What you need to know about cloud backup

Exclusive The Department for Transport (DfT) has "unwittingly" misled the public over the benefits of speed cameras for the last four years.

That was the shock admission yesterday by a DfT spokeswoman, when finally cornered by the Department’s own research. She also told us that they have finally agreed to put matters right by adding an explanation to future public statements.

The claim has been repeated frequently by official spokespersons and road safety campaigners, and on the DfT’s own Think! road safety website. It has also been regularly questioned by speed camera opponents, who point out that other effects should be taken into account. These include overall "trend" improvements in the KSI rate, other road safety measures put in place at accident black spots, and a statistical quirk known as "regression to the mean".

That last factor is important. Scientists and statisticians have long been aware that whenever something out of the ordinary happens – from a plague of frogs to a spate of road accidents - it is probably just that: a freak, a fluke, an anomaly. Anything that you do in that location after the event will look as though it makes a difference – but it hasn’t. The figures would have returned to a more normal average rate of their own accord anyway.

When the DfT first started claiming such a high benefit for speed cameras, respected academics Dr Linda Mountain of Liverpool University and Mike Maher, Professor of the Mathematical Analysis of Transport Systems at Leeds, objected. The DfT took notice, and the 2005 report included an appendix supplied by this pair showing in meticulous detail how the effect of speed cameras was almost certainly less than half the 42 per cent quoted.

The Reg has raised this with the Department for Transport many times since. On each occasion we have been fobbed off. In one instance, a spokesman told us that statistical analysis was no more than "a matter of opinion".

So far, so stonewalled - until this week, when the DfT decided that speed cams were no longer quite so deserving of central government support. Once more we drew their attention to this issue, expecting to be brushed off again. At long last, a spokeswoman told us: "This was basically an oversight and it will be corrected."

We checked: she confirmed. In future, DfT websites will still contain the 42 per cent claim, but there will be further explanation – or so we have now been assured.

A quick check of the Think! site shows the figure has already been changed. ®

Agentless Backup is Not a Myth