Education objectives for traffic and transportation engineers are not discussed because of the raging controversy surrounding the safety benefits to be derived from modifying the traffic system to better accommodate bicycles. Similarly, this section contains no discussion of educational objectives for legislators, officials of governmental agencies, traffic-safety researchers, school administrators, classroom instructors, or the many other persons who may benefit from education about the incidence, consequences, and causes of bicycle accidents. The discussion of educational objectives for these persons must be left for another time and another report.

The reader should keep in mind that the educational objectives discussed in the following pages are based almost entirely on a consideration of the behavioral causes of bicycle/motor-vehicle accidents. Hopefully, the type of education that will effect a reduction in bicycle/motor-vehicle accidents will also serve to reduce NMV accidents. However, it is almost certain that the set of educational objectives presented here will have to be expanded once detailed data on the causes of NMV accidents become available.

SOURCES OF CONTROVERSY ABOUT EDUCATIONAL OBJECTIVES

A careful review of bicycle-safety education programs developed in the recent past reveals a great many differences in the objectives that the programs were designed to accomplish [6]. Many of the differences are the direct result of the lack of valid information about the specific knowledge and skill deficiencies that lead to bicycle accidents. Other reasons for the differences are discussed below.

MOTIVES OTHER THAN PROMOTING SAFETY

It is apparent from an examination of existing educational materials that the promotion of safety was not always the sole motive in operation when the materials were developed. Other apparent motives include: promoting greater bicycle usage, increasing bicyclists' ability to ride efficiently, modifying patterns of bicycle usage, and modifying attitudes toward the bicycle and bicycle users. There are few who believe that the inclusion of such topics in an education program will serve to reduce accidents. Rather, they are included because of the belief that (a) greater and more efficient use of the bicycle will result in societal benefits that are at least as great as reducing accidents, and (b) a bicycle-safety education program provides a convenient vehicle for promoting bicycling and teaching bicyclists to ride more efficiently. Those who oppose the inclusion of such topics argue that the time and resources available for safety education are so limited that a safety-education program should be limited to the topics and activities that are known to reduce accidents.

FAILURE TO DEFINE UNDERLYING RATIONALE

It is believed that much of the controversy about educational objectives stems from the failure to define the rationale that led to the selection of a specific educational objective. This is particularly true when the objective is to enhance rudimentary knowledge or skills that a student must possess before he can be taught more complex and more directly relevant concepts or skills. For example, on first examination, it may be difficult to see why a bicycle-safety education program would include material about the functioning of the human visual system. However, the usefulness of this information quickly becomes apparent when it is explained that some knowledge of the functioning of the visual system is required to teach students why bicyclists sometimes fail to observe clearly visible cues to hazard, why motorists sometimes fail to observe clearly visible bicyclists, and so on. It is believed that much of the controversy about educational objectives would vanish if each educational objective was accompanied by a brief description of the rationale that led to the establishment of that objective. To be effective, the description should be complete enough to enable readers to easily perceive the links) between an educational objective and the accident-producing behavior to be modified.

ASSUMPTIONS ABOUT THE TARGET GROUP

There appears to be almost universal agreement that bicyclists in general and juvenile bicyclists in particular constitute the primary target group for bicycle-safety education. However, there is no universal agreement about the specific age at which bicycle-safety education must be introduced. Some persons believe that bicycle-safety education should commence in kindergarten and continue in every grade through high school. Others believe that it is futile to attempt to educate very young children, so have developed programs only for older children. Obviously, the specific objectives of a program are going to vary greatly as a function of the age at which bicycle-safety education is to be introduced.

To complicate matters even more, the objectives of programs developed for the same young age group often differ because of differences in opinion about what young children can and cannot be taught. Child development experts agree that certain concepts simply cannot be learned by children before they reach a certain maturational level, no matter how much effort is expended in trying to teach the child the concepts. However, child development is not well enough understood to enable even the most knowledgeable experts to define exactly how old the average child must be before he can be taught a given safety-related concept. Estimates about the earliest age at which a child can be taught a given concept may vary over a range of several years. Concepts that fall in this range of uncertainty may be included or excluded from an educational program designed for a specific age group, depending on the opinion and biases of the program developer. Controversies of this type can be resolved only through research.

