Essay Writing Guide
A Scholarship essay is a sustained argument, built under time pressure, in response to a question you've never seen. This page breaks that skill into trainable parts.
1. How to unpack a question
Spend the first five minutes on the question itself. Most failed Scholarship essays fail here — they answer a question that wasn't asked.
- Circle the instruction word. Evaluate, discuss, to what extent, how far do you agree — these demand a weighed position, not a survey.
- Define the contestable terms. If the question says "media power", "authenticity", "the audience", or "democratised" — those words are doing heavy lifting, and your essay should briefly stake out what you take them to mean. Defining terms is argument.
- Find the tension. Scholarship questions almost always contain a built-in conflict (freedom vs control, global vs local, art vs commerce). Name the tension explicitly — it gives your essay its spine.
- Ask "compared to what? for whom? since when?" A claim like "media has never been more powerful" hides comparisons. Surfacing them generates instant analytical material.
- Choose your ground. Decide which concepts and examples from your preparation genuinely serve this question — and let go of the ones that don't, however much you love them.
2. How to build an argument
An argument is a position plus the reasoning that earns it. To build one quickly in the exam:
- Take a side, with nuance. "Yes, but with two important limits" is a stronger position than either pure agreement or fence-sitting. Nuance is not indecision — it's precision about where and how far your claim holds.
- Plan paragraphs as moves, not topics. Each paragraph should advance the argument: establish, extend, complicate, counter, resolve. If two paragraphs could be swapped without loss, they're topics, not moves.
- Build in the counter-argument. Dedicate material to the strongest case against your position, then show why your position survives it (or how it must be modified). This single habit separates Scholarship scripts from Level 3 scripts.
- Drive towards a destination. Know your final sentence's idea before you write your first. Everything en route should be pulling towards it.
3. How to write a thesis
Your thesis is your answer to the question, compressed into one or two sentences, positioned at the end of your introduction. A Scholarship thesis is specific, arguable, and weighted — someone informed could disagree with it, and it signals how your argument will be qualified.
✗ Weak thesis
"Social media has both positive and negative effects on society, which this essay will discuss."
Why it fails: no position, no specificity, announces rather than argues. Could open any essay ever written.
✓ Strong thesis
"Social media has redistributed the power to speak while concentrating the power to be heard: participation has been democratised, but visibility is now governed by engagement-driven algorithms whose logic favours platforms over publics."
Why it works: takes a side, contains a distinction (speaking vs being heard), names a mechanism, and forecasts the essay's structure.
4. The paragraph framework
Use this reusable six-part shape — P‑D‑M‑T‑E‑L — until it becomes instinct. Not every paragraph needs all six in this order, but every paragraph needs most of them.
- Point — one claim that advances your thesis. ("Algorithmic curation has inverted the traditional gatekeeping model.")
- Development — unpack the claim: define, explain the mechanism, draw the distinction. (How does an algorithm's gatekeeping differ from an editor's?)
- Media evidence — a specific example, precisely described and analysed, not retold. (A concrete case of algorithmic amplification or suppression.)
- Theory/concept — the tool that deepens the analysis. (Hall, Castells, gatekeeping theory — applied to this evidence.)
- Evaluation — the "so what": significance, limits, who's affected, how much it matters. (Is this a marginal shift or a structural one? What's the counter-case?)
- Link back — close the loop to the question's actual words and your thesis.
Timing guide: in a planned response, aim for four to six of these paragraphs plus introduction and conclusion. Depth beats coverage: one fully developed P‑D‑M‑T‑E‑L paragraph outscores three descriptive ones.
Test a paragraph now: paste one into the Paragraph Coach — it checks instantly for each of the six moves and flags what's missing, so you can self-edit before anyone marks it.
5. How to integrate media examples
- Be precise. Name the text, the scene, the campaign, the platform decision, the date if it matters. "In the opening sequence of…" beats "In many films…" every time.
- Analyse, never retell. Assume the examiner knows the text or doesn't need the plot. Your sentences should be about how it works and what it shows, not what happens.
- Make the example earn its place. Before writing it, finish this sentence: "This example proves my point because…" If you can't, swap it.
