Project Activities
In Phase I of the project in 2024, we tested the AI Helper in four middle and high schools. It was found that the Generative AI technology with teacher-customized rubrics and topic-specific resources uploaded by the teacher, the AI Helper provided students with accurate ratings and constructive reviews to revise their work. Pilot research results showed evidence of active student engagement with the AI Helper tool and improvements in writing skills. Students acknowledged the usefulness of the feedback. Teachers noted the ways in which generative AI can magnify the teacher’s voice, making it available immediately and to all students. Through the medium of the AI Helper, 1-to-n teacher became a 1-to-1 teacher simultaneously for the whole class.
In Phase II, the CGScholar AI Helper will be developed into a full-scale, classroom-ready writing and learning assistant. The project adds two major innovations: (1) an AI toolbar that supports students during drafting, offering brainstorming, research, outlining, expansion, summarization, flow checking, and proofreading; and (2) an AI-assisted rubric generator that enables teachers to quickly create detailed, standards-aligned rubrics from their own lesson materials. A key feature is the ability to monitor the balance between student and AI contributions, addressing concerns about AI misuse while fostering ethical and effective writing practices. Development will follow agile “cyber-social research” methods with iterative two-week design, release, and evaluation cycles. Pilot testing will proceed in three stages: (1) usability studies with small groups in controlled settings, (2) feasibility trials in diverse high school classrooms across districts (targeting 200–1,000 students), and (3) a large-scale study with comparison groups. Professional development modules (“GenAI for Writing and Subject Area Teachers”) will support teachers’ integration of the tool. The end goal is a scalable, web-based system that improves student writing outcomes, supports teacher workload, and provides transparent tracking of learner progress and AI contributions.
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.