[Up: Table of Contents]
[Previous: Section IV]
[Next: Section V, part 2]
[This long section has been divided into three parts
for Internet publication]
|Bicycle/motor-vehicle accidents exhibit great diversity in the situations in which
they occur and the reasons for which they occur. When every case is viewed as a unique
event, the universe of bicycle/motor-vehicle accidents presents an overwhelmingly complex
picture to even the most capable researcher. The nature of the problem, and therefore
approaches to reducing the problem, simply cannot be comprehended without structuring the
universe of accidents in some meaningful way. The traditional approach to structuring
information about accidents is to examine the distribution of data one or two variables at
a time. The descriptive data presented in Section IV typifies the traditional analytic
approach. This approach is useful and necessary, but descriptive data seldom provide the
type of structure that stimulates innovative ideas about accident countermeasures.
Another approach to structuring a complex universe of objects or events is to develop a classification scheme that enables one to subdivide the universe of cases into mutually exclusive "sets" by grouping together objects or events that exhibit commonality in one or more of their attributes. Classification schemes have been developed and used since the days of the early Greeks (Crowson, 1969), and much of the progress in the physical and biological sciences can be attributed to this tool for scientific inquiry (Sokal, 1974). More recently, classification schemes have been developed and successfully used in the study of pedestrian accidents (Snyder & Knoblauch, 1971) and alcohol-related motor-vehicle accidents (Perchonok, 1975).
This section describes the results of a classification of bicycle/motor-vehicle accidents. To convey the full range of similarities and differences among accident cases, a hierarchical classification system was developed that consisted of problem classes, types, and subtypes. Problem classes reflect commonality at the most general level. Problem types represent variations of accidents within the same class, and subtypes represent variations of accidents within the same type. Problem types generally provide the most useful definition of a problem for which specific countermeasures can be tailored; but for some kinds of countermeasures, problem classes or problem subtypes may constitute a more meaningful problem definition.
ORGANIZATION AND CONTENT
For ease of exposition, problem types within the same class are discussed together in a separate subsection. Each subsection begins with a brief description of the distinguishing characteristics of the problem class and the similarities and differences among the problem types within that class. Then, each problem type and subtype in the class is described in turn.
The descriptions of Problem Types 1 through 25 are accompanied by perspective drawings that illustrate the traffic contexts in which the accidents occur and the proximal pre-crash paths of both vehicles. Some drawings illustrate two or more subtypes of the same problem type. The illustration of subtypes is accomplished by showing a separate set of vehicles (a bicycle and a motor vehicle) for each subtype. Each illustration shows the percentage of fatal and the percentage of non-fatal accidents accounted for by the problem type that is illustrated. When two or more subtypes are illustrated, percentage values are shown in close proximity to each vehicle set. These values show the percentage of cases within the problem type that is accounted for by each subtype; the combined percentage values for the subtypes shown on each illustration total 100%. Although the illustrations provide a useful aid in understanding how accidents of a given type occur, the reader is cautioned against using the illustrations to draw inferences about the characteristics of the roadway(s), the presence or absence of visual obstructions, the exact impact points, the exact collision points, and so on. The problem-type descriptions for each class are followed by a discussion of the educational countermeasures that appear to have potential for reducing the incidence of one or more problem types within that class.
CLASS A PROBLEM TYPES: BICYCLE RIDEOUT -- DRIVEWAY, ALLEY, AND OTHER MID-BLOCK
Table 14 lists the generic titles of the four Class A problem types and shows. the proportions of cases in the fatal and non-fatal samples that were classified into each problem type. The proportion of cases in the total class is shown at the bottom of the table.