ASSUMPTIONS ABOUT RESOURCES AND CONSTRAINTS

Perhaps the most important source of controversy about educational objectives is the assumptions that are made about the resources that can be devoted to bicycle-safety education or, conversely, the constraints within which a program must operate. Examples of assumptions that may have an important impact on educational objectives include:

• The agency who is responsible for developing and implementing the program (schools, police departments, bicycling organizations, civic groups, etc.).

• The amount of time students will be available for education and training.

• The funds and other resources available for developing educational materials and activities.

• The educational media and method (the distribution of reading materials, public-service radio or television spot announcements, one-shot rodeos, classroom training, on-the-road training, training in driving simulators, or some combination of these).

• The knowledge and sophistication of the instructional personnel.

Until now, there has been a strong tendency to tailor educational programs to the capabilities and limitations of the organization that will be responsible for administering the program. Since organizations differ greatly in their capabilities and limitations, it is not surprising to find important differences in the specific educational objectives they have adopted for their programs. Tailoring educational objectives to the capabilities and limitations of an organization is a dangerous practice that can lead to programs which are so incomplete and superficial that they have little or no impact on accidents. A better approach is to define the objectives of an ideal program and then identify the organizations who are most capable of administering that program. The same is true for educational methods and media. That is, the educational methods and media to be employed should be dictated by the educational objectives rather than attempting to tailor educational objectives to a specific method or media.

MULTIPLE EDUCATIONAL STRATEGIES

For most types of accidents, there are several educational strategies that may prove effective in reducing accident likelihood. One strategy might focus on education that would induce bicyclists to avoid riding in high-hazard areas; another strategy might focus on education that would induce bicyclists to modify their speed and path at high-hazard locations; a third strategy might focus on training that would increase the bicyclists' ability to make emergency stops and turns. Differences in educational strategy account for some of the differences in the objectives of contemporary bicycle-safety education programs. The controversy about such differences would be greatly reduced if (a) the various education strategies were defined more explicitly, and (b) the rationale for selecting one strategy over another was explained by educational program developers.

OBJECTIVES OF BICYCLIST EDUCATION

Because an accident is the end product of a sequential chain of events, it is possible that the causal chain may be broken and the accident avoided by modifying any one of the events in the causal chain. The implication of this assertion is that there may be several different educational solutions for the same type of accident. When discussing educational objectives, it is important that the full range of potential educational solutions be considered, and it is important that they be considered in an organized fashion. To provide an organizational framework for discussing the full range of educational objectives, the bicycle-riding task has been divided into three sets of functions. The first set of functions -- the Preparatory-Phase functions -- are those ordinarily accomplished before the bicyclist departs on a trip. The second set of functions -- the Anticipatory-Phase functions -- are those required to select a safe course (both path and speed) through an area. The third set of functions -- the Reactive-Phase functions -- are those required to respond to a specific threat in the environment.

For each of the three sets of functions, educational objectives are discussed at two levels of specificity. At the most general level (Level I), the objectives are defined in broad behavioral terms. Level II objectives are defined in terms of the knowledge and skills that must be enhanced and the values that must be modified in order to achieve the behavioral changes specified by the Level I objectives. If the educational objectives defined here prove valid and meaningful, it will be necessary to define objectives at a third level of specificity. Objectives at Level III would define the specific methods and techniques required to accomplish the Level II objectives. An effort is presently underway to conduct the research, development, and evaluation needed to define objectives at the third level of specificity. The effort is being funded by the National Highway Traffic Safety Administration (NHTSA) and should be completed by late 1979 or early 1980. Readers who have ideas about specific methods and techniques for accomplishing the Level II objectives defined below are encouraged to convey their ideas to NHTSA directly or to convey them indirectly through the AAA Foundation for Traffic Safety.

