A US study has shown that digital billboards do not represent a traffic hazard.
The purpose of the study was to examine the statistical relationship between certain digital billboards and traffic safety, and to determine if any correlation exists.
A study area was identified in Ohio, data was collected, and an analysis was made. Specifically, this study analysed the traffic and accident data near seven existing digital billboards on the 132.07 miles of Interstate routes in Cuyahoga County, Ohio. These seven billboards were located along Interstate routes I-77, I-90, I-271, and I-480.
In July 2005, the seven billboards were converted to digital from conventional format; a total of 335 million vehicles drove by these seven billboards in that year.
The analysis had two parts. In the first part, the temporal analysis, the occurrence of traffic accidents near the digital billboards is examined for an equal length of time before and after July 2005, for the purpose of establishing if traffic accidents occurred more or less frequently with the presence of the digital billboards. The second part, the spatial analysis, established statistical correlation coefficients between the digital billboards and accidents.
Correlation coefficients are statistical measures of the “association” between two sets of data, for example, billboards and traffic accidents. The results were analysed for various scenarios between accident density to sign density (the number of billboards), to Viewer Reaction Distance (the distance from a billboard that a driver was potentially within the “influence” of a billboard), and to sign proximity (the distance from the accident to the nearest billboard). In each scenario, the study considered accident data, with and without the bias from interchanges or known causes.
The conclusions of this study of Cuyahoga County indicate the following.
• At each of the digital billboards, and for periods of 12 months before and after the conversion (a total of 24 months), the accident statistics and metrics are consistent, exhibiting statistically insignificant variations. The same conclusion also applies for periods of 18 months before and after the conversion (a total of 36 months). The metrics include the total number of accidents in any given month, the average number of accidents over the 12- and 18-month periods, the peak number of accidents in any given month, and the number of accident-free months. These conclusions account for variations in traffic-volume and vehicle-miles traveled.
• The correlation coefficients demonstrate no statistical relationship between vehicular accidents and billboards (including conventional and the seven, digital billboards). Also, these correlation coefficients strongly suggest no causal relationship between the billboards and vehicular accidents.
• Accidents occur with or without billboards (digital or conventional). The accident statistics on sections of Interstate routes near billboards are comparable to the accident statistics on similar sections that have no billboards. The overall conclusion of this study is that digital billboards have no statistical relationship with the occurrence of accidents. The frequency of traffic accidents may be much more likely attributable to, and correlated with, other factors, such as DUIs, deer hits, adverse weather conditions, excessive speeding, inter alia.