PROBLEM CLASS A--BICYCLE RIDEOUT:
DRIVEWAY, ALLEY, AND OTHER MID-BLOCK
|TYPE 1||BICYCLE RIDEOUT: RESIDENTIAL DRIVEWAY/ALLEY, PRE-CRASH PATH PERPENDICULAR TO ROADWAY||6.7%||5.7%|
|TYPE 2||BICYCLE RIDEOUT: COMMERCIAL DRIVEWAY/ALLEY, PRE-CRASH PATH PERPENDICULAR TO ROADWAY||2.4%||3.2%|
|TYPE 3||BICYCLE RIDEOUT: DRIVEWAY/ALLEY APRON, PRE-CRASH PATH PARALLEL TO ROADWAY||2.4%||2.5%|
|TYPE 4||BICYCLE RIDEOUT: ENTRY OVER SHOULDER/CURB||3.6%||2.5%|
|TOTAL||CLASS (N: FATAL = 25; NON-FATAL = 105)||15.1%||13.9%|
|All Class A accidents occurred at a mid-block location shortly after the bicyclist
entered the roadway from a driveway, alley, or over a curb or shoulder. In almost every
case, the bicyclist entered the roadway without slowing, stopping, or searching for
oncoming traffic. Because of the bicyclist's sub-optimal pre-crash course (path and/or
speed), the motorist had insufficient time to avoid the accident once the bicyclist became
visible and the bicyclist's intended path became apparent to the
motorist. The function failures and contributing factors are similar for the four Class A
problem types. The main differences among the problem types are the type of location at
which the bicyclist entered the roadway, the factors that served to limit the operator's
preview time, and the bicyclist target
Problem Type 1 (6.7% Fatal; 5.7% Non-Fatal)
Figure 8 illustrates the traffic context and critical actions for Problem Type 1. Accidents of this type occur when the bicyclist rides straight out of a residential driveway or alley and collides with a motor vehicle approaching from the left or right. Figure 8 shows that 72% of the collisions occurred in the first half of the roadway (the half nearest the point at which the bicyclist entered the roadway); the remaining 28% occurred in the second half of the roadway.
Illustration of Problem Type 1,
Bicycle Rideout: Residential Driveway/Alley,
Pre-Crash Path Perpendicular to Roadway
|Problem Type 1 includes only the bicycle rideout accidents that occurred at the
junction of a roadway and a residential driveway (48%), a residential alley (33%), or a
driveway serving a rural residence (19%). Seventy-nine percent of the cases occurred on a
two-lane urban street with light traffic
and a posted speed limit of 25 MPH or less; 19% occurred on a two-lane rural roadway, and
two percent occurred on an urban street with more than two lanes.
Accidents of this type occurred almost exclusively during daytime hours, and the frequency
of occurrence was greatest in the afternoon; 95% of the cases occurred during the daytime
and 84% occurred between 2:00 PM and 7:00 PM.
A visual obstruction was a contributing factor in 63% of the accidents; parked motor vehicles and vegetation were the most common types of obstructing objects. When the operators' views were not obstructed, the accident was usually the result of one or both operator's failure to search in the direction of the other vehicle until an accident was imminent. In about nine percent of the cases, the motorist observed the bicyclist early enough to have avoided the accident but proceeded with the assumption that the bicyclist would slow or stop before entering the roadway.
The motorist's failure to search in the bicyclist's direction was usually due to his expectation that all traffic entering the roadway from intersecting driveways and alleys would yield the right of way. In short, the motorist did not search in the bicyclist's direction because he saw no necessity to do so in that traffic context. The factors that contributed to the bicyclist's failure to search are more numerous and complex. The most common contributing factors revealed by the interviews include:
Accidents of this type nearly always occurred close to the bicyclist's home; many occurred as the bicyclist was exiting the driveway serving his own residence. Consequently, most bicyclists were thoroughly familiar with the physical and operational characteristics of the accident location. Mainly because of his familiarity with the area, the bicyclist did not consider either the environment or his actions to be particularly hazardous. Therefore, risk assessment rather than risk acceptance must be considered an important factor for Problem Type 1. Although the bicyclists' actions would be perceived as risk-taking behavior by adults, it would be misleading to suggest that the bicyclists who were involved in this type of accident were any more willing to engage in risk-taking activities than the general population of bicyclists in the same age group.
Problem Type 1 involved bicyclists who were younger than those involved in any other problem type. The median age of the bicyclists was 9.8 years, and about five percent were five years of age or younger. Fewer than five percent of the bicyclists were 16 years of age or older.
Problem Type 2 (2.4% Fatal; 3.2% Non-Fatal)
As is shown in Figure 9, Problem Type 2 occurred in much the same way as Problem Type 1. The distinguishing characteristic of Problem Type 2 is that all the collisions occurred at the junction of a roadway and a commercial driveway (75%) or alley (25%). That is, the bicyclist rode straight out of a commercial driveway or alley into the approaching motor vehicle's path.
Illustration of Problem Type 2, Bicycle Rideout:
Commercial Driveway/Alley, Pre-Crash Path perpendicular to Roadway
|(NOTE: The building was drawn in the above illustration to indicate that this type of accident occurs at the junction of a commercial rather than a residential driveway/alley. Although a building sometimes obstructed the operator's view in accidents of this type, buildings were not the most frequent type of obstructing object.)|
|The accidents occurred with about equal frequency on two-lane urban streets (54%) and
urban streets with more than two lanes (42%). But, in either case, the roadway was usually
carrying moderate to heavy traffic at the time the accident occurred. Accidents of this
type nearly always occurred during the daytime (96%) and the frequency was clearly
greatest between 2:00 PM and 5:00 PM (58.4%).
