Bicycles/Pedestrians

Safety Toolbox: Problem identification

What to look for

Risks and trends

View a summary of bicyclist and pedestrian injury rates for the Bay Area.

The analysis of pedestrian and bicyclist-involved collisions is challenging for several reasons:

Pedestrian and bicyclist-involved collisions typically account for a very small fraction of all collisions;
A number of collisions go unreported;
There is limited data on the number of pedestrians and bicyclists (the "denominator problem"); and,
Unlike with vehicle-vehicle collisions, there are no standard formulas for calculating collision rates.

The typical response to the dearth of data and guidance on how to analyze pedestrian and bicyclist-involved collisions has been to focus on locations that have experienced the highest number of collisions. In the scale of pedestrians and bicyclist-involved collisions, high collision location means one or more collisions per year. The following presents alternative parameters that may be considered in addition to high collision locations. They focus on the basic questions of who, what, when, and where.

A few points to keep in mind:

Most of the analyses described below are best undertaken on a wide-area basis.
A combination of tabular and spatial analyses is advisable.
Data may need to be manipulated prior to conducting an analysis. In some cases, data may need to be grouped, for example when looking at age. In other cases, data may need to be "cleaned", for example when looking at vehicle code violations where small variances in data entry could result in erroneous results.

Who?

The primary goal is to find out if there are patterns in the demographics of victims and offenders. Some examples of this type of analysis include:

If using SWITRS data, the analyst may look for patterns in party-at-fault and victim demographics: age group, gender, and impairment (physical or due to drugs/alcohol).
A finer level of detail may be obtained from actual police reports (CHP 555), where race/ethnicity and zip code are also available. Zip code analysis can inform whether victims or offenders are mostly local residents or from out of the area. Analysts may also overlay findings with census tract information to determine if there are patterns associated with environmental justice concerns, such as level of education, poverty, etc.
Data pertaining to severity of injuries from SWITRS may also be cross-referenced against information from public health databases. In trauma data, a scoring system is used to quantify the extent of injuries to particular body regions (head, face, chest, abdomen, extremities including pelvis, external) and to calculate overall summary or severity scores. The values of these scores can be used to correlate with a person’s risk of mortality. For example, each type of injury encountered is assigned a value from 1 to 6, correlating to: minor injury, moderate injury, serious injury, severe but not life-threatening injury, critical with uncertain survival, and unsurvivable injury. Injury severity is also directly correlated to the cost of medical care and indirect costs due to reduced productivity and quality of life lost. Therefore, data on injury severity can be used to estimate the total costs of injury.

What?

The goal is to find out if there are patterns in the types of collisions. If using SWITRS, there are several parameters that may be investigated: primary collision factor, crash type, movement preceding collision, and pedestrian's actions. Following are typical California Vehicle Code violations associated with pedestrian and bicyclist-involved collisions, categorized based on which party (motorist, bicyclist, or pedestrians) was found to be at fault. This can be used as a starting point for the analysis involving primary collision factors:

Common bicyclist-at-fault violations

Common motorist-at-fault violations (involving bicyclists)

Common pedestrian-at-fault violations

Common motorist-at-fault violations (involving pedestrians)

Relevant SWITRS crash types for pedestrian and bicyclist-involved collisions include but are not limited to: vehicle/pedestrian, sideswipe, and broadside.

Relevant SWITRS designations for movements preceding bicyclist-involved collisions include but are not limited to: making right turn, changing lanes, parking maneuver, entering traffic, parked, and traveling wrong way. The movements of a pedestrian prior to a collision are designated in the pedestrian's actions field of SWITRS.

When?

The goal is to find out whether there are patterns to the conditions under which pedestrian and bicyclist-involved collisions occur. Following are some parameters that may be investigated, including findings from a comprehensive collision analysis conducted for four Bay Area cities:

Time of day - Most collisions occurred between 2 p.m. and 7 p.m.
Day of the week - Most collisions occurred on weekdays.
Time of year - Majority of bicyclist-involved collisions occurred during the Summer months and pedestrian-involved collisions during the Winter months.
Lighting conditions - Most collisions occurred in daylight hours.
Weather and Surface conditions - Most collisions occurred during dry weather and surface conditions.
Road conditions - Most collisions occurred when there were no unusual roadway conditions, i.e. no construction.

Where?

The goal is to find out whether there are patterns to the locations where pedestrian and bicyclist-involved collisions occur. This type of analysis is best undertaken using a GIS, which can easily reveal higher concentrations of collisions. It may also be undertaken as a secondary level of analysis if and when patterns are identified under the who, what, and when categories. Once collisions are mapped, following are some examples of parameters to consider:

High-incident locations
Collision concentration in high-use areas such as downtown areas, schools, senior centers
Dispersion of certain types of collisions over wide areas. For example, red light running collisions that are scattered throughout a jurisdiction may indicate the need to adopt policies regarding all-red phases at signalized intersections.
Intersection versus non-intersection collisions
Facility types: arterials, freeway interchanges, collectors, local streets
Environment: rural versus suburban versus urban locations
Roadway alignment, especially east/west streets where the sun may play a factor