Accident data and traffic flows, with and without roadworks present, have been collected, along with details of the roadworks, for a number of years in the county of Norfolk and the royal county of Berkshire. Consideration of the statistical requirements of the data set, and the likely costs of assembling the data, led to the proposal to collect some data `retrospectively' about past roadworks, and some data `current-passively' which involved the two counties assembling data about roadworks as they took place (as opposed to `current-actively' which would have involved the consultants collecting data about the roadworks as they occurred, which would have incurred greater cost). A sample of 155 roadworks was assembled. Of the 155 sites, 37 were in Berkshire and 118 in Norfolk. Taking the two counties together, 26 roadworks were on trunk roads, 54 on a-class roads, 45 on b-class roads, and 30 on c-class roads. Most of the 34 accidents recorded at roadworks occurred on single-carriageway roads. For single-carriageway trunk and a-class roads, an increase in the accident rate due to roadworks, over a length of road equal to the works zone and 1,200 metres either side, of 172% was identified, which was statistically significant at the 99% level of confidence. (A)
Samenvatting