- Range signals breadth. Across a whole response, draw from different forms (film + platform + journalism + your own production), different scales (global + Aotearoa), and different eras where relevant.
- One example, many uses. A deeply known example can serve representation, industry, audience, and technology questions — if you know its production context, reception, and form, not just its content. Build this in your Example Bank.
6. How to use theory without dumping it
Theory dumping — summarising a theorist for half a paragraph before mentioning any media — is the most common weakness in nearly-there scripts. The fix is structural:
- Lead with the example, bring theory in to explain it. "When the trailer reframes the female lead for its YouTube thumbnail, Mulvey's gaze is operating at the level of marketing, not just cinematography" — theory and evidence in one sentence.
- Use the theory's idea, not its biography. The examiner doesn't need the year, the book, or the theorist's life story. They need the concept doing analytical work.
- Test it. "X explains A but struggles with B" is worth more than perfect agreement. See the Theory Toolkit for ready-made tensions.
7. How to compare examples
Comparison is an evaluation machine: differences between cases generate judgements automatically. To compare well:
- Compare on a dimension, not in general. Pick the axis: funding model, representation strategy, audience relationship, formal conventions. "X and Y differ in how they monetise attention" gives you a paragraph; "X and Y are different" gives you nothing.
- Use contrast to isolate causes. Two texts in the same genre with different funding (public vs commercial) let you argue what funding does. Same story on two platforms lets you argue what form does.
- Let the comparison conclude something. End with what the difference means for the question — don't leave the two cases sitting side by side.
8. How to evaluate significance
Evaluation is judgement with criteria. Whenever you make a claim, run it through these questions and write the answers into your paragraph:
- Scale — does this affect a niche, a nation, or the whole media ecosystem?
- Depth — is this a surface change (new label, same logic) or a structural one (new power relations)?
- Distribution — who gains, who loses, who is unaffected? Effects are rarely uniform.
- Durability — trend or blip? Would this claim have been true ten years ago? Will it be true in ten?
- Counterweight — what is the strongest fact pulling the other way, and does your claim survive it?
9. Conclusions that do more than repeat
A conclusion that restates the introduction wastes your final impression. Instead, in roughly this order:
- Deliver the verdict — your weighed answer to the question, now carrying the accumulated force of your evidence. It should be sharper than your thesis, not a photocopy of it.
- Acknowledge the strongest qualification — the condition under which your answer would change. This reads as intellectual honesty, because it is.
- Widen the frame — why this question matters beyond the exam: what it implies for audiences, democracy, culture, or what's coming next. One or two sentences; no grand clichés about "the digital age".
10. Weak vs strong: paragraph models
✗ Weak paragraph (description + theory dump)
"Stuart Hall came up with the encoding/decoding model in 1973. It says producers encode meanings and audiences decode them in three ways: preferred, negotiated and oppositional. An example of this is reality TV. Reality TV shows are encoded by producers and decoded by audiences in different ways. This shows that audiences are active. Another theorist is Henry Jenkins, who talks about participatory culture…"
Problems: biography instead of application; example named but never analysed; "this shows" claims with no shown mechanism; second theorist arrives before the first has done any work.
✓ Strong paragraph (P‑D‑M‑T‑E‑L)
"Audience power is real, but it operates within boundaries set by producers — and increasingly by platforms. (Point) The question is not whether audiences make their own meanings, but how much those meanings can cost the producer. (Development) When a reality franchise's editing constructs a contestant as a 'villain', viewers routinely contest the edit — fan accounts re-cut footage to expose manipulative sequencing, generating oppositional readings at scale. (Media evidence) This is Hall's decoding made visible and collective: the preferred reading is not just refused but publicly dismantled. (Theory) Yet the franchise profits from precisely this contestation — outrage drives engagement, and engagement is the product. Audience resistance, in platform economics, is just another form of attention. (Evaluation) Audience interpretation is free; its consequences are owned — which is why 'active audiences' have not translated into audience power over what gets made. (Link)"
11. Common mistakes checklist
Run every practice essay against this list. Print this section and keep it beside you while writing.