Organisation
National Young Rider Forum and Esitu Solutions
Amount awarded
£51,594
Completed
2025
Uploaded to Knowledge Centre
2 December 2025
Given that hazard perception is measurable, trainable, and a known predictor of crash risk, this project set out to create a motorcycle-specific hazard test designed to assess riders’ ability to identify hazards in the road ahead.
A valid hazard test will lay the groundwork for developing training interventions and other road safety resources for these vulnerable road users. The project team chose to create its test using real world video footage, to ensure clips reflected genuine hazards.
To identify the ideal camera position for the test, pilot filming involved recording motorcycle footage from three camera viewpoints: bike-mounted, head-mounted, and ‘drone follow’ views. A focus group consisting of academic experts, police representatives, and road safety stakeholders provided advice on the viewpoint selection, hazard types, and early test designs.
Based on the group’s recommendations, the final footage was filmed using the bike mounted viewpoint to ensure maximum visibility of upcoming hazards. Resultant footage was subsequently reviewed by transport psychologists and a further expert focus group to select the most valuable clips. Three different measures of hazard skill were captured across 18 final clips: Hazard perception (press a button when a hazard develops), hazard prediction (the clip occludes at hazard onset and participants are asked ‘What happens next?’), and hazard avoidance (clip occludes at a critical decision point and participants are asked ‘What would you do next?’). The two latter measures required riders to select an answerfrom four on-screen options. This three-metric approach captured different
aspects of a rider’s hazard awareness.
Participants were recruited online via the participant recruitment panel Prolific and completed the test through a learning management system, EsituDrive. Following data cleaning, the project team had a total sample of 125 motorcycle riders and 72 car drivers.
Of the three metrics recorded, the hazard perception scores of the motorcycle riders were significantly related to the number of self-reported collisions in the last three years, with those drivers who reported more collisions performing worse on this measure.
The hazard avoidance and hazard perception metrics did not relate to any of our proxy measures of safety, implying that those tests would require further iteration to be included in future assessments.
The project team also found that hazard prediction scores were higher than hazard perception scores. This may indicate that the prediction questions were too easy or that the perception windows were set too severely. Despite this, the hazard perception metric remained the most valid in predicting real-world safety outcomes.
Some limitations were identified, including the positive relationship between response frequency (number of clicks made) and hazard perception scores. Despite confirmation of this well-known methodological problem, the link between perception scores and crash history for riders suggests the test does tap into hazard perception skill and is not just an artifact of frequent clicking.
Click the following link to read the full report:
https://www.roadsafetytrust.org.uk/small-grants-awarded/national-young-rider-forum