In 39% of the cases, the motorist's preview time was critically limited by a visual obstruction. Parked motor vehicles, fences, and walls were the most common types of visual obstructions. The remaining 61% of the cases occurred even though the visibility conditions were good and the operators had a clear view of the other vehicle long before the collision occurred. About eight percent of the motorists observed the bicyclist in time to have avoided the accident but incorrectly assumed that the bicyclist would stop or turn at the junction. In about 42% of the cases, however, the motorist failed to search in the direction of the clearly visible bicyclist because he assumed that all traffic entering the roadway from intersecting driveways would yield to him.
The bicyclist's sub-optimal course and his failure to search were the result of a wide range of different factors. The most common are listed below.
There were few cases in which the presence of information overload could clearly be established from the interview data. That is, few bicyclists believed that their information processing capacity was severely taxed by the information processing requirements that existed at the time of the accident. Even so, it is believed that a substantial portion of the bicyclists were heavily loaded (if not overloaded) by the task of entering a heavily trafficked, multiple-lane roadway, and that information overload or attentional conflict often contributed to the bicyclist's search failure.
Although the bicyclists involved in Type 2 accidents were usually juveniles, there was a substantial number who were in their late teens or older. The median age of the bicyclists for this problem type was 13.8 years; five percent of the bicyclists were seven years of age or younger and five percent were 25 years of age or older.
Problem Type 3 (2.4% Fatal; 2.5% Non-Fatal)
Problem Type 3 is similar in many respects to Problem Types 1 and 2. As is illustrated in Figure 10, the distinguishing characteristic of Problem Type 3 is that the bicyclist entered the roadway from a parallel sidewalk by way of a driveway apron. About three-fourths of the collisions occurred in the near lane(s) and one-fourth occurred in the far lane(s). Problem Type 3 includes accident cases that occurred at either a residential or a commercial driveway, but most accidents (89%) occurred at a residential driveway. (in this respect, Problem Type 3 is most similar to Problem Type 1.) Eighty-four percent of the collisions occurred on a two-lane residential street; the remaining 16% occurred on a roadway with more than two lanes. Eighty-nine percent of the accidents occurred during the daytime; 63% occurred between 2:00 PM and 7:00 PM.
Illustration of Problem Type 3,
Bicycle Rideout: Driveway/ Alley, Pre-Crash Path Parallel to Roadway
|Like the previous two problem types, there were many cases (47%) in which the
bicyclist's pre-crash course combined with visual obstructions to limit the motorist's
preview time to such an extent that there was no chance to avoid the accident once the
bicyclist emerged from behind the obstructing object. In 22% of the cases, however, the
motorist observed the bicyclist early enough to have avoided the
accident, but incorrectly assumed that the bicyclist would continue riding on the
sidewalk. In 17% of the cases, the bicyclist was visible, but the motorist failed to
search in his direction because he assumed that all intersecting traffic would yield to
Even when visual obstructions were present, there were many instances in which the bicyclist could have observed the motor vehicle early enough to have avoided the accident. Thus, search failures accounted for 72% of the bicyclist's precipitating function failures. Most of the bicyclists' search failures were due to the presence of some type of distractor. The most frequent distractors were interacting with another person (36%), play activity (27%), and non-traffic-related mental activity (18%). In 18% of the cases, the bicyclist failed to search because he incorrectly assumed that a riding companion would search for hazards and select a safe course through the accident area.
The bicyclists who were involved in Type 3 accidents were slightly older than those involved in Type 1 accidents but were younger than those involved in Type 2 accidents. For Problem Type 3, the median age of the bicyclists was 11.5 years; about five percent of the bicyclists were five years of age or younger and about five percent were 16 years of age or older:
Problem Type 4 (3.6% Fatal; 2.5% Non-Fatal)
All Type 4 accidents occurred shortly after a bicyclist entered the roadway over a curb (74%) or shoulder (26%) at a mid-block location. Thirty-seven percent of the bicyclists stopped or slowed before entering the roadway; the remaining bicyclists made no attempt to slow their speed. As is shown in Figure 11, the bicyclist's pre-crash path was sometimes parallel to the roadway (42%) and sometimes perpendicular to it (58%). This type of accident most often occurred on a two-lane urban street (74%), but occasionally occurred on an urban street with more than two lanes (10%) or on a rural roadway (16%). Ninety-five percent of the accidents occurred during the daytime; 68% occurred between 3:00 PM and 6:00 PM.
