ESL writing rubrics that actually work for ELL students

My first ESL writing rubric, made for a grade-7 class in Saigon, was an embarrassment in retrospect. It had four columns, each labelled "Excellent / Good / Adequate / Needs improvement," and almost every comment I wrote in the box translated to either "watch your verb tenses" or "good effort." The student writing wildly improved that year, but I am not sure my rubric had anything to do with it. The rubric was a scoring shell. The teaching happened in conversation.

I have built a lot of rubrics since. Most of them have been bad. The good ones share three things, and the bad ones share three different things, and once I started noticing the pattern I could not unsee it. This post is the thing I wish someone had handed me on day one in Korea. Three CEFR-aligned writing rubrics you can defend in a parent meeting, with the reasoning that goes with them.

The fluency-vs-accuracy trap

Most ESL writing rubrics over-weight grammar and spelling. Some of that is honest. Many ESL programmes are explicitly working on accuracy. Some of it is laziness: grammar errors are easy to count, ideas are hard to assess. The problem is what happens to your student.

Icy Lee’s longitudinal case study of an ESL writing programme found that mandatory form-focused corrective feedback combined with grammar-heavy rubrics produced what she calls "hyperfocus on accuracy." Students improved at surface-level grammar, but they also stopped taking risks with their ideas. They wrote shorter, safer essays. They valued the marginal comments more than the rubric scores, because the rubric scores told them only that they had failed to be perfect. (Lee & Mak, Assessing Writing)

That should worry any ESL teacher. The whole point of a rubric is to be a teaching tool, not just a scoring instrument. If your rubric is making students write less, the rubric is wrong. The fix is not to drop accuracy, it is to make sure each criterion on the rubric is doing its share of the work and no more.

The four pillars I now use, no matter the level

After a lot of trial and error I settled on the same four-criterion structure that IELTS, Cambridge, and most reputable ESL exam boards use. The names vary, but the substance does not. Here is the version I use in my own rubrics:

1. Task response. Did the student do what was actually asked? Answered the question, hit the word count, addressed all parts of the prompt, used the genre that was requested. This is the criterion most often missing from home-grown rubrics, and the most important one.
2. Coherence and cohesion. Are the ideas in an order a reader can follow? Are paragraphs structured? Are connectors used appropriately, or mechanically? The IELTS public band descriptors call Band 9 cohesion writing that "uses cohesion in such a way that it attracts no attention" (British Council, 2023), which is a beautifully simple bar.
3. Lexical range and accuracy. Is the vocabulary appropriate to the level, used precisely, and varied where it counts? This is where you score word-choice mistakes, but also reward students for taking risks with new vocabulary.
4. Grammatical range and accuracy. Same logic as lexis but for structure. Did they attempt complex grammar? Did they get most of it right? Are the remaining errors interfering with meaning, or just minor slips?

Notice that two of the four criteria are about content and structure, and two are about language. That balance is the whole game. It is what stops the rubric from becoming a grammar-policing instrument.

What the CEFR descriptors actually say at B1, B2, and C1

For my own rubrics I anchor each level’s descriptors to the Council of Europe’s 2020 CEFR Companion Volume, which remains the live standard for 2026 (no revision announced). Here is the overall written-production descriptor for each level, paraphrased for parent-meeting language:

• B1. Can write straightforward connected texts on familiar topics within their field of interest, linking shorter discrete elements into a linear sequence. Translation: "the student can write a clear paragraph or two about something they already know about."
• B2. Can write clear, detailed texts on a variety of subjects related to their field, synthesising and evaluating information and arguments from a number of sources. Translation: "the student can write a structured essay that uses ideas from more than one place."
• C1. Can write clear, well-structured texts on complex subjects, underlining the salient issues, expanding and supporting points of view at length with subsidiary points, reasons and relevant examples, and rounding off with an appropriate conclusion. Translation: "the student can write a real essay, with a thesis, body, and conclusion, on a topic they have to think hard about."

Source for the original descriptor language is the Council of Europe Companion Volume 2020, which is the document I would actually quote from in a parent meeting. The translations above are my own.

Three rubric templates, level by level

Each template uses the same four pillars. The descriptor for each band changes per level. Score each criterion 1 to 5; total out of 20. The descriptors below are for the top band of each rubric (i.e. the version that earns full marks). Bottom band reads as "did not attempt" or "interfered with meaning throughout."

