2026/03/12 by Mei Tan, Lena Phalen, Dorottya Demszky · 2 voices
Computer Science · Decision Sciences · Psychology · Social Sciences · #Psychometric Methodologies and Testing #Reading and Literacy Development #Writing and Handwriting Education #cs.CL #cs.HC
paper · pdf · doi:10.1145/3785022.3785113
arxiv published 2026/03/12 · arxiv updated 2026/03/12 · openalex created_date 2026/03/18 · openalex publication_date 2026/04/25 · openalex updated_date 2026/07/29
Effective personalized feedback is critical to students' literacy development. Though LLM-powered tools now promise to automate such feedback at scale, LLMs are not language-neutral: they privilege standard academic English and reproduce social stereotypes, raising concerns about how "personalization" shapes the feedback students receive. We examine how four widely used LLMs (GPT-4o, GPT-3.5-turbo, Llama-3.3 70B, Llama-3.1 8B) adapt written feedback in response to student attributes. Using 600 eighth-grade persuasive essays from the PERSUADE dataset, we generated feedback under prompt conditions embedding gender, race/ethnicity, learning needs, achievement, and motivation. We analyze lexical shifts across model outputs by adapting the Marked Words framework. Our results reveal systematic, stereotype-aligned shifts in feedback conditioned on presumed student attributes--even when essay content was identical. Feedback for students marked by race, language, or disability often exhibited positive feedback bias and feedback withholding bias--overuse of praise, less substantive critique, and assumptions of limited ability. Across attributes, models tailored not only what content was emphasized but also how writing was judged and how students were addressed. We term these instructional orientations Marked Pedagogies and highlight the need for transparency and accountability in automated feedback tools.