The motorist's preview time was critically limited by visual obstructions in 41% of the cases; a parked motor vehicle was the most common type of visual obstruction. In 32% of the cases, the motorist observed the bicyclist well in advance and could easily have avoided the accident had he known that the bicyclist would enter the roadway. In the remaining 21% of the cases, the motorist failed to search in the bicyclist's direction and therefore failed to observe the bicyclist (clearly visible) until it was too late to avoid the accident.
The objects that obstructed the motorist's view also obstructed the bicyclist's view in many instances (26%), but in the majority of cases, the bicyclist made no attempt to search in the motorist's direction before entering the roadway (53%). Of the factors that were found to contribute to the bicyclists' function failures, 67% were found to be distractions of one type of another. A wide range of distractors were revealed by the data, but there was no single type of distractor that was clearly more important than any other.
Illustration of Problem Type 4,
Bicycle Rideout: Entry Over Shoulder/Curb
|Surprisingly, there were few bicyclists who reported that they were distracted by the
act of riding over the curb or shoulder. It seems almost certain that most bicyclists'
attention would be focused on the curb/shoulder they are preparing to ride over; the
closer the bicyclist's position to the curb/shoulder, the more his scan would be directed
downward. Thus, although not directly supported by the data, it seems reasonable to assume
that the bicyclist's failure to search was often due, in part, to the distractions
inherent in the act of riding over a curb or shoulder.
EDUCATIONAL COUNTERMEASURES FOR CLASS A PROBLEM TYPES
It seems clear that the education and training of motorists and bicyclists would prove effective in reducing the incidence of all four problem types within Class A. However, it is also possible that educating and training could prove effective for the parents of juvenile bicyclists, law enforcement officers, and bicycle-design engineers. The objectives of an education and training program for each of these groups is discussed briefly below.
If education and training of bicyclists is to be effective in reducing Class A accidents, it must be administered at a very early age -- preferably in kindergarten and certainly not later than the fourth grade. For instance, consider the age of the bicyclist for Problem Type 1. The data show that more than five percent of the Type 1 accidents involved bicyclists who were five years of age or younger, and 25% of the cases involved bicyclists who were younger than eight years of age. The age of the 5th and 25th centile bicyclist for the other three Class A problem types is only one or two years older than for Problem Type 1. Clearly, the requirement to impart, to very young children, the knowledge and skills necessary to avoid Class A accidents represents a formidable task.
There were very few instances in which a bicyclist rode into a motor vehicle's path because he misjudged the motor vehicle's approach velocity. Therefore, it seems reasonable to assume that most Class A accidents would be avoided if the bicyclist could be taught to stop at the edge of the roadway and search carefully for oncoming motor vehicles. In fact, substantial gains would probably be achieved if the bicyclist could merely be induced to stop at the junction or slow his speed considerably, thereby giving the motorist sufficient time to observe the bicyclist and initiate evasive action. To counter Class A accidents, an ideal educational program for young bicyclists would accomplish at least the following:
This study revealed no indication that the motorists who were involved in Class A accidents were atypical in their skills or their concern for safety. Even so, it is possible that some accidents of this type could be avoided if the general motoring public was informed of the frequency with which Class A accidents occur, where they occur, and the reasons for which they occur. The main objectives of an education and training program for the general motoring public would be to:
The education of parents of bicyclists in the target group could result in parents' assuming more responsibility for the bicyclists' training and, more importantly, a greater degree of parental control of where and how young bicyclists are permitted to ride. Casual observation indicates that most parents generally recognize that riding a bicycle may be dangerous for very young children, but few parents appear to have a clear understanding of the types of locations where bicycle/motor-vehicle accidents occur or the types of bicyclist actions that most often lead to such accidents. It is altogether possible that misinformed parents may be giving their children instructions that are counterproductive. For instance, the instruction to "ride close to home" may cause the bicyclist to ride in an area that is less safe than available alternative areas.
The main objective of a parent-education program is to inform parents of the frequency with which Class A accidents occur, how they occur, and why they occur. If parents are to be effective in educating their children, they must have a clear understanding of the function failures and contributing factors that lead to an accident. It is particularly important that parents understand that quiet neighborhood streets and thorough familiarity with the area do not ensure the bicyclist's safety.
Law Enforcement Officers
Educating patrol officers about the importance of Class A accidents and the reasons for which they occur could prove useful in curtailing the behavior that leads to these types of accidents. That is, an understanding that many bicycle/motor-vehicle accidents occur as the bicyclist enters the roadway would increase the likelihood that an officer would observe and issue citations to bicyclists who enter the roadway in an unsafe manner. However, an education and training program for law enforcement officers must be preceded by the passage of ordinances that make unsafe entry into the roadway unlawful.