The discussion of educational objectives follows a brief discussion of the target groups for bicyclist education.

COMMENTS ABOUT THE TARGET GROUPS FOR BICYCLIST EDUCATION

Decisions about the target groups for bicyclist education must be based on a joint consideration of two factors: (a) the age distribution of the accident population of bicyclists, and (b) the relationship between educational costs and the age of the bicyclists being educated. On one hand, there is a need to introduce education at an early enough age so that substantial numbers of accidents will not have occurred before bicyclists receive the education. On the other hand, there is a desire to delay education for as long as possible because the ease and efficiency of accomplishing the educational objectives tends to increase as a function of the age of the student. That is, within limits, it is easier to teach complex concepts and skills to older children than to younger children. The following paragraphs address this dilemma.

Consider first the age distribution of the bicycling population. The curves in Figure 33 show the proportion of fatal and non-fatal bicycle/motor-vehicle accidents that occur after each age from four to 18 years. The proportion of accidents occurring after a specified age provides an indication of the maximum payoff that could be realized from education introduced at that age. For instance, Figure 33 shows that only 36% of the fatal accidents and 21% of the non-fatal accidents involve a bicyclist who is older than 18 years of age. Thus, the maximum benefit that could be realized from education introduced at this age level is a 36% reduction in fatal and a 21% reduction in non-fatal accidents.

The finding that accident rate is highest for bicyclists between 11 and 13 years of age has led some persons to conclude that education should be focused on bicyclists within this age group. Examination of Figure 33 shows that the potential payoff would be considerably reduced if education was delayed until bicyclists had reached the age of 11, 12, or 13 years. For instance, if education was delayed until the age of 13, the potential payoff would be reduced to 64% for fatal and 52% for non-fatal accidents. The potential payoff would be higher if the education was introduced at age 11; but, even so, it would be only 75% and 70% for fatal and non-fatal accidents, respectively.

The data in Figure 33 make it clear that a major educational effort must be introduced at the elementary school level if bicycle-safety education is to have a significant impact on accidents. However, this conclusion gives rise to several important questions that are difficult to answer with the data presently available.

• What is the earliest age at which children can be taught the necessary concepts, principles, and skills?

• How long can children of different ages be expected to retain the concepts, principles, and skills they are taught?

• If retention is a problem, what are the requirements for education to refresh students' recollection of key concepts and principles in the years after they have completed a comprehensive education program?

• What is to be done to educate bicyclists who are older than the educational target group at the time a bicycle-safety education program is first introduced?

*PS = Preschool

** K = Kindergarten

The ease and efficiency of almost any type of education is importantly determined by the language skills of the student, particularly the student's ability to read and write. According to experts in elementary education with whom the author has discussed this problem, the language skills of a typical student represent a serious barrier to efficient education until the student has completed the third grade (about nine years of age). For this reason, most elementary education experts identified the fourth grade as the earliest age at which bicycle-safety education could be accomplished with reasonable efficiency. Some experts believed that education should not be introduced until the fifth grade, but few expressed the view that it would be effective to introduce a truly comprehensive education program earlier than the fourth grade.

The data show that the potential payoff of a comprehensive educational program introduced at the fourth-grade level would be about 87%. But, what about the 13% of the accidents that will have already occurred before the education is introduced? When attempting to formulate an answer to this question, the data were examined to determine the type of accidents that involve bicyclists who have not yet reached the fourth grade (bicyclists younger than about nine years of age). It was found that bicyclists younger than nine years of age are involved in a relatively small number of different types of accidents. About 60% of the accidents are "bicyclist-rideout" accidents that occur at the junction of a street and driveway/alley, at an intersection controlled by a stop sign, or at an uncontrolled intersection. Another 15% of the accidents occur when a bicyclist makes an unexpected left-hand turn into the path of an overtaking motor vehicle. This finding suggests the possibility of a limited educational program that would focus solely on the behavioral errors that contribute to "bicycle-rideout" and "unexpected-left-turn" accidents. Such a program would have maximum potential payoff if it was introduced at the kindergarten level; but because of the difficulty of educating kindergartners, it is believed that the program would be most effective if introduced at the first-grade level.

