How to structure a first-class essay

Most students who are stuck on a 2:1 are not stuck because of their writing style. They are stuck because their essays read as a tour of the topic rather than as an argument about it. The marker finishes the last page knowing what you read, but not what you think, and there is no band descriptor anywhere in the UK that awards a first for coverage alone.

Structure is what fixes that, and structure is learnable in an afternoon. This guide walks through the shape of a first-class essay from the argument outwards: how to decide what you are claiming, how to build paragraphs that carry weight, how to sequence them so each one earns the next, and how to run a 60-minute structural edit before you submit.

Start with the argument, not the introduction

The single most expensive habit in undergraduate writing is opening a blank document and typing the introduction. You cannot introduce an argument you have not made yet, so what comes out is background: a definition, a bit of history, a sentence about how the topic is 'widely debated'. Two hundred words gone and nothing claimed.

Instead, start by answering the question in one sentence, out loud, badly. If the question is 'To what extent did the 1834 Poor Law Amendment Act reflect Benthamite thinking?', your first answer might be 'Mostly, but the administrative bits more than the moral bits.' That is unpolished, and it is already an argument — it has a position, a qualification and an implied structure.

From there you are no longer writing about a topic; you are defending a claim. Every subsequent decision — which sources to use, which counter-argument to concede, how long the historiography section should be — becomes easy, because you can ask a single question of each one: does this help me defend the claim?

The one-sentence thesis test

A usable thesis has three properties: it is contestable, it is specific, and it is provable with the evidence you can actually reach. Run your sentence through all three. 'Social media affects political participation' fails the first test — nobody disagrees. 'Social media platforms increase the volume of political expression while narrowing its deliberative quality' passes: a reasonable, well-read person could argue against it.

Specificity matters because it constrains you usefully. A vague thesis lets you wander; a sharp one tells you what is off-topic. If your claim is about deliberative quality, the section you drafted on advertising revenue is interesting but irrelevant, and cutting it will raise your mark rather than lower it.

The third test is the one students skip. A thesis you cannot evidence in 2,500 words with the reading available to you is not ambition, it is a trap. Scale the claim to the word count. It is far better to prove a modest point completely than to gesture at a grand one.

Build paragraphs as evidence units, not PEEL boxes

PEEL — point, evidence, explain, link — is a serviceable school framework, but at degree level it produces flat paragraphs because it treats evidence as a single item to be quoted and explained. Strong paragraphs usually hold two or three pieces of evidence in tension and use the tension to make a point neither source makes alone.

A better shape is: claim, evidence, complication, resolution. State what this paragraph establishes. Give the evidence. Introduce the thing that does not fit — a counter-example, a methodological limit, a source that disagrees. Then resolve it in a way that leaves your claim standing but more precisely stated than it was at the start.

That final move is where the marks live, because it demonstrates judgement rather than recall. Notice too that the paragraph should end on your voice, not a citation. If your last sentence is a quotation, the reader's last impression is of someone else's thinking.

• Claim: one sentence, in your own words, that could stand as a bullet in your plan.
• Evidence: the specific study, case, dataset or text — named, dated and precise.
• Complication: what the evidence does not cover, or who reads it differently.
• Resolution: the sharpened version of your claim, plus one clause pointing forward.

Sequence sections so each one earns the next

Weak essays are ordered by topic; strong essays are ordered by dependency. Ask of every section: what must the reader already accept for this to land? Put that section first. The result is an essay that feels like it is going somewhere, because it is.

A reliable default for a 2,500-word argument is four body sections of roughly 500 words: establish the terms of the debate, make your strongest positive case, take the most serious objection seriously, then extend or qualify. Notice that the counter-argument sits third, not last — you want to answer it and then keep building, rather than ending on someone else's point.

Use signposting sparingly but deliberately. One clause at the start of each section that names the move ('Having established X, the harder question is Y') is worth more than a paragraph of 'this essay will now discuss'. Markers read quickly; they are grateful for a map, and irritated by a table of contents in prose.

Write the introduction last, and make it do four jobs

Once the body exists, the introduction takes fifteen minutes because you finally know what you are introducing. A good one does exactly four things and nothing else: it defines the problem in a way that makes it feel worth solving, it states your thesis explicitly, it declares the scope you have chosen, and it previews the route.

Scope declarations are underused and disproportionately rewarded. A sentence such as 'This essay considers only NHS commissioning in England after 2012, since devolved arrangements differ materially' converts an omission into a decision. Markers cannot penalise you for not covering Scotland if you have told them why you are not.

Resist the urge to open with a grand generalisation about human nature or the pace of technological change. Open with the specific problem. 'Whether an algorithm can hold a duty of care is now a live question in three jurisdictions' earns attention; 'Technology has changed the world' does not.

Conclusions that add value rather than repeat

A conclusion that summarises is a wasted 200 words. The marker has just read the essay. What they want at the end is the answer stated with the confidence the evidence now supports, plus the intellectual consequence of it being true.

So restate the thesis in sharper terms than the introduction could — you have earned the precision. Then do one of three things: name what follows for the wider debate, identify the specific evidence that would change your mind, or flag the limitation that most constrains your claim. Any of the three signals a writer who knows where the edges of their argument are.

Never introduce new evidence, and never end with an apology. 'Further research is needed' is only worth writing if you say what research, and what it would settle.

A 60-minute structural edit before you submit

Structural problems are invisible when you read your own draft in order, because you supply the missing logic from memory. The fix is to stop reading it as prose and start reading it as architecture.

Copy the first sentence of every paragraph into a new document, in order. Read only that list. It should read like a coherent argument in miniature. Where it jumps, you have found a missing step; where two lines say the same thing, you have found a paragraph to merge. This single technique catches more marks than any amount of sentence-level polishing.

• Extract every topic sentence into a list and read it on its own — does it argue?
• Highlight every sentence that is purely descriptive; if a paragraph is more than half highlighted, rewrite it.
• Check each paragraph ends in your voice rather than a quotation.
• Count words per section — a section at half the length of its neighbours is usually underdeveloped, not concise.
• Read the introduction and conclusion back to back; they should agree, with the conclusion sharper.
• Finally, check every claim that would need a citation actually has one.