FEEDTYPE: a benchmark of teacher and LLM feedback annotated for feedback focus across three writing domains¶
Product or Programme · All products and programmes
Evidence · 3 claims (2 for, 1 mixed) · 1 study (1 design),q2· 0 of 1 report an effect size · 3 claims rest on one study
Description¶
A released corpus of paragraph-level teacher- and LLM-generated writing feedback with feedback-focus annotations, scores, and student revisions across composition, ESL, and fiction courses, created by Norah Almousa et al.
Components¶
- FEEDTYPE: a benchmark of teacher and LLM feedback annotated for feedback focus across three writing domains: FEEDTYPE is a released corpus of paragraph-level teacher and LLM-generated feedback on student essays, drawn from the SEFORA corpus across composition, ESL, and fiction courses, with feedback focus type annotations, scores, and revisions. The authors state it is the only corpus that "jointly provides scores, paragraph-level feedback, feedback focus type annotations, and revisions across all three domains." It contains 1,966 annotated paragraphs and feedback from six LLMs under three prompting strategies. (Norah Almousa et al. (2026))
Claims¶
- LLMs cover most teacher feedback focus types but fail to match teacher feedback focus distributions, with LearnLM closest [+W]
- LLM feedback adaptivity to student performance occurs only in frequently produced focus types and only in large or closed models [~W]
- Gemini 2.0 Flash validated as a reliable LLM annotator of feedback focus types [+W]
Related Products and Programmes¶
- Estonian reflective-writing AI-feedback corpus of 2988 homework triplets from 283 students
- TACTCorpus and TACTBench
- Corpus of 500 annotated competency-based assessments in higher education for developing and evaluating the Teacher Evaluator
Key Sources¶
- Norah Almousa, Shayan Peyghambari Oskoui, Raquel Coelho, Gayle Rogers, Xiang Lorraine Li, Diane Litman. (2026). Evaluating Feedback Focus and Pedagogical Adaptivity in LLM-Generated Feedback on Student Writing. arXiv preprint. https://arxiv.org/abs/2609.28026
Page metadata
| Field | Value |
|---|---|
type |
product |
id |
feedtype-a-benchmark-of-teacher-and-llm-feedback-annotated-for-feedback-focus-across-three |
title |
FEEDTYPE: a benchmark of teacher and LLM feedback annotated for feedback focus across three writing domains |
description |
A released corpus of paragraph-level teacher- and LLM-generated writing feedback with feedback-focus annotations, scores, and student revisions across composition, ESL, and fiction courses, created by Norah Almousa et al. |
status |
review |
product_kind |
dataset |
generated |
by process:settle-candidates on 2026-10-11 |
trust |
unverified — no one has confirmed this page's content |
Frontmatter is the page's machine-readable record — see the Schema & Guide.