ETS's own AI scores your real essay
Most students do not know this: every GRE AWA essay is scored by both a human rater and ETS's proprietary AI system called e-rater. If the two scores differ by more than one point, a second human rater is brought in. If they agree, the average is your score.
e-rater evaluates essays on criteria including vocabulary sophistication, grammatical complexity, sentence variety, and argument organisation — which are the exact same things ChatGPT evaluates when you ask it to score your essay. This is why AI feedback for AWA preparation is unusually accurate compared to other exam sections.
The essay structure that scores 5.0+
The GRE Issue essay has one task: state a position and defend it with specific, relevant examples. The 5.0+ structure that ETS's own rubric rewards is consistent:
- Introduction (2–3 sentences): Acknowledge complexity of the issue. State your position clearly. Preview your main reasoning.
- Body paragraph 1 (4–5 sentences): Strongest supporting argument with a specific real-world example — historical, scientific, or personal. Concrete examples beat abstract claims every time.
- Body paragraph 2 (4–5 sentences): Second supporting argument with another specific example from a different domain than paragraph 1.
- Body paragraph 3 — concession (3–4 sentences): Acknowledge the strongest counter-argument. Refute it or qualify it. This is what separates 4.5 from 5.0+.
- Conclusion (2 sentences): Restate position in new language. One broader implication or condition.
The ChatGPT prompts that build AWA score
3 mistakes that cap AWA at 4.0
| Mistake | What it looks like | How to fix it |
|---|---|---|
| Vague examples | "Many companies have benefited from innovation." No names, no specifics. | Name specific companies, events, or people. "Tesla's pivot to over-the-air software updates in 2012 demonstrates..." is instantly stronger. |
| No counter-argument | Essay only presents the side you agree with. Reads as one-dimensional. | Dedicate one full paragraph to the strongest opposing view. Acknowledge it genuinely before refuting it. |
| Mechanical transitions | Every paragraph starts with "Furthermore," "Moreover," "In addition," | Vary your transitions: "This pattern becomes clearer when..." / "A contrasting perspective emerges in..." / "What makes this especially significant is..." |
Get the complete AI guide — free
50+ prompts, study calendars, and strategies for IELTS and GRE. Sent directly via DM.
Frequently asked questions
No — AWA is reported separately from the 260–340 composite score. Most programs require a minimum AWA (typically 4.0) but it does not add to or subtract from your Verbal and Quant scores. However, for competitive programs and fellowships, a 5.0+ AWA is a differentiator that signals writing ability beyond what the composite score shows.
450–550 words is the optimal range. ETS research shows that essays in this range consistently score higher than shorter essays, while essays over 600 words tend to be less focused and score no higher. In 30 minutes, 500 words is very achievable — that is roughly 17 words per minute.
No — and this actively hurts you. The real exam requires you to write under time pressure without any assistance. If you memorise AI-generated essays, you show up unprepared for the blank page. Use ChatGPT to score YOUR essays and fix YOUR weaknesses. That is what builds the actual skill.
Most students see significant improvement after 10–15 timed practice essays with AI feedback. That is 2–3 essays per week for 5 weeks. The improvement curve is steep — essays 8–10 are usually dramatically better than essays 1–3 for students who review feedback seriously.