Feedback as Process_Data

This study examined how AI, peer, expert, and instructor-guideline feedback relate to students’ scientific writing and reasoning in undergraduate biology. Data from 104 students included rubric-scored essays, scientific reasoning assessments, and written reflections. Final essay performance was similar across conditions, with feedback type explaining little variance. However, scientific reasoning showed domain-specific differences with small to moderate effects and some ceiling effects. Qualitative results revealed distinct engagement patterns: guidelines supported task understanding, peer feedback fostered evaluative judgment, AI supported organization, and expert feedback supported discipline-aligned revision. Overall, feedback influenced interpretation and revision processes more than final performance outcomes.