Methodology · Updated 11 May 2026

What we see in OET letters now, and how we adjusted.

The OET criteria have not changed since 2018, but since 2024 our predictions have varied more against official results than they did in 2014–2023. Here is what we see in the letters we correct, what we recalibrated, and how we report grade predictions you can actually act on.

11,000+ letters reviewed since 2014 Calibrated against student result reports Reviewed by Dr Mariam, PhD in English

Why this page exists

If you have failed OET writing and the prep service you used predicted a pass, you are not alone. We hear from students every week whose practice correction said B and whose official result came back C+ or below. The honest reasons for that gap are rarely explained to the student.

There are three causes — and we will lay them out plainly so you can judge for yourself whether the service you are using accounts for them.

OET marking trends shift over time. A service calibrated in 2021 will systematically drift from 2026 results.
Even trained markers can differ by one band on Genre & Style or Conciseness & Clarity for the same letter. That is normal rubric application, not a flaw — but it must be reflected in how grade predictions are reported.
The official examiner may weight a single criterion decisively. Purpose at 2/3 instead of 3/3, or a Content score of 4 instead of 6, can move a letter from B to C without changing the surface text much.

These three causes do not go away. What changes is whether the correction service you choose accounts for them — or pretends they do not exist.

What we see in 2024–2026 letters

Three patterns have shown up consistently in our cohort data from 2024 onwards. Each one corresponds to one of the six official OET writing criteria. We see them in submitted result reports, in students returning after a retake, and in the variance between what a careful reading predicts and what the exam actually returns.

Pattern 1

Conciseness & Clarity often costs a grade

Wordy openings, hedged closing requests, and irrelevant biographical detail ("retired teacher", "lives with wife") are among the most frequent reasons a letter loses marks on this criterion.

Pattern 2

Clinical relevance under Content matters

Including information the receiver does not need costs marks. A physiotherapy referral that lists every comorbidity loses more marks than a tighter letter that names only what is clinically relevant.

Pattern 3

Genre & Style shows variance between markers

The same letter sometimes scores 5 and sometimes 6 on Genre & Style. Borderline letters can sit on a grade boundary, which is why we report a range rather than a single grade.

What we changed in our methodology

Spotting the pattern is the easy part. Adjusting how you report grades is the part most services skip. Here is what we did.

We re-anchor the rubric against every cohort of student result reports

When a student returns an official OET result, we map it against the prediction we made on their last submitted letter. Patterns across cohorts feed into the next month's marking calibration. This is the feedback loop that keeps the service from drifting.

We report a range — not a single grade — when a letter sits in a volatile zone

On letters that fall close to a grade boundary, you receive a prediction like "C+ to B-" along with a sentence on which criteria are most likely to swing the outcome. A confident single-grade prediction on a borderline letter is dishonest. A range you can plan against is useful.

We name the criterion most likely to cost you the grade

Every correction now flags the single criterion where a half-band adjustment would change the result. If Conciseness & Clarity is the swing, you focus your retake practice there — not on the criteria that are already strong.

We do not promise pass guarantees

No prep service can guarantee a result, and any that does is misrepresenting how the exam works. What we can promise is that the feedback you receive is calibrated against the latest OET marking patterns and reviewed by a named expert — not generated by AI and not based on a 2021 rubric.

Who actually marks your letter

Every correction at Writing Correction Service is reviewed by Dr Mariam or her trained OET writing team.

PhD in English — formal academic background in the language being assessed.
20+ years preparing students for English-language examinations, including IELTS and OET.
~7 years specifically focused on OET writing.
11,000+ OET letters reviewed since 2014 — the data set behind our calibration methodology.

We will not call our team "certified OET examiners" because that is not a public credential anyone can verify. The credentials above are. They are the basis of the marking you will receive.

Get a grade you can act on

A single corrected letter from $15. Detailed feedback against all six OET criteria, a calibrated score range when the letter sits on a boundary, and the specific criterion most likely to swing your result. Returned within 24, 48 or 72 hours.

Submit a letter — from $15 See pack pricing

No pass guarantees. Honest grading, calibrated against the latest OET marking patterns. Reviewed by Dr Mariam or her trained OET writing team.

Frequently asked questions

Has OET writing marking changed in 2026?

The criteria and their descriptors have been unchanged since August 2018, and OET has announced no change to how Writing is marked. What candidates feel as strictness is usually Purpose, selection of content and conciseness, which the descriptors have always required. In the letters we correct, Conciseness & Clarity and Genre & Style are the criteria that most often cost a grade, and results on letters of similar quality can vary between attempts.

Why do OET grade predictions sometimes not match the official result?

Three reasons. First, every prediction is an estimate against a published rubric — even trained markers reading the same letter can differ by one band on some criteria. Second, the official exam marker may weight a single criterion (often Purpose or Conciseness & Clarity) more decisively than a prep-service marker. Third, a service that never compares its predictions with official result reports will drift away from real outcomes. We recalibrate against the latest result reports submitted by our students and give a range — not a single grade — when a letter sits in a volatile zone.

How accurate are your OET grade predictions?

Until 2024, the median variance between our predicted scaled score and the official OET result across our submitted students was within roughly ten score points (out of 500). Since 2024 the variance has widened, so we now explicitly report a range (e.g. 'C+ to B-') on letters that fall near a grade boundary, and we explain the specific criteria that are most likely to swing the outcome. This is more useful for retake planning than a single confident-sounding grade that turns out wrong.

What patterns do you see in OET letters in 2024–2026?

Three patterns have emerged consistently in our 2024–2026 cohort data. (1) Conciseness & Clarity often costs a grade — wordy openings, hedged closing requests, and irrelevant biographical detail are the most frequent causes. (2) Clinical relevance under Content matters — including information the receiver does not need costs marks. (3) Genre & Style shows variance between markers, with the same letter sometimes scoring 5 and sometimes 6. We have adjusted our marking methodology to reflect these patterns.

How do I know my correction reflects current marking, not old patterns?

Every correction at Writing Correction Service is marked by Dr Mariam (PhD in English, 20+ years of exam-prep experience) or her trained writing team, working from a marking rubric we update after every cohort of student result reports. Our methodology is described on this page. If your predicted grade does not match the official result, we ask you to share the official report so we can recalibrate — this feedback loop is how we have stayed accurate for over a decade.

Stuck below Grade B on OET writing? Dr Mariam's expert OET writing correction service marks every letter against all six OET writing criteria and returns line-by-line corrections within 24–72 hours.

See how it works