When Pamayla E. Darbyshire and Carl Beitsayadeh sat down to tackle one of higher education's thorniest challenges—how to keep online students from falling through the cracks—they started with an observation that sounds almost obvious, yet一直被忽视。"AI in education should begin with the learner experience," Darbyshire said. That principle now anchors a groundbreaking framework published by these University of Phoenix scholars that could quietly transform how universities support students who never set foot on campus.
Darbyshire, who holds a doctorate in health administration and is a clinical nurse specialist, along with data analyst Beitsayadeh, have developed a 16-stage model that weaves together two powerful technologies—generative AI and predictive analytics—into what they call a "closed-loop support system." Their paper, "Enhancing Student Success through GAI and Predictive Analytics," appeared in the International Journal for Educational Media and Technology, drawing on research presented at the 2025 Teaching, Colleges, and Community Worldwide Online Conference.
The framework addresses a persistent problem: universities often treat predictive analytics and generative AI as separate tools, deploying them in isolation. Darbyshire and Beitsayadeh's model brings them together in a continuous cycle. Here's how it would work in practice: institutional data systems—student information platforms, learning management software, analytics tools—would scan for warning signs like disengagement, late assignments, or slipping grades. Those signals would be translated into risk tiers, flagging students who might need help. Then generative AI could step in with personalized messages, practice quizzes, study plans, or resource recommendations tailored to that individual student's situation. Crucially, human faculty would review these AI-generated suggestions, adding context and judgment before any intervention reaches the student.
"For online learners, timely support matters," Darbyshire noted. "The goal is not to replace the human relationship in learning, but to help educators respond with greater context, clarity and care."
The framework doesn't just imagine a better support system—it maps exactly how to build one, breaking the integration process into 16 stages organized around four phases: data and modeling, risk-aligned interventions, monitoring and feedback, and institutional refinement. It also spells out practical requirements: secure, interoperable data systems; clear policies for data access; faculty training programs; and governance structures that continuously evaluate whether AI tools are accurate, fair, and actually helping students.
Beitsayadeh emphasized that this represents a fundamental shift in how universities think about AI. "This framework brings them together within a single adaptive system, where data-informed insights, AI-enabled support, faculty judgment and institutional oversight operate as interconnected parts of a continuous improvement cycle." For the millions of students enrolled in online programs—many of them working adults balancing jobs and families—earlier identification of struggle, paired with faster, more personalized support, could make the difference between a degree and a dropout.