eRevise+RF: A Writing Evaluation System for Assessing Student Essay Revisions and Providing Formative Feedback

TL;DR

The paper introduces eRevise+RF, an enhanced automated writing evaluation system designed to assess student revision quality and provide formative feedback on evidence usage and revision success within argumentative essays. The backend combines evidence-scoring indicators ($NPE$, $SPC$) with revision classifiers to guide feedback, while the frontend supports students, teachers, and administrators across three drafts aligned to a Response-to-Text Assessment. In deployment with 406 students across PA and LA, the system achieves substantial NLP evaluation performance and demonstrates measurable improvements in essay quality and revision effectiveness, including a demonstrated case study. The work advances AWE by integrating multi-draft revision analysis with actionable feedback, offering practical benefits for improving young students’ argumentative writing skills, and lays groundwork for broader deployments and future enhancements such as incorporating LLM-assisted feedback alongside expert guidance.

Abstract

The ability to revise essays in response to feedback is important for students' writing success. An automated writing evaluation (AWE) system that supports students in revising their essays is thus essential. We present eRevise+RF, an enhanced AWE system for assessing student essay revisions (e.g., changes made to an essay to improve its quality in response to essay feedback) and providing revision feedback. We deployed the system with 6 teachers and 406 students across 3 schools in Pennsylvania and Louisiana. The results confirmed its effectiveness in (1) assessing student essays in terms of evidence usage, (2) extracting evidence and reasoning revisions across essays, and (3) determining revision success in responding to feedback. The evaluation also suggested eRevise+RF is a helpful system for young students to improve their argumentative writing skills through revision and formative feedback.

Figures (6)

• Figure 1: The eRevise+RF system usage pipeline, where students work on three essay drafts, and receive evidence use feedback and revision feedback.
• Figure 2: System architecture, including AES and AES+RF backend systems and user frontend interface. The AES system generates evidence use feedback based on scoring indicators (NPE and SPC scores) and the AES+RF system provides revision feedback based on both scoring indicators and revision classifiers.
• Figure 3: Revision feedback tree, where solid squares are unsuccessful revision feedback focused on helping the student try again, and the dotted squares are successful revision feedback that advances the student to a new evidence usage skill.
• Figure 4: The example of a MVP essay, revising from draft1 to draft2, then from draft2 to draft3. The purple marks deletion, yellow and green marks addition.
• Figure 5: eRevise+RF System Student Interface: Main Page.