IGCSE Computer Science 0478: Half Your Grade Is Written in Pseudocode

IGCSE Computer Science attracts a particular kind of student: the one who already codes. They arrive confident, coast through the theory, and are genuinely surprised when the grade comes back at a B.

It happens because 0478 is two papers that reward opposite habits. Paper 1 is 60% recall. Paper 2 is 60% application. Being good at one does nothing for the other, and each is worth exactly half the qualification.

There is also no tiering to hide behind: both papers assess the full grade range, A* to G. Everyone sits the same exam — unusual among IGCSEs, where maths and the sciences all split Core from Extended, and worth knowing before you read anything about how IGCSE tiering works.

The inversion, in one picture

Paper 1 — Computer Systems

1 hour 45 minutes, 75 marks, 50% of the grade, and 60% of its marks are AO1 — knowledge and understanding. This is the paper that rewards knowing things precisely.

The content sits in six areas of the syllabus:

• Data representation
  Binary, hexadecimal, character sets, images, sound, compression. Conversion questions are free marks if you have drilled them and a slow disaster if you haven't.
• Data transmission
  Packets, transmission modes, USB, error checking. Parity, checksums and check digits are asked repeatedly.
• Hardware
  The CPU and the fetch–decode–execute cycle, memory, storage, embedded systems, logic gates in context.
• Software
  System and application software, operating systems, interrupts, translators — compilers versus interpreters.
• The internet and its uses
  How the web works, cyber-security threats and the specific countermeasures for each. Security questions want the named threat matched to the named defence.
• Automated and emerging technologies
  Automated systems, robotics, artificial intelligence. Short, and often the least revised.

Paper 2 — Algorithms, Programming and Logic

Same length, same marks, same weight — and the objectives flip: 60% AO2, application in context. You are asked to do things, not recall them.

The content covers algorithm design and problem-solving, programming, databases and Boolean logic. And it finishes with the question that decides a lot of grades.

The 15-mark unseen scenario

The final question in Paper 2 is a 15-mark unseen scenario: you are given a context you have not met and asked to write an algorithm for it, in pseudocode or program code. Fifteen marks is a fifth of the paper and a tenth of the whole qualification, in one question, with no worked example to lean on.

Two things make the difference here, and neither is talent:

1. Fluency in one language, plus Cambridge's pseudocode
   You may answer in pseudocode or in program code. Whichever you choose, the version that scores is the one you can write quickly and correctly under pressure. Cambridge publishes its pseudocode conventions — learn them as a dialect, because examiners read them as one.
2. A standard structure you always start from
   Declare, initialise, input with validation, process with the right loop, output. Students who have a default skeleton write something creditworthy within a minute; students who start from a blank page lose time deciding how to begin.
3. Trace tables until they are boring
   Tracing is the most reliably examined skill in the paper and the most mechanical to master. It is also how you check your own algorithm before the marks are lost.
4. Write the validation
   Range checks, type checks, length checks, presence checks. Mark schemes award them explicitly and candidates skip them constantly under time pressure.

Boolean logic and databases: small topics, reliable marks

Two areas repay revision out of proportion to their size. Boolean logic — logic gates, truth tables, logic expressions — is entirely learnable and entirely predictable; there is no reason to lose these marks. Databases — single-table concepts, primary keys, data types and SQL queries — is similarly bounded, and SQL in 0478 is a small, closed set of patterns rather than a language you must master.

Both are exactly the kind of content where twenty focused questions produce more improvement than an hour of reading, which is the general case we made in how many past papers you actually need.

A revision plan that matches the weightings

What we hold for 0478

Computer Science is one of the platform's larger untapped banks: 6,391 tagged questions across 142 topics, with 192 revision-note sets and materials. The topic tree follows the syllabus exactly — data representation, data transmission, hardware, software, the internet and its uses, automated and emerging technologies, algorithm design and problem-solving, programming, databases, Boolean logic — with a separate exam technique branch, which for this subject means the algorithm-writing patterns rather than general advice.

Instant marking matters more here than in most subjects, for a specific reason: an algorithm you believe works and an algorithm that earns marks are different objects, and the gap is only visible against a mark scheme. Open the Computer Science 0478 library, and if Paper 2 is the weaker half, start at algorithm design and problem-solving and programming rather than anywhere else.

If you are choosing subjects rather than revising them, choosing IGCSE subjects backwards from your destination and which IGCSEs are actually hardest are the two posts worth reading first — and ICT 0417 is the subject most often confused with this one, which we cover separately — along with which ICT topics come up most. For the underlying habit, spaced retrieval is what keeps Paper 1's definitions available months after you learned them, and command words decide how much of your Paper 1 answer actually scores.

FAQ

Split your revision the way Cambridge splits the marks — half theory, half algorithms — and write code by hand until the 15-mark question stops being frightening. Open the Computer Science 0478 library and create a PapaMarks account.