Beyond Behavioral Adaptation: Evaluating the Added Value of EEG in Personalized Language Interfaces

Neuroadaptive interfaces personalize human–computer interaction using brain activity, but whether neural measurements provide useful information beyond behavioral indicators remains unclear, particularly in language comprehension tasks. We propose a framework to investigate the added value of EEG in adaptive communication support. Written messages are presented through combinations of sentence segmentation, pictograms, and keyword labels, with presentation choices informed by individual preferences and task performance. A behavior-adaptive policy serves as the baseline for comparison with a neuroadaptive policy incorporating calibrated EEG estimates under signal-quality and uncertainty constraints. The research plan includes interface validation, behavioral feasibility testing, and an EEG-based comparison using held-out sessions. Evaluation focuses on comprehension accuracy, response time, subjective effort, and adaptation behavior. The objective is to determine whether neural information improves assistance decisions and user outcomes beyond behavioral adaptation.