A limited educational program introduced at the first grade and a comprehensive educational program introduced at the fourth grade would have a combined potential payoff in excess of 90%. The actual payoff would be a function of the program's effectiveness in achieving the desired behavioral changes and the extent to which the educational material is retained by the students. Even if a program proved highly effective in achieving the desired behavioral changes, it is unlikely that the effects would be long lasting without subsequent education to refresh bicyclists' recollection of key concepts and principles. if a comprehensive program is introduced at the fourth-grade level, it is believed that "refresher" education should be introduced at least every other year thereafter, through the tenth grade. It must be emphasized, however, that the recommendation about the grade levels for administering refresher education is based upon precious little empirical information. Once a comprehensive program has been developed, this issue can be resolved empirically by retesting students each year for several years after they have received the education. With such test data in hand, it would be a relatively simple matter to judge when retention has degraded enough to warrant refresher education.

In summary, the author's "best guesses" about the target groups for bicyclist education are as follows:

• First graders (six-year olds) -- limited program aimed at "bicyclist-rideout" and "unexpected-left-turn" accidents.

• Fourth graders (nine-year olds) -- comprehensive education program aimed at all types of accidents, including NMV accidents.

• Sixth graders (11-year olds) -- reinforcement education aimed at all types of accidents. Eighth graders (13-year olds) -- reinforcement education aimed at all types of accidents.

• Tenth graders (15-year olds) -- reinforcement education with emphasis on Problem Types 6, 8, 9, 10, 13, 16, 17, 19, 22, 23, and 24.

Assuming that it is possible to introduce a program to educate the target groups defined above, it is then appropriate to ask, What is to be done to educate bicyclists who are older than the primary target group (fourth graders) at the time the program is first introduced? The answer to this question depends almost entirely on the funding available. Ideally, it would be possible to commence a long-term educational program by educating the entire population of bicyclists during the first year and educating only the primary target group in each subsequent year. However, a one-shot program to educate the entire population of bicyclists in a single year would involve monumental costs and countless logistics problems. In all likelihood, it would be impossible to obtain the funds needed to accomplish such an ambitious task. This means that educating the population must be accomplished over a number of years.

In the author's view, the education of each year's crop of first and fourth graders should be considered first priority. If additional funds can be obtained, they should be spent providing comprehensive education to as large a group of older bicyclists as is possible with the funds available, rather than providing limited education to every bicyclist older than the primary target group. A decision to exclude some bicyclists from a safety-education program may seem callous, but it would be far worse to decide upon expending the limited educational resources on a program that would provide only superficial education to large numbers of bicyclists.

EDUCATION TO ENHANCE PREPARATORY-PHASE FUNCTIONS

By definition, the Preparatory Phase of a bicycling trip commences when the operator makes a decision to execute a trip and terminates at the point at which the operator begins the task of selecting a course through a particular area. During the Preparatory Phase, a bicyclist must evaluate his own capability and that of his vehicle to complete the anticipated trip under the environmental conditions that will be encountered during the trip. In addition, the bicyclist must evaluate alternate routes to his destination and select the one that best suits his momentary needs. Education aimed at the Preparatory-Phase functions is based upon the assumption that bicycle accidents can be prevented by education that would increase bicyclists' ability and inclination to perform the Preparatory-Phase functions. The Level I and Level II educational objectives for enhancing the performance of Preparatory-Phase functions are listed in Table 22 and discussed below. The discussion includes objectives that are clearly important as well as some whose importance is marginal or has yet to be determined.

Increase knowledge of procedures and criteria for evaluating the bicycle's state of repair.

Increase knowledge of procedures and criteria for evaluating the adequacy of safety equipment for the contemplated trip.

Increase knowledge of procedures and criteria for determining whether the bicycle fits the rider.

Increase knowledge of risk associated with riding a bicycle that has mechanical defects, is ill-fitting, and/or is not equipped with needed accessories.

