How to get an A or A* in Cambridge IGCSE Computer Science (0478)

Translate theory into a sequence of actions

When revising a system, explain what enters, what happens and what leaves. For data transmission, identify how the data is represented, sent and checked. For an automated system, trace the path from sensor input through processing to an actuator response.

Avoid collecting disconnected definitions. Compare closely related ideas such as authentication and encryption, or validation and verification. Give a situation where one is useful and explain why the other does not solve exactly the same problem.

Write and trace small algorithms frequently

Start with a clear input, processing rule and expected output. Build short routines that total values, count items meeting a condition or find a maximum. Trace the values of each variable after every iteration; do not rely on what you intended the algorithm to do.

Use real programming practice to test your logic, but also rehearse the exam’s required representation. In the 2026–2028 syllabus, coding solutions generally require pseudocode; the 15-mark scenario question has an exception allowing the specified programming languages. Read the official rules rather than assuming any code is acceptable.

Test where an algorithm is most likely to fail

For each input rule, choose ordinary valid data, extreme valid values and values just outside the permitted boundary. If zero items can occur, test that case too. Work out the expected result before running or tracing the solution.

An algorithm that works once is not necessarily correct. Check loop start and end values, initialisation, division by zero and conditions using greater-than versus greater-than-or-equal. Give variables meaningful names so these mistakes are easier to see.

Turn a long scenario into manageable requirements

Read the problem and list every required input, calculation, validation and output. Group related tasks before writing a solution. A plan prevents you from spending most of the time polishing one section while missing another requirement completely.

After writing, trace a small example through the entire solution. Check the data structure sizes and indices. Then compare your output with what the user requested, including labels or summaries. Correct arithmetic is not enough if the result answers a different question.

Keep number systems and logic active

Practise conversions and logical expressions in short sessions without depending on a calculator. Show the intermediate steps so you can locate errors. In a truth table, use a consistent input order and calculate intermediate columns for a complicated expression.

Alternate these exercises with written theory and programming. A mixed mock reveals whether you can switch between them under pressure. Record errors by type: knowledge, interpretation, tracing, syntax convention or incomplete requirements, then repair the category that keeps recurring.

Common mistakes—and how to fix them

Your four-week revision plan

Use this as a focused revision cycle alongside your lessons. If you are still learning the course, give unfamiliar topics more time before attempting full papers.

1. Week 1

   Audit systems and logic

   • Explain data representation and hardware processes from memory.
   • Practise number-system conversions and complete truth tables.
2. Week 2

   Build reliable algorithms

   • Write short search, count and total routines.
   • Trace them and test invalid or boundary inputs.
3. Week 3

   Decompose scenarios

   • Turn long tasks into a requirements checklist.
   • Write complete solutions using the representation required for your exam year.
4. Week 4

   Rehearse and repair

   • Attempt both theory and algorithm questions under time limits.
   • Correct repeated logic errors and retest with new examples.

Your questions, answered