How studies were done
Four facets that cut across every topic: the method, the instruments, the automation level studied, and any standing dataset analysed. Browse these when the question is how rather than about what — "which studies used the DRT", "which were on-road", "which looked at Level 2".
Read the counts carefully. These facets are sparsely declared. A value showing few papers usually means few papers declared it, not that few studies did it — every count below is given against how many papers carry that facet at all.
Method
How the study was conducted. 19870 of 62594 papers declare this facet (31.7%).
- narrative review — Non-systematic review or theoretical synthesis. 6445
- on road instrumented — Real-vehicle, public roads, controlled protocol, experimenter present. 3755
- crash database analysis — Analysis of pre-existing crash records (FARS, CRSS, GES, in-depth). 3635
- survey questionnaire — Self-report instrument as primary data. 3185
- simulation modeling — Computational/agent-based modeling, no human participants. 3037
- simulator — Driving simulator of any fidelity (desktop to motion-base). 2803
- lab experiment — Non-driving lab task (PRP, OSPAN, n-back, perceptual psychophysics). 2446
- retrospective observational — Analysis of pre-existing non-crash records (telematics, claims data). 2124
- other — None of the above. Note required. 2029
- naturalistic study — Instrumented vehicles in regular use, no experimenter intervention. 2028
- case study in depth — One or few cases analyzed in depth. 1681
- tools paper — Paper introduces a tool, method, or platform (e.g., DReyeVR). 1588
- field operational test — Large-fleet naturalistic deployment, often with intervention arm. 1441
- test track closed course — Closed course / proving ground / private track. 1341
- qualitative interview focus group — Interview, focus group, or grounded-theory study. 748
- dataset paper — Paper introduces or describes a dataset. 641
- meta analysis systematic review — Quantitative synthesis or PRISMA-style review. 11
Instruments & Measures
What the study measured with. 9327 of 62594 papers declare this facet (14.9%).
- driver facing camera — Driver-facing camera (gaze, head, behavior). 2350
- subjective scale other — Other validated questionnaire (SART, Trust scales, Driving Anger Scale, etc). 1514
- eeg — Electroencephalography. 1352
- subjective scale nasa tlx — NASA-TLX questionnaire. 1308
- eye tracker — Any eye-tracking system (head-mounted or remote). Brand goes in summary, not the tag. 1222
- subjective scale kss — Karolinska Sleepiness Scale. 1072
- pupillometry — Pupil-diameter measurement (subset of eye_tracker; flagged separately for relevance). 1057
- eeg or eda gsr — EEG and/or EDA/GSR co-occurring (papers using both physiological measures). 908
- respiration — Respiration belt or sensor. 811
- vog device — Visual Occlusion Glasses (ISO 16673). 751
- wearable smartwatch — Consumer wearable (HRV, accelerometry). 743
- ecg — Electrocardiogram. 652
- eda gsr — Electrodermal activity / galvanic skin response. 645
- other — Hardware not in vocabulary. Note required. 637
- subjective scale dssq — DSSQ questionnaire. 510
- vehicle can bus — CAN-bus extraction (steering, throttle, brake, speed). 365
- fnirs — Functional near-infrared spectroscopy. 362
- head tracker — Head-pose tracking (camera or magnetic). 312
- drt device — Detection Response Task device (ISO 17488). 158
- lidar radar external — External-world sensors (sensor fusion data). 87
- dashcam video — Forward / cabin video. 81
- gps imu — GPS / inertial measurement. 77
Automation level
The SAE level of automation studied. 8040 of 62594 papers declare this facet (12.8%).
- manual l0 — No automation. 3632
- not applicable — Paper not about automation level (most distraction/cognitive work). 3080
- mixed levels — Paper compares two or more levels. 1629
- l1 advisory — One assistance feature (ACC or LKA, not both). 1546
- l3 conditional — System handles DDT, driver is fallback. 707
- l4 high automation — System handles DDT in defined ODD, no driver fallback needed. 293
- l2 supervised — Two assistance features active simultaneously, driver supervises. 244
- l5 full automation — System handles DDT everywhere. 32
Dataset
A standing dataset the paper analyses. 5322 of 62594 papers declare this facet (8.5%).
- shrp2 100car analyses — SHRP2 + 100-Car combined analyses / cross-study comparisons. 2141
- mit avt — MIT Advanced Vehicle Technology study. 1539
- shrp2 — SHRP2 Naturalistic Driving Study data. 1056
- fars — NHTSA Fatality Analysis Reporting System. 729
- udrive — UDRIVE European naturalistic study. 520
- 100 car study — VTTI 100-Car Naturalistic Driving Study. 421
- ges crss — NHTSA General Estimates System / Crash Report Sampling System. 391
- nds other named — Other named naturalistic dataset (specify in note). 313
- custom dataset — Authors' own data not from a standing dataset. 134
- iihs data — IIHS database (HLDI, real-world performance). 36
- oem proprietary — OEM-internal data (Tesla fleet, Volvo Drive Me, etc.). 36
- other — Dataset not in vocabulary. Note required. 20