Question 1

When should I start doing OCR A-Level Computer Science past papers?

Accepted Answer

Starting past-paper practice earlier than most students expect tends to pay off, and the approach works best in two distinct phases. As you work through each topic in our notes, we'd suggest using our by-topic exam practice to test yourself on past questions straight away, since this kind of spaced retrieval helps the content stick far better than saving everything until spring.

Then, in the final few weeks before your mocks and again before the May/June series, we'd suggest completing two or three full papers. Both Component 01 and Component 02 are each two hours and thirty minutes long, so building that stamina before your mocks means you go into the real thing feeling genuinely prepared.

Question 2

How many past papers should I do?

Accepted Answer

Knowing exactly how many to aim for can feel tricky, but with first exams in 2017, OCR has built up several years of papers across both Component 01 (Computer Systems) and Component 02 (Algorithms and Programming), so there is plenty to work with. We'd suggest starting with the most recent series first, since question styles can shift over time, and then working back through earlier papers as your confidence grows. Our by-topic exam practice is a great way to target specific areas before you move on to full papers.

Question 3

Should I do past papers under timed conditions?

Accepted Answer

Yes! Working through past papers under timed conditions is genuinely valuable preparation for OCR A-Level Computer Science, because both Component 01 (Computer Systems) and Component 02 (Algorithms and Programming) each run for 2 hours and 30 minutes, and building stamina for that duration matters.

We'd suggest using spaced retrieval with our notes and flashcards to build solid content knowledge first, then moving into timed practice once you feel confident on a topic. That way you're testing genuine recall rather than working through unfamiliar material under pressure.

Our exam practice tools let you work through questions by topic, so you can focus on weaker areas before attempting a full paper. When you do attempt a full paper, remember that no calculators are permitted in either exam, so replicating that condition from the start will help it feel natural on the day.

Question 4

What should I do if I get a low mark on a past paper?

Accepted Answer

A low mark on a past paper is more useful than it might feel right now, because it's showing you exactly where the gaps are. We'd suggest going back through each question you dropped marks on and identifying whether the issue was content knowledge, algorithm-writing technique, or how you structured your extended responses for the levels-of-response questions in paper 02.

Once you know the pattern, our notes and flashcards can help you revisit the specific topics from Component 01 or Component 02 where the marks slipped. Spaced retrieval is particularly effective here: returning to those weak areas at increasing intervals helps the content stick far better than a single re-read.

After working through the content, we'd then suggest using our exam practice to try a fresh paper and see how the extra preparation transfers to your score.

Question 5

How do I mark my own past paper using the OCR mark scheme?

Accepted Answer

Self-marking OCR H446 past papers is most accurate when you understand that the mark scheme uses more than one approach. For shorter questions, work through each mark point in turn and check your answer clearly matches it. For higher-tariff extended-response questions, paper 02 uses a levels-of-response mark scheme, which means you pick the band that best fits your answer as a whole rather than ticking off individual points. When applying levels-of-response, read all the band descriptors and choose whichever one your response meets more of, then place yourself within that band based on how convincingly you hit those descriptors. Our exam practice section on Cognito includes model answers that help you see what a strong response looks like, making it much easier to calibrate your self-marking honestly.