Reliable but multi-dimensional cognitive demand in operating partially automated vehicles
DOI: 10.1186/s41235-024-00591-5
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 addresses the critical need for reliable, objective measures of cognitive demand in real-world driving environments, specifically when operating partially automated vehicles (Level 2 automation). While physiological and behavioral metrics are well-established in controlled laboratory settings, their stability and applicability in noisy, high-stakes highway conditions remain under-researched. The authors aimed to determine whether five common cognitive demand measures maintain test–retest reliability across multiple real-world driving occasions and to investigate whether these measures tap into a single general construct or distinct, multi-dimensional aspects of cognitive workload. The researchers conducted a secondary data analysis involving 71 participants (aged 21–64) who drove four different partially automated vehicles on actual highways across four separate occasions. Simultaneous data collection captured peripheral nervous system activity (heart rate in beats per minute and heart rate variability via RMSSD), central nervous system activity (EEG alpha power in the 8–12 Hz band), and behavioral performance (reaction time and hit-rate on the Detection Response Task). Data were collected during both partial automation and manual driving modes. The study employed intraclass correlation coefficients (ICC1) to assess test–retest reliability for each measure individually. To evaluate the structural relationship between measures, the authors calculated internal consistency coefficients, including omega hierarchical (for a general factor) and omega total (for a multi-factor structure), alongside intercorrelation analyses. Results demonstrated that all five measures exhibited acceptable to excellent test–retest reliability in real-world conditions. EEG alpha power showed excellent reliability (ICC = 0.96), heart rate and heart rate variability showed good reliability (ICCs of 0.80 and 0.72, respectively), and Detection Response Task metrics showed moderate reliability (ICCs of 0.69 and 0.62). Despite this individual stability, intercorrelations among the measures were generally low, particularly between cardiovascular and EEG metrics. Psychometric analysis revealed that a general factor model (omega hierarchical) fit the data poorly, whereas a multi-factorial model (omega total) yielded significantly higher internal consistency. This indicates that the measures do not reflect a single, unified cognitive demand construct but rather capture separable, domain-specific components of workload. The findings confirm that psychophysiological and behavioral measures can be reliably deployed in real-world automation research, providing confidence in their use for monitoring driver engagement. However, the low intercorrelations and superior fit of multi-factor models suggest that cognitive demand during automated driving is a multi-dimensional construct. Consequently, relying on a single metric may provide an incomplete picture of driver state. The study concludes that a combination of multi-modal measures is necessary to comprehensively capture the varied aspects of cognitive demand, supporting the integration of these tools in future real-world assessments of automated vehicle safety and human-machine interaction.
Key finding
In real-world Level-2 automated highway driving, EEG alpha power shows excellent and cardiovascular measures show good test-retest reliability across four occasions, while DRT reaction time and hit-rate show moderate reliability; low intercorrelations and superior fit of a multi-factor (omega total) over a general-factor (omega hierarchical) model indicate cognitive demand is best treated as a multi-dimensional construct in applied automation research.
Methodology
on_road
Sample size: N=71 drivers (25 female), aged 21-64 (M=40.32, SD=13.37); each tested across 4 vehicles.
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. Discovered via unpaywall on 2026-05-07 (7 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 5 | 2026-05-28 |
| archive | success | — | — | — | 2 | 2026-05-07 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-06 |
| promote | success | — | — | — | 1 | 2026-05-06 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 17 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
- workload measurement
- mental demand
- stress driving
- cognitive capacity variation
- situational awareness
- visual occlusion
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
- Methodological Resource: validation psychometrics, measurement protocol