About Reed in Partnership

Reed in Partnership is a national public service provider working with public sector commissioners across public health, employment support, assessment, justice, advice, and skills. More than 250,000 people have benefited from its healthy lifestyle services, including delivery of the NHS Diabetes Prevention Programme, and around 350,000 people have been supported to move from unemployment into work.

Reed in Partnership also manages driving theory test centres for two-thirds of the Driving and Vehicle Standards Agency (DVSA) network, delivering around 2.5 million theory tests each year across two-thirds of the UK. Inclusion is central to this service, with tailored support provided to ensure people with health conditions, disabilities, or learning differences can access assessments on an equal basis.

Alongside its assessment role, Reed in Partnership has a strong interest in interventions that support behaviour change and improve safety outcomes, particularly for young people.

Project Overview

Reed in Partnership commissioned Esitu Solutions to design, deliver, and evaluate a virtual reality (VR) driver training module aimed at young pre-drivers, learner drivers, and recently qualified drivers. The project explored how immersive, evidence-based hazard perception and hazard prediction training could be embedded within live road safety provision and delivered at scale. This builds on Reed in Partnership’s wider work in supporting young driver safety.

The training was conducted as part of Bedfordshire Road Safety Partnership’s MORE course, a free, voluntary programme for young people aged 16–30. The VR module replaced an existing session and was delivered within the constraints of a live, multi-activity training environment.

The Challenge

As a driving theory test centre provider for the DVSA, Reed in Partnership is committed to contributing to the elimination of traffic fatalities and serious injuries on our roads. Because a disproportionate number of young drivers are involved in collisions, the organisation wanted to focus on this group and explore innovative ways of supporting driver safety as part of the Safe System approach.

Reed in Partnership saw a gap in the support available to learner and novice drivers, particularly around understanding how hazards develop and how risk can be anticipated in real driving situations. This interest also dovetailed with their wider commitment to supporting behaviour change in other sectors, as well as their broader agenda of improving outcomes for young people.

The organisation was particularly interested in:

• Exploring innovation in training delivery, and were excited at the prospect of using VR as an engaging tool
• Supporting behaviour change, not just knowledge acquisition
• Understanding how immersive training might engage young people more effectively
• Ensuring approaches were accessible to learners with special educational needs and disabilities (SEND)

Why Hazard Perception and Prediction?

Hazard perception is the ability to identify and respond to developing hazards early enough to avoid a collision. It is the only higher-order cognitive driving skill that has been consistently linked to crash risk and forms a core part of the UK licensing process.

For this project, we adopted a simpler and more transparent response mechanism than traditional hazard perception tests. Whereas the DVSA test requires participants to press a button the moment they detect a developing hazard, the VR module used a hazard prediction test. Each clip cuts to black at the point of hazard onset, and participants are asked: “What happens next?” They are then shown four on-screen options, only one of which correctly describes the next event. Choosing the correct answer awards one point, creating a clear and intuitive scoring system that young drivers found easy to understand.

This approach was particularly well suited to group delivery across multiple VR headsets, where individual reaction-time measurements would have been impractical. Importantly, a hazard prediction test is not a different skill: it is simply another, and arguably a better, way of measuring hazard perception, capturing the same underlying cognitive processes involved in anticipating and avoiding risk on the road.

The Solution

Esitu Solutions brought specialist expertise in hazard perception research and training design, with the underlying VR clips already supported by academic research. This meant the project could focus on practical deployment, learner engagement, and evaluation, rather than on content validity.

The one-hour VR training module was developed by Esitu’s traffic and transport psychologists and structured around three components:

1. An introduction to hazard perception and the DVSA hazard perception test
2. An explanation of the key cognitive processes involved in detecting hazards
3. Explicit guidance on spotting hazard “clues” to support prediction and avoidance

Participants experienced 360-degree hazard prediction clips through VR headsets, followed by error-based feedback explaining where they should have been looking and why. The module was delivered alongside other activities within the MORE course, with one session dedicated specifically to SEND participants.

Evidence and Evaluation

The VR training was evaluated through a live project involving 160 participants across three course deliveries. Participants completed questionnaires before and after the training to assess their knowledge of hazard perception, understanding of hazard perception processes, and confidence in their own skills. Hazard prediction performance was also measured using a short assessment within the session.

The evaluation showed clear improvements in:

• Knowledge of the DVSA hazard perception test
• Understanding of the processes involved in hazard perception
• Confidence in future hazard-spotting ability

These improvements were observed across both neurotypical participants and those with SEND, although differences between the groups highlighted the need for tailored approaches to better support diverse learners.

Due to the short session length and the constraints of delivering training in a live course environment, no significant pre- and post-training improvement in hazard prediction scores was observed. The evaluation highlighted that more extensive assessment and longer training duration would be required to robustly measure behavioural skill change.

Overall, the findings demonstrated that evidence-based VR training can be successfully embedded into real-world road safety provision and can meaningfully improve understanding and confidence among young and novice drivers.

What Reed in Partnership Said

What We Learned

Delivering hazard prediction training through VR proved to be highly engaging for young people and provided valuable insight into how training design, delivery context, and session length influence learning outcomes.

The project highlighted the importance of:

• Allowing sufficient time for training and assessment
• Providing strong train-the-trainer support
• Adapting delivery to better support SEND learners
• Selecting robust and reliable VR platforms

What’s Next

Reed in Partnership and Esitu are exploring opportunities to build on this project, including refining the training structure, extending session duration, and investigating wider deployment. Both organisations recognise a genuine gap in engaging, evidence-based support for learner and novice drivers and see strong potential for VR-based hazard training to contribute to improved understanding, confidence, and safer decision-making over time.