Increase knowledge of maintenance, adjustments, and repairs that should be accomplished by a parent or professional bicycle mechanic.

Increase knowledge of the effect of darkness on accident likelihood.

Increase knowledge of the necessity for lighting equipment when night riding cannot be avoided.

Increase awareness of the effect of specific knowledge and skill deficiencies on accident likelihood.

Increase awareness of the knowledge and skills required to complete various types of trips with reasonable safety.

Perform Safety Check of Bicycle

Mechanical condition. Nearly every bicycle-safety program in existence stresses the importance of performing a safety check of the bicycle before departing on a trip. Some programs merely attempt to induce the bicyclist to perform the safety check, whereas others are designed to teach bicyclists how the safety check is to be accomplished. If education is to be introduced at the fourth-grade level, it seems certain that many bicyclists will not possess the knowledge required to perform a safety check. Thus, if young bicyclists are to be taught to perform an effective safety check, it will be necessary to increase their knowledge of:

• The parts of a bicycle and their functions.

• The procedures and criteria for evaluating the state of repair of each bicycle part.

• The procedures and criteria for evaluating the adequacy of safety equipment for the contemplated trip.

• The procedures and criteria for determining whether the bicycle fits the rider.

Assuming a bicyclist can be taught how to perform the safety check, education also will be required to induce him to do so on a regular basis. This will require education to increase bicyclists' knowledge of the risk associated with riding a bicycle that has mechanical defects, is ill fitting, is not equipped with the necessary safety equipment, or a combination of these. In short, the bicyclist must be educated about the extent to which accident likelihood is increased when they fail to perform a safety check before departing on their trip.

Although there is evidence that a substantial proportion of bicycles have defects which are potentially accident producing (see Figure 6 in Section IV), it was found that a bicycle defect contributed to less than three percent of all bicycle/motor-vehicle accidents. The main contributor was defective brakes, which was a factor in about one percent of the bicycle/motor-vehicle accidents. The role of bicycle defects in NMV accidents is not known for certain, but many experts feel strongly that mechanical defects contribute to a substantial portion of NMV accidents. In summary, teaching bicyclists to check the mechanical condition of their bicycle could have a small, but significant, impact on bicycle/motor-vehicle accidents and possibly a far greater impact on NMV accidents.

Bicycle size/fit. The literature contains a great deal of instructional material that has been designed to teach bicyclists how to select the proper size bicycle and how to adjust the bicycle handlebars and seat to fit the rider. This fact reflects the common belief that many bicyclists ride ill-fitting bicycles, that a bicyclist's ability to maintain proper control is seriously degraded when riding an ill-fitting bicycle, and that accidents frequently result from bicyclists riding ill-fitting bicycles. These views are so logically appealing that it is difficult to argue with them; yet, there is little recent empirical data to support them. The author knows of no systematic research that has been conducted to determine the number of bicyclists who ride ill-fitting bicycles or to assess the extent to which control is degraded when bicycle fit is non-optimal. In Section V, it was mentioned that riding an oversized bicycle was a contributing factor in about one percent of all bicycle/motor-vehicle accidents. Although it seems probable that non-optimal fit would contribute to an even larger proportion of NMV accidents, no data have been located to support or refute this assumption. In addition to fit, as measured by the operator's ability to reach the pedals and handlebars, a small number of accidents were noted in which the bicyclist's hands were too small to grasp caliper-brake handles. Although only one or two bicycle/motor-vehicle accidents of this kind were noted, it is possible that this aspect of bicycle fit accounts for an important number of NMV accidents.

After reading the above paragraph, the reader may be surprised that teaching bicyclists to check the size and fit of their bicycle has been included as an educational objective. Although such education appears to have the potential for eliminating only a small number of bicycle/motor-vehicle accidents, it is the type of education that can be accomplished quickly and in a straightforward manner. The low cost of teaching bicyclists to check bicycle size and fit, combined with the possibility that such education would serve to reduce NMV accidents, seems sufficient justification for the inclusion of this educational objective.