A first step in the development of methods to increase the vertical dimension and conspicuity of bicycles would be to educate bicycle-design engineers about the need for such devices. Thus, persons who are involved directly or indirectly with bicycle design should be educated on the importance of Class A accidents and the nature of the accident-generation process for these types of accidents.
CLASS B PROBLEM TYPES: BICYCLE RIDEOUT -- CONTROLLED INTERSECTION
Table 15 lists the problem types within Class B and shows the relative frequency with which they occurred. The distinguishing characteristic of all Class B problem types is that the bicyclist entered a controlled intersection in an unsafe and usually unlawful manner. In all Class B accidents, the motorist and bicyclist were traveling on orthogonal legs of the intersection.
PROBLEM CLASS B -- BICYCLE RIDEOUT: CONTROLLED INTERSECTION
|TYPE 5||BICYCLE RIDEOUT: INTERSECTION CONTROLLED BY SIGN||7.8%||10.2%|
|TYPE 6||BICYCLE RIDEOUT: INTERSECTION CONTROLLED BY SIGNAL, SIGNAL PHASE CHANGE||.6%||3.1%|
|TYPE 7||BICYCLE RIDEOUT: INTERSECTION CONTROLLED BY SIGNAL, MULTIPLE THREAT||2.4%||
|OTHER CLASS B||BICYCLE RIDEOUT: INTERSECTION CONTROLLED BY SIGNAL, OTHER||1.2%||1.7%|
|TOTAL||CLASS (N: FATAL = 20; NON-FATAL = 128)||12.0%||17.0%|
Problem Type 5 (7.8% Fatal; 10.2% Non-Fatal)
Problem Type 5 includes "bicycle rideout" accidents that occurred at a signed intersection. The approach leg traveled by the bicyclist was controlled by a "stop" sign in 96% of the cases and a "yield" sign in only four percent of the cases. The approach leg on which the motorist was traveling was uncontrolled, except for three percent of the cases which occurred at an intersection controlled by a four-way stop sign. Eighty-two percent of the bicyclists entered the intersection without slowing or stopping; 18% slowed significantly or stopped at the intersection before riding into the path of the oncoming motor vehicle. About six percent of the motorists were traveling at a speed that exceeded the posted limit, but in the remaining cases, the motorist's speed was judged to be well within the normal range.
Seventy-five percent of the cases occurred at the junction of a pair of two-lane streets. In 17% of the cases, the motorist was traveling on a four-lane street and the bicyclist was traveling on a two-lane street. The remaining cases occurred at the junction of a pair of four-lane streets (4%) or at the junction of a pair of two-lane rural roadways (4%). Most accidents occurred during the daytime (94%) and they occurred with about the same frequency throughout the period between 7:00 AM and 7:00 PM.
Figure 12 shows that 22% of the bicyclists were riding facing traffic prior to the accident. Riding facing traffic was an important contributing factor because it decreased the likelihood that the bicyclist would be detected by the motorist in this situation. But, the most critical factor was the bicyclist's failure to slow or stop at the junction. That is, riding facing traffic contributed to the accident only because the bicyclist failed to stop at the junction.
It can be seen in Figure 12 that almost two-thirds of the collisions occurred before the bicyclist reached the center of the roadway. This finding can be attributed to the fact that motorists approaching from the left, in the near traffic lane(s), have very little time to initiate evasive action once it becomes apparent that the bicyclist does not intend to stop. Motorists approaching from the right have more time to respond because the bicyclist must travel across an entire traffic lane before he intersects the motor vehicle's path.
Seven percent of the cases classified into Problem Type 5 were "multiple-threat" accidents -- a variation of Problem Type 5 that is not portrayed in Figure 12. In these cases, a motorist observed the bicyclist and slowed or stopped to let him pass. The bicyclist observed the motorist slow or stop, assumed it was safe to cross the roadway, and proceeded into the intersection where he collided with a second motor vehicle. Every case of this type occurred in California where motorists are accustomed to yielding the right of way to pedestrians. Apparently, the motorists in these cases treated the bicyclist as a pedestrian rather than as a vehicle operator.
Illustration of Problem Type 5,
Bicycle Rideout: Intersection Controlled by Sign
|The motorist's view of the bicyclist was obstructed in about 31% of the cases-usually
by vegetation. It was surprising to find that parked motor vehicles obstructed the
operator's view in only three percent of the cases. About five percent of the motorists
failed to detect the approaching bicyclist because of darkness, inadequate bicycle
lighting, or both. In all of the cases that involved obstructions or degraded visibility,
it was judged that the motorist's preview time was critically limited and that the
accident was imminent at the point at which the bicyclist could first have been
The motorist had sufficient preview time to have avoided the accident in the majority of cases. The motorist failed to search in the direction of the bicyclist (clearly visible) in about 40% of the cases. The motorist's search failure was usually because he assumed that all intersecting traffic would yield the right of way to him, or because the bicyclist was riding in an unexpected location (wrong side of street). In 13% of the cases, the motorist observed the bicyclist soon enough to have avoided the accident, but failed to initiate evasive action because he assumed the bicyclist would slow, stop, or turn at the intersection.
