IGCSE Computer Science (0478): Where Logical Thinkers Still Lose Marks

IGCSE Computer Science is one of the fastest-growing subject choices among Kenyan IGCSE students, and one where strong logical thinking doesn't automatically translate into strong exam marks. The paper rewards a specific kind of precision. Here's what the syllabus covers and where students actually lose marks.

Which Code You're Actually Sitting

0478 is the standard Cambridge IGCSE Computer Science syllabus, graded A* to G. A separate code, 0984, covers the same content under the newer 9-1 grading scale. The content and skills tested are identical between the two, only the grade scale differs. Confirm with your school which code your entry is registered under, since it affects how your eventual grade is reported, not what you need to study.

The Two Halves of the Syllabus

The subject splits into two genuinely different skill areas that need separate revision approaches:

• Theory (computer systems, networks, data representation, algorithms and logic). This half rewards precise, technical vocabulary. "Binary" and "denary" conversions, logic gates, and how data is actually stored and transmitted are tested with exact expected answers, not general understanding.
• Practical programming (algorithm design and problem-solving, usually in pseudocode, plus a programming project component depending on the exam board pathway). This half rewards structured, working logic, not just "an answer that produces the right output." Markers check the actual steps and structure of an algorithm, not only whether a final result seems plausible.

Where Marks Actually Get Lost

Three patterns show up consistently:

• Pseudocode syntax carelessness. Cambridge pseudocode has specific conventions (how loops, conditionals and variable declarations are written). A student who codes correctly in Python but doesn't translate that logic into the exam's expected pseudocode conventions loses marks for presentation even when the underlying logic is sound.
• Confusing "how" with "why." Theory questions often ask students to explain why a particular technical choice is made (why use a particular data structure, why a network uses a specific protocol) rather than just describe how something works. Students who only memorise definitions struggle with the reasoning-based questions.
• Under-practising trace tables. Algorithm-tracing questions, where a student manually steps through pseudocode and records variable values at each stage, are a distinct skill from writing algorithms and need separate, deliberate practice. Many students who can write working code still make careless errors tracing through someone else's.

A Practical Revision Split

Given the two-halves structure, revising strictly subject-by-subject through the syllabus is less effective than splitting practice time between theory recall (with exact technical vocabulary drilled specifically), algorithm writing in the exam's actual pseudocode conventions, and dedicated trace-table practice as its own skill. Past papers, marked strictly against the official mark scheme rather than "does this look about right," reveal exactly which of the three areas needs more work.

Where This Fits

Computer Science sits within the applied sciences that Kenyan students often pair with Mathematics and one of the physical sciences. For general science exam technique that applies across subjects (command words, data-response questions), see our Physics 0625 A* method guide, much of the command-word discipline transfers directly.