Virtual Reality vs. Real-World Learning: A Comprehensive Neurocognitive Analysis of Brain Activity and Cognitive Outcomes in Kinetic and Spatial Tasks

This comprehensive pilot study presents an in-depth neurocognitive comparison of learning and creative pursuits across immersive Virtual Reality (VR) environments, flat-screen desktop interfaces, and real-world physical settings, with particular emphasis on brain activity monitoring using advanced Electroencephalography (EEG) techniques. In terms of brain informatics and artificial intelligence applications, the study explores the neurological underpinnings of spatial cognition and kinetic learning tasks, analysing their different impact on learning efficacy and cognitive load. Two different task categories: pottery (motor kinetic) and spatial thinking tasks were completed by nine healthy adults (ages 21–35) in three different environments: real world, a flat-screen desktop, and an immersive virtual reality environment utilizing Oculus Quest 2. EEG signals were acquired at 512 Hz using a 24-channel configuration following international 10–20 standards, with comprehensive preprocessing including notch filtering (50 Hz), band-pass filtering (0.1–50 Hz), Independent Component Analysis (ICA), and Common Average Referencing (CAR). Behavioural assessments utilized psychometrically balanced multiple-choice examinations administered post-experiment for each learning environment. In VR-based spatial tasks, EEG spectral analysis revealed substantial elevations in theta (4–7 Hz) and alpha (8–12 Hz) activity, particularly in the parietal and occipital regions, indicating heightened visual engagement and cognitive load. On the other hand, pottery tasks in the real world showed more somatosensory activation and delta band activity, which means that sensorimotor feedback was more integrated. Finding out, performance scores were much lower in VR settings (mean = 3.2/20) than in flat-screen delivery, (mean = 6.8/20). This means that cognitive stress or attention problems could be why more neuronal activation doesn’t always lead to better learning outcomes. This study involves creating basic datasets for EEG-informed, neuroadaptive VR learning systems that can adapt in real time and keep track of cognitive states. This will directly help AI applications that use neuroscience to improve educational technologies.