The bicyclist's speed control at the intersection is a critical factor in explaining his role in Type 5 accidents. The classification of cases in terms of the bicyclist's speed control at the junction revealed the following variations or subtypes for Problem Type 5:
The bicyclist's function failures are discussed for each of these variations of Problem Type 5.
First, consider the accidents in which the bicyclist stopped at the junction and concluded that it was safe to proceed (13%). More than half of these accidents were multiple-threat accidents (described above); the remainder involved a bicyclist who failed to search properly (3%) or who misjudged the motor-vehicle's approach speed (3%). Next, consider the cases in which the bicyclist slowed significantly and concluded it was safe to proceed (5%). These accidents were due to the bicyclist's failure to search effectively or his failure to take into account the presence of visual obstructions.
Finally, consider the accidents in which the bicyclist clearly failed to slow his speed. In 7.8% of the cases, the bicyclist attempted to stop at the junction but was unable to do so because of a skill deficiency, defective brakes, wet caliper brakes, wet pavement, or a combination of these. The bicyclist in these cases misjudged his ability to manipulate the brakes or misjudged stopping distance under the conditions that existed at the time of the accident. In 74% of the cases, the bicyclist made no attempt to stop or slow prior to entering the intersection. The interview data clearly show that the bicyclist's failure to stop or slow at the intersection was not the result of his failure to observe the stop sign. The accidents nearly always occurred at an intersection through which the bicyclist had ridden many times before the accident, so most bicyclists knew perfectly well that a sign was present at that location. Furthermore, it is clear that the bicyclist's failure to stop was not the result of ignorance of the law. Even the youngest bicyclist admitted knowing that the law requires bicyclists to stop for stop signs and to yield the right of way at intersections controlled by a yield sign. So, failure to observe traffic signs and ignorance of the law definitely are not important contributing factors for Problem Type 5.
Of the bicyclists who failed to slow or stop, it was judged that nearly 70% could have avoided the collision if they had searched in the direction of the motor vehicle prior to entering the intersection. In the remaining cases, because of the combined effects of the bicyclist's speed and an obstructed view, it was judged that the bicyclist could not have avoided the accident at the point where the motor vehicle first could have been observed. The bicyclist's failure to slow or stop and his failure to search must be explained in terms of the following factors:
Although a variety of factors contributed to the bicyclist's failure to stop at the intersection, it appears that faulty risk assessment was an overriding factor in most cases. This opinion is based upon three facts. First, most accidents occurred at a relatively safe-appearing intersection; in most cases, the operators were traveling residential roadways on which both traffic volume and operator speeds were low. Secondly, most accidents occurred at an intersection that the bicyclist had ridden through many times before the accident -- probably without stopping in many instances. Third, the bicyclists' self-ratings provided no indication that their actions were due to a high willingness to accept risks. For these reasons, it seems reasonable to assume that the overriding reason for most bicyclists' failure to stop was their expectation that the roadway would be void of traffic. Although few bicyclists admitted to this fact during the interviews, it is to be expected that bicyclists would be reluctant to report such an unrealistic expectation.
Although bicyclists of all ages frequently fail to stop or slow at signed intersections, Type 5 accidents nearly always involved a juvenile bicyclist. The median age of the bicyclists involved in this type of accident was 11.8 years; less than 25% of the bicyclists were older than 14 years of age and about five percent of the bicyclists were older than 18 years of age.
Problem Type 6 (.6% Fatal; 3.1% Non-Fatal)
All accident cases classified into Problem Type 6 occurred at a signalized intersection. Eighty-three percent of the accidents occurred as the bicyclist was crossing an intersecting street with four or more traffic lanes. Although the majority of these accidents occurred during the daytime, 17% occurred during darkness. About 70% of all Type 6 accidents occurred during the period between 1:00 PM and 7:00 PM.
The distinguishing characteristic of Problem Type 6 is that the bicyclist entered the intersection as the signal phase was changing and failed to clear the intersection before the signal turned red. In all cases, the motorist entered the intersection after the signal controlling his approach had turned green. Problem Type 6 does not include cases in which the bicyclist entered the intersection more than one or two seconds after the onset of the red-signal phase. In addition, Problem Type 6 does not include "multiple-threat" accidents. Multiple-threat accidents were classified into Problem Type 7 and are described below. As is shown in Figure 13, 38% of the collisions occurred before the bicyclist reached the center of the roadway he was crossing; the remaining 62% occurred in the second half of the roadway the bicyclist was crossing.
