Human Factors

Dual-task bimanual tracking amplifies dominant-hand prioritization due to gaze bias; enforcing central fixation reduces asymmetry without performance cost, informing selection, training, and cockpit display design.

Users who struggle to detect deception in one digital domain tend to struggle across others; cognitive reflection and digital literacy modestly improve detection, while domain-switching slows responses and can shift response tendencies.

Gestural interface performance depends on user, action, target, and environment interactions; for practitioners, effective design and evaluation requires accounting for these combined factors rather than isolated metrics.

Practice with an input device improves complex tracking regardless of device. However, mismatches harm simple tasks. Selection, learning ability and task adaptation may better predict performance than prior device experience.

Greater system transparency improves human accuracy of automation use. Allowing operators to choose transparency levels does not enhance accuracy compared to providing a fixed transparency setting.

Forward (day-to-night) rotations significantly increase fatigue, degrade sleep, and impact vigilance and delay adaptation; for operations, minimize rotating shifts and prioritize fatigue risk management for safety-critical performance.

HRV and accelerometer data can distinguish mental from physical stress, with combined stress producing stronger, more persistent physiological responses in ICU nurses.

Dual-task interference increases when tasks share modality and as task load rises, especially for detection tasks. Effects depend on load type and modality interactions.

Forward (day-to-night) rotations significantly increase fatigue, degrade sleep, and impact vigilance and delay adaptation; for operations, minimize rotating shifts and prioritize fatigue risk management for safety-critical performance.

Avoiding low back flexion reduces spinal flexion without reducing strength, but high individual variability shows simple directives are insufficient. Personalized coaching is needed to ensure safe and effective lifting performance.

Students’ trust in AI is miscalibrated: nonexperts over-rely on incorrect outputs, while explanation complexity increases trust. This helps experts but can be harmful for nonexperts.

Study quantifies how network latency impacts communication efficiency in video conferencing, showing a 2.6× increase in communication time per delay unit, which demonstrates that low-latency intent signaling significantly mitigates performance degradation.

Study identifies cybersecurity vulnerabilities in robotic-assisted surgery, highlighting training gaps, communication issues, and stakeholder disconnects, and introduces an FMECA-based tool to assess risks and improve patient safety.

Multimodal cues reduce workload under degraded vision, while physiological signals, especially respiration and fNIRS, effectively predict mental workload, supporting adaptive pilot systems.

Drivers reduce trust and reliance under high risk and after automation failure, but trust recovers over time; failure timing does not affect trust dynamics.

Short exposure to optimal behaviors (collecting large amounts of evidence and using machine learning aid) increased these behaviors later, improving performance but not sustaining optimal levels when operators regain decisional freedom.

Older adults discontinued smart speakers for different reasons, including disappointment, low relevance, privacy concerns, and independence, supporting a taxonomy of distinct technology disuse trajectories.

Time pressure non-linearly influences decisions to interrupt a teammate, with individuals biased toward interrupting unless the teammate’s workload is high; however, responses to interruptions follow a rational, linear sensitivity to relative time pressure.

Passive exoskeletons prevented low-back tissue creep and reduced muscle strain during prolonged stooping, helping maintain spinal stability and lowering potential risk of low back pain development.

Simplified runway status lights reduced runway incursion risk and workload, especially when traffic was hard to see, with pilots generally viewing the system as helpful and safety enhancing.

Drivers often disengaged automation when traffic cues and system behavior conflicted with expectations, highlighting the need for clearer feedback and more intuitive overtaking behavior.

Highlighting displays improved reaction time, reduced workload, and increased trust, with surface aiding safety perception and outlines reducing mental demand.

Repeated exoskeleton use shifts attention to tasks and increases self-integration; perception of the device as part of oneself develops over time and differentiates novice versus experienced users.

New road markings and signs reduce speed, improve positioning, and increase safe riding behaviors among motorcyclists, with sustained effects and no negative behavioral impacts observed.

Selecting moving targets using a crossing technique was associated with greater advantages than pointing and clicking. However, the advantages of crossing decline when targets become larger or move slower.

A 10-attribute taxonomy was developed to review 103 human-agent teaming (HAT) testbeds; most are simple dyadic teaming with human-led structures. More flexible, multi-agent, and ecologically valid testbeds are needed.

The ankle exoskeleton did not change novice users’ street crossing decisions, but it slightly increased workload; practitioners should evaluate cognitive fit, not just physical assistance, in urban mobility tasks.

Increased spatial separation of information presented on augmented reality (AR) head-mounted displays is associated with increased response times. Practitioners need to keep in mind that head movements can be costly.