GCSE Computer Science: What the 'Algorithms' Topic Actually Covers — and Why Students Find It Harder Than Programming

When parents hear "algorithms" in GCSE Computer Science, many picture their child typing code into a computer. In the specification, it means something broader and more abstract: the thinking that comes before the code is written, plus the ability to describe, trace and compare different ways of solving a problem.

That distinction matters because algorithms are examined on the written papers, not through practical coding. A student who happily writes a working Python program at home can still lose marks on algorithm questions because the skills being tested are different ones.

What "algorithms" means in the specification

An algorithm is a precise sequence of steps that solves a problem or completes a task. In GCSE Computer Science, the topic covers four main things:

• Designing algorithms: working out a step-by-step method for a given problem.
• Expressing algorithms: writing those steps using flowcharts, pseudocode or structured English.
• Tracing algorithms: following an algorithm by hand to work out what it does with particular inputs, often using a trace table.
• Evaluating algorithms: comparing two methods, spotting errors, or commenting on efficiency.

Students also need to know a small set of standard algorithms. Across the main English and Welsh boards this typically includes:

• Searching: linear search and binary search.
• Sorting: bubble sort, merge sort and insertion sort (the exact combination depends on the board).

Parents can check the detail on each board's own specification page: AQA 8525, OCR J277 and Eduqas (WJEC in Wales) all publish the current GCSE Computer Science specification on their websites. All three include algorithms as a named topic area, and all three assess it through written exams rather than coursework.

How it's actually tested

Algorithm questions appear on written papers and tend to take one of these forms:

1. "Complete the flowchart" or "Complete the pseudocode" — a partially written algorithm with gaps.
2. "Trace the algorithm" — a trace table where the student fills in the value of each variable after every step.
3. "Write an algorithm" — usually in pseudocode, to solve a short described problem.
4. "Compare" or "Explain" — for example, why binary search is faster than linear search, or when it cannot be used.
5. "Find and correct the error" — a given algorithm that doesn't quite work.

A crucial point: pseudocode is not a real programming language. Each exam board uses its own style, and students are expected to write in a way that is clear and unambiguous rather than runnable. OCR, for instance, publishes a reference pseudocode guide for J277. AQA accepts answers in their own pseudocode style or in a recognised high-level language, provided the logic is clear. It is worth checking what the student's own board expects.

Why strong coders can still struggle

This is the part that surprises many parents. A student who enjoys programming can still underperform on algorithm questions, for several reasons.

Coding is interactive; algorithm design is not

When a student writes code at a computer, they get instant feedback. The program runs, an error appears, they fix it and try again. Much of their "thinking" is actually guided by the computer telling them what went wrong.

In an exam, there is no computer. The student has to hold the whole algorithm in their head, predict what it will do, and spot mistakes without ever running it. That is a genuinely different skill, and it has to be practised deliberately.

Pseudocode feels fiddly

Students used to Python often write Python when a question asks for pseudocode, or mix the two. That isn't always penalised, but it can be if the question specifically asks for pseudocode in the board's style, or if the Python relies on features (like built-in sort functions) that sidestep the point of the question.

Tracing requires patience, not cleverness

A trace table question might involve a loop that runs eight times. The marks come from writing out every value, in order, without skipping steps or doing arithmetic in your head. Students who are quick thinkers often rush and make slips. The method is simple; the discipline is harder.

Explaining "why" is a separate skill

"Why is binary search more efficient than linear search?" is not a coding question. The answer involves explaining that binary search halves the search space each time, so it takes far fewer comparisons on large lists — but it requires the data to be sorted first. A student can know how to code both searches without being able to articulate the trade-off in writing.

A small worked example

Here is a typical trace question. Given this pseudocode:

total = 0
FOR i = 1 TO 4
    total = total + i
NEXT i
OUTPUT total

A trace table would look like this:

Output: 10

Simple enough — but under exam pressure, students commonly either forget the starting value of total, miscount the loop, or jump straight to the final answer without filling in the intermediate rows. In many mark schemes, the intermediate rows are where the marks sit.

How to help at home

Parents don't need to teach pseudocode themselves. A few practical things do help:

• Ask them to explain an algorithm out loud, without a computer. If they can talk you through linear search step by step, they understand it. If they can only "show you in Python", they may be relying on the computer.
• Practise trace tables on paper. Past papers from the student's own exam board are the best source, because the pseudocode style will match.
• Ask "what if the list is already sorted?" or "what if it's empty?" — the kind of edge-case questions that evaluation questions are built on.
• Separate the two kinds of revision. Programming practice (at a computer) and algorithm practice (on paper) are both needed. One does not replace the other.

If a student is clearly comfortable with coding but losing marks on written algorithm questions, that's a specific, fixable gap. It usually responds well to targeted practice on the exact style of question their board uses, rather than more general revision. One-to-one tuition can help by working through past-paper algorithm questions in the student's own board's pseudocode and focusing on the written-paper technique that classroom lessons often have less time for. At Luminary Tutoring, GCSE Computer Science lessons cover all exam boards, so the practice matches the specification your child is actually sitting.

A sensible next step is to look at one algorithm question from a recent past paper for your child's board, and ask them to answer it on paper, without a computer. The gap between how confident they feel and how many marks they'd score tells you a lot.