Evaluación multimodal de la carga cognitiva en el desempeño de tareas visomotoras complejas: un estudio sobre el sesgo implícito en la interacción humano-ia

Abstract

This study analyses the multimodal evaluation of cognitive load in complex visuomotor test, specifically focusing on implicit bias within human-IA interaction in the context of Industry 5.0. The research aims to understand how unconscious prejudices, measured through the Implicit Association Test (IAT or TAI, Spanish version), modulate user experience and how intelligent systems can act as emotional regulators. Using a multiple single case experimental design (n=3) three female university students aged 20 were evaluated. The methodology integrated subjective measures via NASA-TLX, physiological responses using Galvanic Skin Response (GSR) and electroencephalography (EEG), and visual behaviour through eye-tracking technology. In addition, a conversational chatbot named EsencIA was developed to provide psychological support based on cognitive-behavioural therapy for participants experiencing high frustration. Results indicate that participants perceive the TAI as a demanding task with medium-high mental load, where the temporal demand and effort are prominent. Quantitative data show a correlation between the intensity of implicit bias and visual exploration patterns, suggesting that more consolidated biases require greater cognitive supervision and increased saccadic movements. These findings highlight the importance of integrating multimodal metrics to understand cognitive and emotional processing, confirming the potential of IA as a support tool in high-demand environment to mitigate stress and improve user interaction.