Illustration of Problem Type 6,
Bicycle Rideout: Intersection controlled by Signal, Signal Phase Change
|In 78% of the cases, the motorist failed to search in the bicyclist's direction until
it was too late to avoid the accident. In the remaining cases, the motorist either (a)
searched adequately but failed to detect the bicyclist because of darkness, inadequate
bicycle lighting, or both (4%), or (b) searched for and detected the bicyclist soon enough
to have avoided the accident but assumed the bicyclist would stop or slow before entering
the motor vehicle's path (13%). The motorist's failure to search in the bicyclist's
direction was due partly to his faulty assumption that all intersecting traffic would
yield to him and partly to information overload. It is clear that the motorist's
information-processing capacity was heavily loaded by the requirement
to watch the signal, search for pedestrian and vehicle traffic, control the speed and
position of his vehicle, and so on.
Nearly 57% of the bicyclists failed to search in the direction of the motor vehicle until an accident was imminent; 30% of the bicyclists observed the motor vehicle but assumed it would stop or remain stationary until the intersection was clear. Only four percent of the accidents were due to an action failure by the bicyclist. The evidence available for this problem type indicates that some bicyclists failed to stop at the intersection because they were unaware that the signal had changed since they last checked it. Other bicyclists knew that the signal had changed but assumed they could clear the intersection before the termination of the amber phase. However, because admitting to trying to beat the red light is more incriminating than admitting to a failure to notice the signal phase change, it is not possible to estimate accurately the relative proportion of the bicyclists who made each type of error. However, it was found that 16% of the bicyclists were following a riding companion whom they assumed would search for hazards and select a safe course.
Because of the complexity of the traffic context and the usually high speed of the bicyclist, it is assumed that information overload contributed to the bicyclist's failure to carefully monitor the traffic signal, to search for approaching traffic, or both.
The relatively low incidence of fatal accidents for Problem Type 6 is due to the low motor-vehicle speeds at impact. Because the collision occurred as the signal phase was changing, the motorist was either accelerating from a stopped position or, more commonly; had slowed to a low speed for the red signal and accelerated when the signal turned green a moment before the collision.
About half the bicyclists involved in Type 6 accidents were juveniles, and half were young adults or adults. The median age of the bicyclists was 16.1 years; about 15%% were 18 years of age or older. Only five percent of the bicyclists were younger than 11 years of age. As a group, the bicyclists involved in Type 6 accidents were considerably older than those involved in any of the problem types discussed previously.
Problem Type 7 (2.4% Fatal; 2.0% Non-Fatal)
Problem Type 7 is highly similar to Problem Type 6 with respect to target location, target period, and the nature of the bicyclist's pre-crash course. Problem Types 6 and 7 differ in one important respect. For Problem Type 7, the bicyclist's decision to proceed across the intersection was influenced by the presence of other motor vehicles that were stopped at the intersection, apparently waiting for the bicyclist to pass. The nature of the accident-generation process for Problem Type 7 is illustrated in Figure 14. It can be seen that 14% of the accidents occurred in the first half of the roadway and involved a bicyclist who was riding facing traffic. The remaining 86% of the cases occurred in the second half of the roadway and involved a bicyclist who was riding on the correct side of the street. In all cases, the bicyclist passed in front of one or more stopped vehicles before colliding with the accident vehicle.
Illustration of Problem Type 7,
Bicycle Rideout: Intersection Controlled by Signal, Multiple Threat
|Standing motor vehicle(s) obstructed the motorist's view of the bicyclist in 53% of
the cases. In these cases, there was no chance for the motorist to initiate successful evasive action once the bicyclist emerged from behind the stopped
vehicles. In 40% of the cases, it was judged that the motorist could have observed the
approaching bicyclist, but he failed to search in the bicyclist's direction. In about
seven percent of the cases, the motorist searched in the bicyclist's direction but failed
to detect the bicyclist because of darkness, inadequate bicycle lighting, or both.
The standing motor vehicle(s) obstructed the bicyclist's view of the approaching motor vehicle in nearly 27% of the cases. Given the speed the bicyclist was traveling prior to the collision, it was judged that there was insufficient time to have avoided the accident once the bicyclist first could have observed the motor vehicle. In 40% of the cases, it was judged that the bicyclist could have observed the approaching motor vehicle early enough to have avoided the accident but failed to search in the direction of the motor vehicle until an accident was imminent. In about one-third of the cases, the motor vehicle was stopped at the intersection and was observed by the bicyclist long before the accident; the bicyclist proceeded with the assumption that the stopped vehicle would remain stationary until he had passed.
