preparation on one paper’s content.

Paper 2’s theory needs precise definitions, not approximate understanding

Sections 14-21 cover material where marks depend on stating things exactly – the specific steps of an algorithm, the exact difference between two similar concepts, the correct terminology for a hardware or networking mechanism. Exam-preparation priority: practise writing definitions from memory and compare them word-for-word against the specification’s wording. “Roughly right” answers in this part of the subject are a well-known source of dropped marks.

Tracing and dry-running are examinable skills in their own right

Both papers reward the ability to work out what a given piece of code does, not just to write code. Exam-preparation priority: practise trace tables deliberately – a column per variable, a row per iteration, written out rather than done mentally. Mental tracing is where confident candidates make avoidable errors, and a written trace table is also creditable working.

Worked routine: a disciplined approach to an on-screen programming task

The routine below is an original model written for this resource, not a reproduction of any official past paper, preliminary material or mark scheme.

Step 1 - read the whole task before writing any code:
Note what the OUTPUT must look like exactly -- format, spacing,
wording. Many marks depend on matching a specified output.

Step 2 - write the smallest thing that runs:
Get a program that compiles and produces some output early, then
build on it. A half-finished program that runs beats a complete
one that does not.

Step 3 - test after each small addition, not at the end:
An error found after three lines is trivial to locate; the same
error found after sixty is not.

Step 4 - if stuck on one sub-task, move on:
Later sub-tasks are usually independently credited. Leave a
comment marking what is unfinished and return if time allows.

Step 5 - save your work as the paper instructs, when it instructs:
Unsaved work is unmarkable work.

Step 2 is the habit that most reliably protects marks: incremental development is both good practice and good exam strategy.

Before/during exam checklist

• Before the exams: practise Paper 1 at a computer in your centre’s language and environment, without internet help; revise Paper 1 and Paper 2 content separately using the specification’s own section numbers; write theory definitions from memory and check them against the specification’s wording; practise written trace tables.
• On the NEA: treat analysis, design, testing evidence and evaluation as mark-bearing in their own right, not as paperwork around the code.
• During Paper 1: read the whole task first and note the exact required output; build incrementally and test as you go; move on from a stuck sub-task; save as instructed.
• During Paper 2: use precise terminology, and write trace tables out rather than tracing mentally.

Self-test

1. Why is paper-only revision insufficient for Paper 1?
2. Which content sections does each paper assess?
3. Why can a modest NEA program outscore a more impressive one?
4. Why is incremental development an exam strategy and not just good practice?

Answers: 1. Because Paper 1 is an on-screen examination in which code is written and run, so fluency with the actual environment – creating, running and debugging files quickly – is part of what determines the mark. 2. Paper 1 covers sections 10-13 plus systematic problem-solving from section 22; Paper 2 covers sections 14-21, with no overlap. 3. Because the NEA credits systematic analysis, design, testing evidence and evaluation, so a well-documented modest project can outscore an impressive one with thin evidence of development. 4. Because a program that runs and does part of the task earns marks, while a more ambitious program that does not compile earns very few – and errors found early are far cheaper to locate than errors found late.

Written against the AQA A-Level Computer Science 7517 specification-at-a-glance and scheme-of- assessment pages (official AQA specification pages, verified 2026-08-28). Marks quoted are the raw per-component marks the board publishes; Paper 1 and Paper 2 raw marks are scaled for the final total. The programming routine above is an original model written for this resource, not a reproduction of any official past paper, preliminary material or mark scheme. Always check the current specification for your examination year at aqa.org.uk.