"Heavy toll: 1 in 4 young drivers involved in crash in their first six months on the road"
Daily Mirror, 20 July 2012
Before parents of teenagers rush out to prepare their cars for a soft impact, let's see about that.
The Press Association tells a different, less surprising story.
They say: "Of drivers who have had an accident, 23% of 18 to 24-year-olds crashed within six months of passing their test." But this is just the start of a messy story.
The figures come from a poll conducted by British Polling Council member Populus.
BPC members are the only polling organisations it is easy to trust because, as we did here, you can ring them up and get hold of the survey questions and answers to judge for yourself.
A copy of the press release was kindly provided to us by the AA Charitable Trust.
With the survey data from Populus, which should be available on their website soon (Update 23/07/2012: the figures are now online, and can be found here), we can see that 19,284 drivers were surveyed.
The Press Association reports it as a survey of 14,000 drivers who have had accidents, following the press release which says "14,229 drivers who have been involved in a car crash were surveyed".
That's fair. 26 per cent of respondents have never been involved in collisions while driving, leaving 14,229 who had.
But 67 per cent of the 365 18-24 year olds surveyed had never had an accident, so that of the 14,000 drivers the survey is based on, a figure that is prominent in both stories, only 119 were aged 18-24.
That's quite a contrast to the impression given by the Mirror completely erroneously talking about "14,000 drivers aged 18 to 24 questioned."
What is going on with novice drivers?
Of the 119 18-24 year olds who had had a crash, 27 crashed in their first six months with a license. 27 out of 365 18-24s in total.
So if we were to rewrite the Mirror's headline based on the polling data, we would conclude that just 7 per cent of young drivers are involved in a crash in their first six months on the road, and that the Mirror was out by a factor of four.
The problem with that is that the specialist research suggests a different conclusion.
Transport Research Labs' Cohort Study of learner and novice drivers, cited by the House of Commons Transport Select Committee in its 2007 report on Novice Drivers, found that "18% of all new drivers were involved in at least one crash within one year of passing their test."
Another piece of TRL research, reported here, was:
"A modelling study based on accident records in Great Britain for the years 1996-1998 and extracts from driver licence records for the same years, separated the effect of age and experience and showed that the effect of experience on accident involvement is… is particularly significant in the early years of driving. The reduction in accident involvement due to experience during the time immediately after passing the driving test was estimated; during the first six months of driving, accident involvement for a driver of any age falls by a factor of 2.3 for male drivers and 1.9 for female drivers, while in the first year it falls by a factor of 3.6 for men and 2.7 for women."
So it is hard to believe that the results of these polls are good evidence of the things they have sought or attracted publicity for.
If you misunderstand the poll as its commissioners and the press seem to have done, the result is too high, and if you understand the poll correctly, the result seems to be unrepresentatively low.
Whether or not the poll is actually useful evidence, the Mirror does not seem to have reported it accurately.
The Mirror's headline can be traced back to the bad press release, so why did the AA Charitable Trust not get its facts straight? We have asked the Trust to clarify or say whether they will correct their release but unfortunately the people who prepared the press release are not available today.
Careless driving, perhaps, from the AA? And a speeding ticket for the paper?