An online hazard prediction test demonstrates differences in the ability to identify hazardous situations between...

Area of Science:

• Cognitive Psychology
• Road Safety Research
• Human Factors in Driving

Background:

• Traditional hazard perception tests show cultural bias and do not reliably differentiate driver experience.
• Speeded response times in existing tests are susceptible to threshold bias, impacting cross-cultural validity.
• An alternative hazard prediction test methodology shows promise for international application.

Purpose of the Study:

• To adapt and validate the hazard prediction test methodology in a new cultural context (Israel).
• To evaluate an online platform for delivering the hazard prediction test.
• To incorporate Signal Detection Theory parameters (criterion and sensitivity) by including hazardous and non-hazardous trials.

Main Methods:

• 74 participants (43 experienced, 31 novices) viewed 26 video clips filmed in Tel Aviv, Israel.
• Clips were occluded before hazards fully materialized, requiring participants to identify and predict traffic scenarios.
• Participants completed a binary hazard identification task followed by a four-option multiple-choice prediction task.

Main Results:

• Experienced drivers significantly outperformed novices on the multiple-choice prediction task when they correctly identified the hazard.
• Novices demonstrated poor prediction accuracy irrespective of hazard identification.
• Driving experience and hazard sensitivity were significant predictors of prediction accuracy; criterion was not.

Conclusions:

• The newly developed online hazard prediction test successfully differentiates between experienced and novice drivers.
• The hazard prediction methodology is suitable for international adaptation and deployment.
• This approach offers a scalable solution for global driver assessment and training.