Implications for Delivery
Combine different training methods to improve learning.
Encourage learners to speak out loud during training. Verbal commentary helps them focus their attention and become more aware of their surroundings.
Give feedback after each task. Explain what the learner did well and what they missed.
Adapt the training to each learner. Consider their level of experience, confidence, and personal learning needs.
Allow learners to repeat exercises. Repeating scenarios helps strengthen memory.
As learners gain experience, shift to more interactive methods. Use driving simulators or computer-based activities to give them hands-on practice in realistic situations.
Background
An essential part of learning to drive is not only to operate a vehicle but also predict and respond to potential hazards. Hazard prediction refers to the driver’s ability to foresee possible dangers based on environmental and behavioural cues, such as a pedestrian nearing a crosswalk or a vehicle inching out from side road. This skill helps prevent crashes and is an important part of safe driving. Hazard prediction is therefore different to hazard perception, in that it is about predicting the consequences of a hazard emerging and then taking action, whilst hazard perception is the skill of recognising a hazard that is present.
Experienced drivers are usually better at hazard prediction than young or novice drivers. This difference contributes to the overrepresentation of novice drivers in crash statistics. With practice, drivers learn to scan the environment better, interpret the signs earlier, and stay one step ahead of danger.
Importantly, hazard perception and prediction are now seen not only as reactive skills but also as proactive components of driving, similar to situational awareness. They include perceiving a hazard, understanding its severity, and projecting what might happen next if no action is taken. This aligns closely with the “What Happens Next” paradigm widely used in driver training and evaluation.
What the Research Shows
Research shows that targeted training can improve hazard prediction skills, especially in learner drivers. The literature outlines various training formats, including:
Commentary driving: This method asks learners to say out loud what they see and what they would do while watching driving scenes or using a simulator. It helps building hazard scanning habits and makes the driver focus more clearly on what’s happening around them. This improves their awareness and ability to spot danger early.
Video-Based Training and “What Happens Next (WHN)” Tasks: Video scenarios are often used in training because they show real-life driving situations in a safe and repeatable way. In “What Happens Next” (WHN) tasks, the video is paused at a key moment, and the learner is asked to guess what might happen next. Some programmes also slow down videos or pause them to help learners focus and think more clearly. This helps drivers think ahead by helping them notice dangers as they develop and become quicker and more accurate through repeated practice.
Implicit Learning: These methods aim to develop skills through exposure, rather than direct instruction. Learners repeatedly engage with hazard scenarios, without being explicitly told what to look for. Over time, learners become more attuned to patterns and early warning signs.
Computer-Based Training (CBT): Computer-based training uses interactive tools to help learners improve hazard perception and prediction skills. It includes images, videos, and quizzes to make learning more engaging. Learners might be asked to click on places where a hazard could appear or choose the best response to a driving situation. CBT is flexible, easy to repeat, and can be a useful addition to practical, on-road training.
A review of many studies shows that the best way to teach hazard prediction is by using a mix of methods (instruction, expert demonstration, active practice). Training that uses different tools lead to better outcomes like faster hazard reaction times and more accurate hazard detection.
However, some questions remain. It’s still not clear how long the benefits of training last or how directly they affect crash rates. Some research also shows that training may work differently for different groups, such as between men and women, which suggests a more personalised approach could be helpful.
In summary, the evidence supports the integration of hazard prediction into all stages of driver training. Structured, interactive, and repeated practice (ideally combining multiple methods) can help drivers spot and respond to dangers earlier, making roads safer for everyone.
Sources
Cao, S., Samuel, S., Murzello, Y., Ding, W., Zhang, X., & Niu, J. (2022). Hazard perception in driving: a systematic literature review. Transportation research record, 2676(12), 666-690.
https://doi.org/10.1177/03611981221096666
Crundall, D. (2016). Hazard prediction discriminates between novice and experienced drivers. Accident Analysis & Prevention, 86, 47-58.
https://doi.org/10.1016/j.aap.2015.10.006
Šeibokaitė, L., Endriulaitienė, A., Markšaitytė, R., & Slavinskienė, J. (2022). Improvement of hazard prediction accuracy after training: Moderation effect of driving self-efficacy and road safety attitudes of learner drivers. Safety science, 151, 105742.
https://doi.org/10.1016/j.ssci.2022.105742