Speaker
Description
Rogo, I. S.
FEDERAL UNIVERSITY OF EDUCATION ZARIA
ibrahimsadirogo@gmail.com
Track: AI, Emerging Technologies and Future Learning
The integration of Generative AI and intelligent tutoring systems (ITS) into higher education has fundamentally altered student-content interaction, yet its cognitive and psychological overhead remains under-explored. This study investigates the real-time cognitive dynamics specifically germane, intrinsic, and extraneous cognitive load—alongside metacognitive monitoring and dynamic feedback processing when students interact with adaptive AI tutors. Utilizing a multi-modal data collection framework combining psychometric inventories, real-time eye-tracking, and interaction-log analytics, we monitored undergraduate students engaged in complex problem-solving tasks with generative conversational agents. Findings indicate that while real-time, personalized AI feedback reduces extraneous cognitive load and accelerates error recovery, excessive conversational AI assistance can inadvertently suppress metacognitive calibration, leading to over-reliance and surface-level information processing. Conversely, scaffolding AI feedback with delayed-prompt interventions significantly heightened metacognitive awareness and deep conceptual retention without causing cognitive overload. By mapping the exact cognitive trade-offs involved in dynamic human-AI learning interactions, this paper establishes an empirical foundation for designing psychologically optimized AI tutoring systems that foster autonomous, higher-order thinking while safeguarding cognitive well-being in digital higher education ecosystems.