How Writing Is Scored

How does partial credit work in PTE Writing?

Every free-text Writing task, Summarize Written Text, Write Essay, Write Email, Summarize Spoken Text, is partial credit scored: you are not marked right or wrong as a whole, and you don't need a flawless response to earn points. Each response is broken into several traits, each with its own point range, and your task score is the sum of what you earn across all of them.

• Content: How well your response addresses the prompt and covers the relevant points. Scored first, and it gates everything else, see the next section.
• Form: Whether your response falls inside the task's word-count range and follows its formatting rules (no all-caps, no bullet points standing in for sentences). Scored second, and it also gates everything else.
• Grammar and Vocabulary: Present on every free-text Writing task in both exams, but the band wording isn't identical: Summarize Written Text and Summarize Spoken Text judge whether errors are absent, present but not disruptive, or severe enough to hinder communication, while Write Essay's Grammar rewards consistent control of complex language, and Write Email's Grammar and Vocabulary are worded around the shorter, everyday register the task calls for.
• Task-specific traits: Write Essay adds Development, Structure and Coherence, General Linguistic Range and Spelling; Write Email adds Email Conventions, Organization and Spelling; Summarize Spoken Text adds its own Spelling trait on top of Content, Form, Grammar and Vocabulary. These task-specific traits are covered in the lessons for each task; see Development, Structure and Coherence and Email Conventions.

Pearson's Score Guides don't publish a single headline maximum for each task, so the figures above are the sum of that task's own trait maxima, not a number quoted directly in either guide.

In practice this means a task score is a sum, not a single judgement call. A Write Essay response, for example, might earn 5 out of 6 on Content, full marks on Form, 5 out of 6 on Development, Structure and Coherence, full marks on Grammar, 5 out of 6 on General Linguistic Range, and full marks on Vocabulary Range and Spelling, seven separate numbers added together, not one overall impression rounded to a grade.

Why can one weak trait zero out an entire response?

Content is always scored first. If there is no response, or the response is off-topic, Content scores 0. Only if Content scores above 0 is Form scored next, and only if both Content and Form score above 0 does the response go on to be rated on its remaining traits. Pearson's own Score Guides state the consequence plainly: when Content or Form scores 0, no score points for the response are given at all.

1. Content is scored first. No response, or a response that is off-topic (including pre-prepared material), scores 0.
2. If Content scores above 0, Form is scored next: word count, capitalisation and punctuation.
3. Only if both Content and Form score above 0 do the remaining traits, Grammar, Vocabulary, and any task-specific traits, get scored at all.

Those are Pearson's own worked examples for each trait. The Form band definitions themselves also list writing in capital letters, using no punctuation, or submitting bullet points instead of sentences as further ways to score Form = 0, the same zero-scoring gate applies regardless of which of these causes it.

This gate applies to every free-text Writing task in both exams: Summarize Written Text, Write Essay, Write Email and Summarize Spoken Text. It does not apply to Write from Dictation or the Fill in the Blanks variants, including Reading and Writing: Fill in the Blanks, which have no Content/Form gate at all: those are scored purely per word or per blank, so a single wrong word costs only that word's point, never the rest of your answer.

Can you use a memorised template?

Pearson's Score Guides state that responses containing pre-prepared or memorised material are classified as an irrelevant response, the same category as writing about the wrong topic entirely. Core's Write Email Content band spells out the risk directly: a response scores at the bottom of the Content scale where it "may not deal properly with the prompt due to significant amounts of pre-prepared/memorized material." A response built mainly from memorised sentences risks that same Content failure, which triggers the zero-scoring gate covered above: no points for the task at all, regardless of how polished the language is.

Is PTE Writing scored by AI, by a person, or both?

Pearson describes its Writing scoring as automated, trained and checked against human expert ratings, the same live human expert review model it uses for Speaking. The current Academic Score Guide names the tasks that additionally get this review on top of automated scoring: Summarize Written Text, Write Essay and Summarize Spoken Text, three of the four Academic Writing tasks. Write Essay gets the most thorough review of all, with an added human check on Development, Structure and Coherence and on General Linguistic Range, not just Content.

Treat this human-review list as a description of current practice rather than a claim about what changed and when. Pearson's own update notes mention the hybrid approach being extended to more question types without naming exactly which tasks moved. The Core Score Guide does not publish an equivalent list naming which Core Writing tasks get this same live review, so don't assume Core works identically to Academic here; write and proofread every response as if it will be checked closely either way.

None of this changes what a score report looks like: both exams still return an overall score plus a score for each of the four communicative skills, all on the same 10–90 scale, the automated-plus-human process behind the scenes doesn't add extra numbers for you to interpret.

Does mixing British and American spelling cost you points?

Both handbooks give test takers the same instruction: you may write in any valid form of English, British or American, but it must be consistent throughout the test, not just within one response. Neither Score Guide names a separate trait for spelling-convention inconsistency, so the practical guidance is to simply pick one convention on test day and stick to it in every Writing task, rather than to try to predict exactly how a mixed response would be marked.

Where a task does have a dedicated Spelling trait, Write Essay, Write Email and Summarize Spoken Text, it is scored by counting actual misspellings, covered in detail in Vocabulary & Spelling.