How to Answer Evaluate Questions in GCSE (9-Mark Structure)

Quick answer: In GCSE, "evaluate" means weigh both sides using specific evidence and then give a clear judgement that answers the question. These questions are levels-marked and usually worth 8, 9 or 12 marks, so the examiner awards a band for the whole answer. Without a supported judgement you are capped mid-band, however much you write.

I do Cambridge IGCSE and A-Levels, and history is my favourite subject, which is exactly where this bit me. In Year 10 I wrote a page and a half for a 12-mark evaluate question, felt quietly smug, and got 5. The comment said plenty of knowledge, little judgement. I had written an "explain" answer to an "evaluate" question. Nothing about my revision was wrong — the shape of my answer was. AQA, Edexcel, OCR and Cambridge IGCSE all define this command word the same way, so fixing it once fixes it in every subject with extended writing.

"Evaluate" is a different job from "explain"

Evaluate asks you to make a decision, not to prove coverage. The command word family runs roughly like this:

• Describe — say what happens. No reasons.
• Explain — say why it happens, as a linked chain of causes.
• Analyse — break something into parts and show how they connect and what follows.
• Evaluate — do the analysis, then decide which side is stronger and say why.

So evaluate contains explain. Explain does not contain evaluate. That is why a beautifully explained answer can still sit stuck in the middle band: it did two thirds of the job. The moment you read the word "evaluate", the last third of your answer is already decided for you — a judgement is coming.

How levels marking works, and why writing more does not help

Levels-marked questions are graded holistically, on best fit. The examiner reads the whole answer, decides which level the reasoning belongs in, then picks a mark inside that level. Nobody is adding a point per fact.

On a typical 9-mark question the marks split across assessment objectives — very roughly 3 for knowledge, 3 for applying it to the context in the question, and 3 for the evaluation itself, though the exact split varies by subject and board. That split is the whole story:

• Facts only: you are competing for about a third of the marks.
• Facts applied to the specific case in the question: about two thirds.
• Facts, applied, then judged: all nine.

An extra paragraph of accurate knowledge on an already knowledge-heavy answer adds nothing, because those marks are banked. One good judgement sentence can add three. That is the trade nobody explains.

The shape that fits the grid

Use five moves per side, then decide.

1. Point — one factor, stated in a sentence that borrows the question's own words.
2. Evidence — a named, numbered, specific detail. A figure, a case study, a date.
3. Chain — two or three sentences of because-therefore reasoning tying that evidence to the point.
4. Mini-judgement — how much this side actually matters here, before you move on.
5. Repeat for the other side, then finish with a decision that names your criterion.

Move 5 is the one almost nobody does. A criterion is the standard you are judging against: cost, time, number of people affected, long-term survival, reliability. "Both matter" is fence-sitting. "X matters more if what you care about is long-term protection, because..." is a judgement, because it says what you weighed on.

A worked 9-mark answer

Question: Evaluate the effectiveness of hard engineering in protecting a coastline you have studied. (9 marks)

Hard engineering can be very effective for the specific stretch it defends. At Mappleton on the Holderness coast, two rock groynes and a rock revetment were installed in 1991 at a cost of around GBP 2 million, holding a coastline that had been retreating at roughly 2 metres a year (point plus specific evidence). Because the groynes trap sediment moving south in longshore drift, the beach in front of the village widened, and a wider beach absorbs more wave energy before it reaches the cliff (chain of reasoning). For the village and the road behind it, that is a clear success (mini-judgement).

However, the same groynes starve the coast downdrift. South of the scheme, erosion rates rose sharply, because sediment that once fed those cliffs is now trapped behind the defences (other side, with a mechanism rather than an assertion). The defence did not remove the erosion. It moved it.

Overall, judged on protecting one settlement over a 30-year window, the scheme worked, and the value of the homes and road saved outweighs the 2 million spent. Judged on protecting the coastline as a whole, it failed, because it transferred the problem to unprotected farmland further south. Effectiveness therefore depends on the scale you choose to measure it at (decision, with a named criterion).

That is about 230 words, and it sits in the top level — not because it holds more facts than a mid-level answer, but because it names a criterion, weighs two mechanisms and decides.

The sentences that move you up a level

Keep three stems in your head and you will never freeze at the end of an answer:

• "On balance, judged on ..., the more important factor is ..., because ..."
• "This matters most in this case because the question specifies ..."
• "X is stronger in the short term, but over ... years Y wins, because ..."

Each one contains the word because and a condition. Generic endings like "both are important in different ways" read as a refusal to answer the question. Examiner reports repeat the same finding every year: the strongest answers spot the command word and then visibly do what it asks.

Timing: what a 9-marker actually deserves

Work at roughly 1 to 1.2 minutes per mark, thinking time included.

• 9 marks: about 10 to 11 minutes, 220 to 280 words, one developed point per side plus the decision.
• 12 marks: about 14 to 15 minutes, two developed points per side, and a conclusion that ranks them against each other.

Spend the first minute writing a four-word plan in the margin — side A, side B, criterion, decision. I lose far more marks to a rushed ending than to a thin opening, and the plan is what fixed that for me.

Mistakes that quietly cap you mid-level

• No judgement at all, or one bolted onto the last line with nothing supporting it.
• Fence-sitting. "Both are important", with no criterion behind it.
• Vague evidence. "Lots of people" and "a big cost" score nothing. Name it, number it.
• Ignoring the context you were handed — a specific place, business, period or person. Applying to that context is an entire assessment objective.
• Writing everything you know. Under levels marking an irrelevant paragraph does not merely fail to add marks, it dilutes the answer the examiner is banding.

The fastest fix is feedback on your own writing, not more reading about technique. Write one real evaluate answer, then run it through Mark my answer to see which part cost you the level — it is nearly always the judgement. Tighten the argument with Essay coach, check the underlying content is accurate with Explain anything, then drill it under time pressure in a Mock exam.

Test yourself

1. A 9-mark question says evaluate. You have written two sides brilliantly and stopped. Roughly what fraction of the marks is still sitting on the table, and why?
2. Turn this into a real judgement: "Both advertising and better ingredients are important for a new cafe."
3. What is a criterion, and why does naming one lift an answer a whole level?

For the level-marking system in full, see how to answer 6-mark science questions.

FAQ

Do I have to argue both sides?

Yes. An answer that only argues one way is not weighing anything, so it cannot reach the top level even when every fact in it is correct. Give the other side genuine strength first, then defeat it with evidence.

Should the judgement go at the start or the end?

Either works, and the best answers often do both — a short verdict early, evaluation running through the middle, then a firmer decision at the end. What loses marks is judging nowhere, or judging only in a final sentence that no earlier paragraph supports.

Is a longer answer a better answer?

No. Levels marking rewards quality of reasoning, not volume. One point explained fully and judged in context beats four points listed and left hanging, and over-writing steals time from the rest of the paper.

Does this work outside Geography and Business?

Yes. Any subject with extended writing uses the same weigh-then-judge logic — history, PE, religious studies, English literature, and the science papers that set evaluate questions. Only the evidence changes.

In short

Evaluate means decide. Two sides, specific evidence, a mechanism for each, then a judgement that names what you are judging against and why it points one way. If you are running out of time, write the decision sentence first — it is worth more than another paragraph of everything you know.