Unlike Problem Type 6, it was found that only 20% of the bicyclists underestimated the length of the amber phase. Most bicyclists were perfectly aware that the amber phase was about to terminate but assumed that all motor-vehicle traffic would remain stationary or yield to them.
The bicyclists age distribution for Problem Type 7 was similar to that for Problem Type 6. The median age of the bicyclists was 15.2 years; about 25% were 16 years of age or older. Only five percent of the bicyclists were younger than 12 years or older than 33 years of age.
Other Class B (1.2% Fatal; 1.7% Non-Fatal)
The sample contained a small number of cases in which the bicyclist entered a signalized intersection well after the onset of the red-signal phase. Because of the small number of such cases and because of the lack of commonality in the accident-generation process, it was not possible to define one or more clear-cut problem types for these cases. Therefore, the cases were classified into "Other Class B."
If the data base for bicycle/motor-vehicle accidents is expanded in the future, it is probable that at least three additional Class B problem types would be revealed. One type would include cases in which a bicycle failure or a skill deficiency prevented the bicyclist from stopping for the red signal. A second type would include cases in which the bicyclist was suffering from a physical or mental impairment (particularly alcohol) and therefore failed to monitor the signal carefully. A third type would include cases in which the bicyclist knowingly failed to stop at the intersection because he assumed he could successfully dodge or otherwise evade approaching motor vehicles. Examples of each of these types of accidents were found among the cases classified into "Other Class B." However, the findings of the present study indicate that such problem types would occur infrequently. The present data, and other samples of accident reports that have been examined by the author, indicate that few bicycle/motor-vehicle accidents occur when bicyclists enter an intersection when the signal is clearly red. Although most readers know that failing to stop for a red signal is not at all uncommon for bicyclists, the bicyclists who engage in this hazardous activity apparently exercise a good deal of caution when doing so.
EDUCATIONAL COUNTERMEASURES FOR CLASS B PROBLEM TYPES
The evidence is clear that Type 5 accidents seldom occur when the bicyclist stops or slows his speed significantly before entering an intersection controlled by a stop or yield sign. Although it is necessary for bicyclists to search for and evaluate the closing velocity of approaching motor vehicles, bicyclists usually perform the search and evaluation functions in an adequate manner when they consider it necessary to slow or stop at an intersection. Thus, a primary goal of countermeasures for Problem Type 5 is to induce bicyclists to slow their speed considerably or, preferably, come to a complete stop before entering a signed intersection. The other objective of countermeasures for Problem Type 5 is to teach bicyclists to avoid multiple-threat accidents at signed intersections.
The objective of countermeasures for Problem Types 6 and 7 is to prevent bicyclists from entering a signalized intersection when it is not possible for them to clear the intersection before the termination of the amber phase. An additional objective for Problem Type 7 is to teach bicyclists and motorists to avoid multiple-threat accidents at signalized intersections. The objective of countermeasures for Other Class B accidents is to prevent bicyclists from entering a signalized intersection against a red signal.
A careful study of the accident-generation process for Problem Types 5, 6, and 7 shows that these accidents were seldom due to the bicyclist's willingness to accept an uncommonly high degree of risk, and were never due to the bicyclist's misunderstanding of the laws governing behavior at controlled intersections. Rather, the bicyclist's critical actions were primarily due to misjudgment of the risk associated with the critical action, misjudgment of the length of the amber phase, failure to recognize a "multiple-threat" situation, competing needs, and momentary distractions. Therefore, an educational program for bicyclists must be developed to accomplish the following objectives:
If the education is to be received before a significant number of accidents already have occurred, education to curtail Type 5 accidents must be introduced during the second or third grade (7- or 8-year-old bicyclists). Education to curtail Types 6 and 7 accidents may be delayed until the fifth or sixth grade (10- or 11-year-old bicyclists) without sustaining significant losses.
An education program that would serve to increase motorists' awareness of multiple-threat situations may prove beneficial in reducing multiple-threat accidents, particularly at signed intersections. Certainly, motorists in standing vehicles should be taught to always check for other approaching motor vehicles before motioning bicyclists to cross in front of them. It may be possible to develop a standardized hand signal or horn signal that motorists can use to inform bicyclists that it is not safe to pass. Also, some benefit may result from educating motorists that slow-moving bicyclists may not have enough time to clear the intersection during the amber phase.
[Up: Table of Contents]
[Previous: Section IV]
[Next: Section V, part 2]
Contents copyright © 1978,