UC Irvine Launches Center to Test AI in Writing Classes

University of California, Irvine is launching a national research center to determine whether artificial intelligence actually helps college students write better or merely does the work for them. The National Center for Writing Research to Improve Teaching Effectiveness with Generative AI will operate for five years under a $10 million grant from the U.S. Department of Education.

The project is unique because it is the first of its kind to focus specifically on writing instruction. Mark Warschauer, an education professor and the center’s director, explained that the goal is to move beyond anecdotal evidence. The center will generate data to help institutions decide how to adopt AI in a way that strengthens learning rather than undermining it.

Testing a specialized writing tool

The research will focus on a specific software platform called PapyrusAI, which was developed by UCI researchers. Unlike general chatbots that can answer any question, this tool is designed specifically for education. It guides students through the process of planning and revising their work, aiming to improve their skills rather than just generating text for them.

Warschauer noted that while previous case studies showed promising results, no one has yet conducted a rigorous randomized controlled trial. This new study will change that by comparing two groups of instructors. One group will use the AI-infused curriculum, while the other will continue with traditional teaching methods.

Measuring real student outcomes

The trial will involve more than 1,000 community college students across 60 classes in three different states. The researchers will look beyond just the quality of the final essay. They will measure whether students improve in writing confidence, gain better literacy in using AI tools, and are more likely to persist in their studies and graduate.

This approach addresses a significant gap in current educational research. By isolating the effect of the AI tool, the center can determine if the technology leads to genuine skill acquisition. This is crucial for understanding the long-term impact of digital assistants on student development.

Navigating regulatory uncertainty

The path to securing this grant was not straightforward. Warschauer explained that the review process took over a year due to administrative changes within the Department of Education. These changes included attempts to move or shut down various research programs, creating a period of uncertainty for the applicants.

Despite these hurdles, UCI’s proposal was the sole award selected through the national competition. The funding signals a federal interest in understanding the practical effects of generative AI in higher education. As reported by GN technics/ai (en-US), the center’s work will provide some of the first hard data on how these tools fit into the classroom.