Multimodal Study of the Effects of Varying Task Load Utilizing EEG, GSR and Eye-Tracking

Born, Jannis; Ramachandran, Babu Ram Naidu; Romero Pinto, Sandra Alejandra; Winkler, Stefan; Ratnam, Rama · 2019 · Crossref

DOI: 10.1101/798496

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the relationship between cognitive task load and performance by simultaneously analyzing multi-modal physiological data and subjective self-reports. The research addresses the challenge of objectively measuring task loading in real-time, a capability essential for dynamic environments where mental and physical demands fluctuate. While previous studies often relied on single modalities like EEG or eye-tracking, this work aims to determine which combinations of physiological signals—specifically electroencephalography (EEG), galvanic skin response (GSR), and eye-tracking—best predict task performance and correlate with perceived workload. The experimental design involved eight participants performing a computer-based visual search task simulating postal code sorting. Participants matched five-digit numbers to one of six numeric ranges under varying conditions of difficulty. Task load was manipulated by altering three binary variables: color consistency, numerical arrangement stability, and response time constraints. This created eight distinct blocks of increasing difficulty, plus a repeat of the most difficult condition. Data were collected using a 32-channel EEG system, an eye-tracker, and GSR sensors, synchronized with participant responses. Subjective workload was assessed using the NASA Task Load Index (TLX) questionnaire administered at discrete intervals. The results indicated that low beta frequency EEG waves (12.5–18 Hz) became more prominent in frontal and parietal regions as cognitive load increased. Eye-tracking data revealed more frequent blinks and increased pupillary dilation under higher load, with blink duration showing a strong correlation with task performance. GSR analysis distinguished between phasic components, which related to specific cognitive workload, and tonic components, which reflected general arousal. Subjective reports via NASA-TLX confirmed increased frustration and mental workload in difficult conditions. Statistical analysis using one-way ANOVA demonstrated that EEG and GSR provided the most reliable correlations with perceived workload levels. When combined, these two modalities were the most informative for predicting performance, outperforming single-modality approaches. The study concludes that while EEG is a strong predictor of task performance, integrating GSR data significantly enhances prediction accuracy. This multimodal approach allows for the isolation of the most informative physiological markers in controlled settings, facilitating the development of robust monitoring systems for real-world work environments. The findings support the use of combined physiological and behavioral metrics to objectively assess cognitive load, offering a pathway to reduce estimation errors inherent in single-modality measurements.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• workload measurement
• neuro workload indices
• mental demand

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data, self report data, behavioral performance data