Twenty years of research worldwide has shown conclusively that mobile phones produce unacceptably high levels of distraction for drivers, regardless of whether they’re held or not. Drivers who are talking on a phone fail to notice some hazards and take longer to react to those they do see. Their focus of visual attention contracts to just the area directly ahead, so they miss peripheral hazards.

Phone conversations are more distracting than passengers, who are aware of the varying demands on the driver and adjust their conversation accordingly (e.g. pausing the conversation to allow the driver to concentrate on a tricky situation). Mobile phone conversation is more mentally demanding because it lacks the non-verbal cues that facilitate face-to-face conversations. The interactive nature of phone conversations makes them more distracting than listening to the radio: to hold a conversation, you have to listen to the other person, respond to what they say, remember what's been said and plan what you are going to say next.

Our research at Sussex (e.g. Briggs, Hole and Land 2011, 2016) has confirmed these findings, and identified some factors in why distraction occurs. The argument in favour of allowing drivers to use hands-free phones assumes that driving is primarily a visual activity whereas conversation is a verbal activity. Because these activities are so different, they should not compete for the same mental processing resources and hence should not interfere with each other. We reasoned that this is a misconception of the nature of conversation.

Many conversations naturally involve visual imagery. Imagine you are driving and your boss phones to ask you where you left something: in order to answer, you might well visualise your office, trying to remember whether you left the item in the filing cabinet, on the desk, etc. Or you might be trying to imagine the expression on your boss' face when you realise you have mislaid it! Perception of the external visual world shares processing resources with the brain mechanisms underlying visual imagery.

As a result, any visual imagery experienced from a phone conversation may directly compete for attention with the driving environment. This is why the driver's visual field exhibits what we have called "visual tunnelling" and other researchers have described as "eye freezing": the driver "switches off" scanning the outside world, in favour of mentally inspecting their inner, imaginary world. Consequently they fail to detect hazards or are slower to respond to them.

In our experiments, people performed a video-based hazard perception task, similar to the one in the U.K. driving test. The primary task for all participants was to watch a series of short clips, filmed from a driver's perspective. Half contained a hazard (such as someone stepping into the road, or a car pulling out of a side-turning into to the driver's path) and half did not. We measured how many hazards each person detected, and how quickly they responded to them.

Additionally, participants were divided into three groups according to the secondary task that they performed. One group did the hazard perception test without distraction. The other two groups did the test while distracted by a simple "sentence verification" task: they listened to statements and decided whether each one was true or false. For one group, the truth of each sentence could be decided without needing to use visual imagery (for example, "a leap year has 366 days"). For the other group, decisions about the statements were helped by using visual imagery (for example, "a ten pound note is larger than a five pound note").

Both of the distracted groups were slower to respond to the hazards, and missed many of them. However the visual imagery group were particularly impaired, with reaction times increased by nearly a second. Additionally, eye tracking measurements showed that the distracted groups' visual field was markedly narrowed, to about a quarter of its span in the undistracted group. Sometimes the distracted participants' eyes fell on hazards that they did not respond to – they made "looked but failed to see" errors, where the visual information seems to fail to reach conscious awareness.

Our findings add to the large existing literature on the effects of mobile phones on drivers. Mobile phones are dangerously distracting, regardless of whether they are hands-free or hand-held. Bear in mind these experiments use alert participants who are actively searching for hazards for a short period of time – quite unlike real life driving conditions, where drivers are generally not expecting emergencies to occur. Yet, despite being on high alert, participants in experiments are clearly impaired in one of the most important aspects of driving, namely hazard perception.

We feel that government policy should be evidence-based, and the evidence is clear on this issue. We would recommend a total ban on the use of any communication device (hands-free phone, hand-held phone or CB radio) when driving. Though difficult to enforce, this would at least send a clear message to the public: phone use while driving is unacceptably dangerous.

WIN 20190917 12 05 00 Pro 2

Dr Gemma Briggs

Senior Lecturer in Psychology at the Open University

Graham hole

Dr Graham Hole

Senior Lecturer in Psychology at the University of Sussex