B1 rubric: writing for first-year MYP, lower-secondary, or pre-intermediate

• Task response (5). Addresses every part of the prompt; word count met; uses the genre asked for (description, narrative, instructions, etc).
• Coherence and cohesion (5). Ideas are in a logical order. Uses paragraphs. Basic connectors (and, but, because, then, so) used correctly.
• Lexical range (5). Uses topic-specific vocabulary from class accurately. Word choice generally matches intended meaning. Errors do not impede understanding.
• Grammatical range (5). Uses present, past, and future tenses appropriately. Subject-verb agreement is mostly correct. Errors occur but do not obscure meaning.

B2 rubric: writing for upper-secondary, IB MYP year 4-5, or intermediate ESL

• Task response (5). Addresses the prompt thoroughly. Develops a position with reasons and at least one extended example. Genre conventions observed.
• Coherence and cohesion (5). Clear paragraph structure with topic sentences. Uses a range of connectors (however, although, therefore, for example). Reader can follow without re-reading.
• Lexical range (5). Vocabulary varied, with attempts at subject-specific or formal lexis. Word-choice errors rare and do not affect meaning. Some idiomatic use attempted.
• Grammatical range (5). Uses complex sentences (relative clauses, conditionals, passive voice) at least once each. Errors with these are present but do not impede communication. Tense control consistent.

C1 rubric: writing for IB DP, A-level, or advanced ESL

• Task response (5). Addresses every part of a complex prompt. Develops a clear thesis with multiple supporting points and counter-arguments where appropriate. Conclusion does meaningful work.
• Coherence and cohesion (5). Sophisticated paragraph structure; ideas linked across paragraphs as well as within. Cohesive devices used naturally rather than mechanically.
• Lexical range (5). Wide range of precise vocabulary, including academic and topic-specific lexis. Effective use of collocations and figurative language where appropriate. Errors limited to slips.
• Grammatical range (5). Confident use of a wide range of complex structures. Errors are rare and minor. Style is appropriate to the genre and register asked for.

What to do in a mixed-level classroom

The reality of every ESL classroom I taught was that I had B1, B2, and the occasional runaway C1 student in the same room. I cannot count the number of Sundays I lost photocopying two versions of the same worksheet. The fix turned out to be using one rubric and varying the prompt complexity, not varying the rubric.

A 2025 Frontiers in Education study on rubric use in EFL classrooms makes the same point: rubrics "improve the learning process, especially in formative assessments" when shared with students before the task, but only if the rubric is consistent enough across the class for students to compare their own work against it (Frontiers in Education, 2025). Use the same rubric, vary the bar.

Practically: I usually picked the rubric for the highest-level student in the class and made the prompt accessible enough that everyone could attempt it. The B1 student scored 8/20 on the C1 rubric. That was honest. They were a B1 student writing on a C1 prompt. Their growth across the year was tracked on the same rubric, which is the only thing that lets you see progress.

Show students the rubric before they write

The cheapest, highest-impact change you can make is to share the rubric with students before they start writing. Icy Lee’s 2024 paper in ELT Journal on co-constructing rubrics with students put it more strongly:

Allowing students to view rubrics before composing their assignments improves writing performance and achievement. (Lee, ELT Journal 2024)

It also makes your grading defensible. When you score a student a 3 on Coherence and Cohesion, they have already seen what a 5 looked like before they started. The grade is not a surprise; it is a calibration.

The piece I cannot solve for you

No rubric, no matter how thoughtfully built, will help you grade thirty essays consistently on a Sunday night. That is a different problem, and it is the one I eventually built software to solve. The rubric is the input; consistency across the batch is the output. If your rubric is correct, the AI grader using it can be honest about where each student lands. If your rubric is the four-column "Excellent / Good / Adequate / Needs improvement" version I used in Saigon, no AI in the world will save you, because there is nothing to anchor on.

Get the rubric right, and the rest is execution.

Sources cited above: Council of Europe (2020) CEFR Companion Volume; British Council/IDP/Cambridge IELTS Writing Band Descriptors (public version, 2023); Cambridge University Press & Assessment, C1 Advanced Handbook for Teachers (2024); Lee & Mak, "Rubrics and corrective feedback in ESL writing," Assessing Writing; Lee, "Implementing rubric co-construction in ESL writing teaching," ELT Journal 2024; Frontiers in Education (2025), "The analytic use of rubrics in writing classes by language students in an EFL context."