Identifying the Electroencephalography Features for Measuring Cognitive Load in Computer-Aided Design

Abstract This article investigates the cognitive load (CL) in the underexplored context of computer-aided design (CAD), focusing on identifying the most effective electroencephalography (EEG) features for measuring CL variations. While previous research has demonstrated that the EEG can distinguish between CL levels in various domains, there is limited understanding of whether existing EEG-based indicators can accurately reflect CL changes in CAD activities. To address this gap, the study first extracted a list of potential EEG-based CL indicators from the literature and then evaluated their applicability to CAD tasks through an experimental study involving 24 engineering designers performing CAD modeling tasks of varying complexity. The experimental study employed two complementary methods: NASA TLX as a subjective measurement and EEG as a psychophysiological measurement, enabling a comprehensive analysis of the CL in CAD tasks by validating and comparing the findings from both methods. The results revealed six EEG features sensitive to changes in the CL, with an increase in parietal alpha task-related power emerging as the most prominent indicator. This study makes a novel contribution by examining the relationship between the EEG-based CL indicators and the NASA TLX scores, highlighting meaningful correlations, and emphasizing the importance of both frequency bands and cortical areas when interpreting EEG signals in relation to CL.