2.4 Analytical Writing Scoring Rubric (0–6) & e-rater

2.4 Analytical Writing Scoring Rubric (0–6) & e-rater

Quick Answer: The GRE Analytical Writing measure is a single 30-minute "Analyze an Issue" essay scored 0–6 in half-point increments. A trained human rater and the ETS e-rater automated engine each assign a score; if they agree, that score (or the average) is yours, and if they diverge meaningfully, a second human rater resolves it. A 4.0 sits at the 63rd percentile and a 5.0 at the 93rd, so most test takers land between 3.0 and 4.5.

How the AW Score Is Produced

The Analytical Writing measure is one task: a 30-minute "Analyze an Issue" essay. The "Analyze an Argument" task was removed September 22, 2023, so every GRE administration now begins with the same single essay prompt drawn from the published ETS Issue Topic pool.

Your essay is scored on a 0–6 scale in half-point increments (6.0, 5.5, 5.0, ..., 0.5, 0.0). Two readers score it independently:

1. A trained human rater assigns a holistic 0–6 score.
2. The ETS e-rater automated scoring engine independently assigns a 0–6 score based on essay features correlated with writing proficiency — grammar, usage, mechanics, development, organization, vocabulary, and fluency.

If the two scores are close (within 0.5 of each other), they are averaged to produce your final score. If they diverge by more than a point, a second human rater scores the essay and the final score is the average of the two human scores. ETS emphasizes that e-rater is not a grammar checker — it produces a holistic score reflecting overall writing quality in response to the task, the same construct the human rater evaluates. The e-rater engine is most reliable on the language-and-conventions dimension and least reliable on the analytical-substance dimension, which is exactly why ETS keeps a human rater in the loop.

The 0–6 Score Level Descriptions

ETS publishes anchored level descriptions in the Guide to the Use of Scores. Knowing them lets you self-score your practice essays honestly instead of guessing.

A score of 6 is described by ETS as sustaining "insightful, well-focused, well-developed analysis of the issue, conveying ideas skillfully." A 4 is "competent" — competent, not excellent. Most successful applicants to competitive programs score 4.0 or above; a 5.0 or higher is already exceptional, and a 6.0 is rare.

Percentile Conversion (July 2022 – June 2025 window)

The percentiles below are the official ETS values for July 2022–June 2025:

Note the steep climb between 4.0 (63rd percentile) and 5.0 (93rd). Half a point of AW score is worth roughly 20–30 percentile ranks in the middle of the distribution, so improving from a 3.5 (40th) to a 4.5 (85th) is the single most efficient percentile gain you can make on the entire GRE.

What the Issue Scoring Guide Actually Rewards

The ETS Issue task scoring guide explicitly looks for five things:

• Articulate a clear position on the issue — you must take a side, not merely describe the prompt.
• Develop with compelling reasons and persuasive examples — generic examples ("for example, in history...") do not score above 4.
• Support with logical analysis — examples must connect back to a thesis through reasoning.
• Acknowledge counterarguments where appropriate — one-sided arguments cap at about a 4.
• Control language and conventions — sentence variety, precise vocabulary, and freedom from errors that interfere with meaning.

Each of those five is weighted holistically — there is no checklist where two columns of vocabulary can rescue an undeveloped thesis. Substance and language carry roughly equal weight in the holistic judgment.

How to Self-Score a Practice Essay

Write a 30-minute practice essay from the published pool, then read it against the level descriptions:

1. Does it take a clear, defensible position? (If not, cap at 3.)
2. Are the reasons specific and the examples concrete? (If vague, cap at 4.)
3. Does it engage at least one counterargument? (If one-sided, cap at 4.)
4. Are there sentence variety and precise vocabulary? (If flat, cap at 4.)
5. Are there grammar errors that interfere with meaning? (If yes, cap at 3.)

The number where you stop is your honest self-score. Most test takers overestimate by one full point the first time they do this, so calibrate against the ETS-anchored language in the table above.

Strategic Implications

Because AW is reported separately from the 130–170 V and Q scores and is not section-level adaptive, the AW essay is a fixed-cost 30 minutes at the very start of the test. Spending those 30 minutes wisely — a clear thesis, two developed reasons with one concrete example each, and one acknowledged counterargument — buys a 4.5 (85th percentile) almost every time. Pushing to a 5.0 or 5.5 requires insight and language control that come from timed practice and honest self-scoring, not from memorizing longer templates.