|15 Feb 2014.
For immediate release.
"It is statistically invalid to calculate an average of percentages by simply adding the highest and lowest and dividing by two. The individual percentage changes have to be weighted according to the actual numbers involved to give a true average. That such an elementary error could be made in a local authority report is both extraordinary and concerning, especially when the figures so derived are used by others to justify speed limit changes that could affect millions of people."Using a Freedom of Information request, Mr Heymer obtained detailed data for all the survey points used to count pedestrian and cyclist numbers in Bristol's two pilot 20mph areas. This showed that counts were made on just one day of the week and one weekend day at each point, before and after the speed limit was changed. All the counts were carried out in August, a school holiday period that is normally avoided for traffic surveys because it is not representative of the rest of the year. These limitations alone mean that the results cannot be relied upon to give a statistically valid picture of the changes that may have taken place.
"It is impossible to quantify the impact these factors could have played in the apparent increases in walking and cycling. If the council had undertaken control counts at the same time, in areas of the city where 30mph speed limits were retained, the changes in the pilot areas could have been compared with those in the control areas. In the absence of such comparisons, and given the limitations of the survey data, the apparent increases in the pilot areas, even when properly calculated, must not be taken as evidence of increased walking and cycling in the 20mph speed limit areas."