FOXGLOVE: Understanding Goal-Oriented and Anchored Writing Feedback from Experts and LLMs on Argumentative Essays
FOXGLOVE compares feedback from trained human instructors and four frontier Large Language Models (LLMs) on argumentative essays. The study found that while both groups focus similarly on argumentative goals, they differ in which specific sentences to highlight and how urgent those points are. LLM feedback is often longer but sometimes less readable, suggesting human judgment remains vital for personalization.