2. Methods
2.1. Location
This study was conducted in February 2009 in a southern
prefecture-level city of China, Zhongshan, which is situated in the
Pearl River Delta, with an area of 1679km2 and a population of
2.49 million. It is composed of 5 districts (and Xiao Lan town) that
can be classified as urban locations and 17 towns classified as rural
locations. In 2008 the city ranked 15th in a list of more than 200
cities in China by per capita income. In November 1999, to restrict
the increasing number of motorcycles, the Zhongshan government
ceased registration of new motorcycles plates. The number of registered
motorcycles has thus remained constant at approximately
350,000 (Zhongshan Statistic Bureau, 2009), with approximately 1
in 7 citizens owning a motorcycle. However this number does not
include the unregistered motorcycles.
2.2. Roadside observation
In February 2010, a cross-sectional on-site observational survey
was implemented based on a random sample of the road hierarchy
in the city. At each of 20 randomly selected observation sites, data
was collected for 3 separate hours at daytime (7 am–7pm).
2.2.1. Site selection
To select the roadside sites, a grid with sections measuring 2 cm
by 2 cmwas placed over a standardmapof Zhongshan and drawn to
a scale of 1 cm= 1.85 km. Each grid box contained an area of approximately
3.42km2. The grid dimensions were 26 lines horizontally
and 18 lines vertically. Each valid grid box containing at least one
intersection in an included part of Zhongshan was numbered. The
selection of sites was based on proportion of the population in
urban and rural areas of the city to avoid the sites over representing
the non-urban areas. In this way 8 sites from urban areas (including
4 districts and Xiao Lang town) and 12 sites from rural areas were
selected. Random numbers were generated to select grid boxes
from which specific intersections were selected. For each site an
alternate site was also selected randomly from the box as a backup.
Researchers visited and evaluated each site prior to the study to
determine whether the site was suitable for the observation. The
availability of traffic lights or stop signs and safety of observers
were the main considerations. If the original site did not meet the
requirements, the researchers would visit and evaluate the alternate
site. If neither the original site nor the alternate site met the
requirements, the researchers then identified another suitable site
that was closest to the original site (the third site). Among the 20
observational sites, 9 alternate sites and 1 third site were used. The
day of week and time of day for site observation were randomly
assigned to sites in such a way that all days of the week and all
daylight hours (7am–7pm) had equal probability of selection