Modeling cognitive load effects of conversation between a passenger and driver
DOI: 10.3758/s13414-017-1337-2
archive: archived pipeline: cataloged verified
Summary
Driving-simulator + cognitive-modeling study (Tillman, Strayer, Eidels, Heathcote; U. Newcastle + U. Utah + U. Tasmania + Vanderbilt; Atten Percept Psychophys 79, 2017, doi:10.3758/s13414-017-1337-2) examining whether DRT (Detection Response Task) slowing under cognitive load reflects loss of information-processing rate, increased response caution, or non-decision-time changes. Pairs of participants performed the ISO-standard DRT while driving a simulator; cognitive load was manipulated by directing the driver-passenger pair to engage in conversation. The single-bound diffusion (Wald) model was fitted to DRT response-time distributions to decompose effects into drift rate, threshold, and non-decision time.
Key finding
DRT slowing under conversational load is driven by increased response caution (threshold), not by reduced rate of information accumulation. This challenges the standard interpretation that DRT directly indexes 'remaining processing capacity' and instead suggests it reflects a strategic shift toward more cautious responding under cognitive load — important for how DRT is interpreted in ISO 17488 / human-factors workload assessment.
Methodology
Driving-simulator study with paired participants (driver + passenger). 40 undergraduates (28F) at U. Newcastle. Driver performed DRT while driving; cognitive load manipulated by conversation between driver and passenger. Single-bound diffusion (Wald) cognitive model fitted to DRT RT distributions to decompose drift rate, response threshold, and non-decision time. Comparison to standard diffusion model.
Sample size: N=40 (28F undergraduates) in pairs.
Quality score: 5 / 5