EEG-informed design reflection: A secondary spectral analysis of neural variation in an AI-augmented interactive installation

This study examines pre- and post-interaction EEG changes associated with an immersive installation experience. Using a publicly available secondary dataset, 30 paired EEG recordings were reprocessed; after quality control, 17 participants were retained for the main analysis and 14 for a stricter sensitivity analysis. Signals from a 32-channel Emotiv Flex system were preprocessed with filtering, average re-referencing, epoch rejection, and Welch-based spectral estimation across five frequency bands. Paired comparisons showed nominally significant post-interaction reductions in beta and gamma power, whereas no corresponding pre–post differences were detected in delta, theta, or alpha; the same beta–gamma pattern was also observed in the sensitivity analysis. The findings are interpreted as design-relevant neural variation associated with the immersive encounter. EEG is treated as a bounded interpretive resource within an AI-augmented immersive art setting, with UDDIE used as a post hoc scaffold for design reflection.