Safety equipment. Education to induce bicyclists to check the type and condition of safety equipment on their bicycle appears to have considerable potential for reducing accidents. The main emphasis should be placed on lighting equipment. It will be recalled from Table 13 (Section ill) that a large proportion of bicyclists involved in night accidents were riding a bicycle with inadequate lighting equipment. For instance, only 13% of the bicycles were equipped with an operational headlight, about 40% were equipped with side reflectors, about 40% were equipped with a front reflector, about 63% were equipped with reflectorized pedals, and 67% were equipped with a rear reflector. Although there is a clear need for the development of more effective bicycle-lighting equipment, it seems reasonable to assume that a substantial proportion of night accidents would be avoided if bicycles were equipped with the best lighting equipment that is presently on the market.

There is virtually no question that a safety-education program should place heavy emphasis on checking the adequacy of a bicycle's lighting equipment before departing on a trip that will involve night riding. Indeed, it is appalling to find that only 13% of the bicycles involved in night accidents were equipped with an operational headlight. Riding at night without a headlight reduces the bicycle's visibility to motorists and also may increase the likelihood of NMV accidents. Standard headlights provide sufficient illumination for the bicyclists to observe major street-surface defects and other large hazards. Headlights with greater than standard power will be required to provide the illumination needed to observe less visually prominent hazards, such as speed bumps, small debris in the roadway, cables across driveway entrances, and so on.

Other types of safety equipment that may decrease accident likelihood include safety flags, baskets and racks, chain guards, handlebar grips or tape, rear-vision mirrors, and auditory warning devices (horn or bell). Unfortunately, there are no data to use in estimating the number of accidents that would be eliminated if bicycles were equipped with such devices. It is known that lack of daytime conspicuity is a contributing factor in many bicycle/motor-vehicle accidents, and it is highly probable that daytime conspicuity would be increased by safety flags. However, the potential benefit of educating bicyclists to equip their bicycles with safety flags cannot be estimated until research is conducted to evaluate their effectiveness. Additional research is also needed to evaluate the benefit of education to equip bicycles with the other safety-equipment items listed above.

In summary, it is highly probable that considerable benefit would derive from educating bicyclists about the need for effective lighting equipment. Educating them about the benefit of other safety-equipment items remains uncertain at this time, but providing such education would certainly not be harmful if it could be accomplished without reducing the time available to educate bicyclists about more important matters.

Bicycle maintenance, adjustments, and repairs. As was stated above, about three percent of all bicycle/motor-vehicle accidents are the direct or indirect result of bicycle defects. Some experts believe that a substantial number of NMV accidents result from bicycle malfunctions, but no data are available to estimate this number accurately. Unless the number of NMV accidents resulting from bicycle defects proves to be large, the accident reduction potential of education about bicycle repair and adjustment must be considered marginal at best. In light of these facts, education about bicycle repair and adjustment would not be cost-effective unless methods can be developed to accomplish this educational objective with only a small expenditure of time and resources.

It is unlikely that cost-effective methods could be developed to teach young bicyclists the full range of skills required to fully maintain their bicycles and repair any malfunction that could arise. However, it is possible that it would be cost-effective to (a) teach bicyclists to perform the most simple maintenance, adjustment, and repair tasks, and (b) teach bicyclists the specific tasks that should be performed by a parent or professional bicycle mechanic. For instance, it should be a relatively simple matter to teach young bicyclists to adjust their seat and handlebars, tighten loose nuts and bolts, replace dead batteries in headlights, clean reflectors, adjust the chain, and perhaps other simple maintenance and repair tasks as well.

As of now, the advisability of educating bicyclists to maintain, adjust, and repair their bicycles remains uncertain. Whether or not teaching such skills should be established as an educational objective will depend on the level of skill that must be acquired, the efficiency of the methods that are developed to teach the skill, and the time and resources needed to accomplish more important educational objectives. In any event, teaching such skills cannot be considered a primary educational objective.

Evaluation of Weather and Lighting Conditions

Decisions about whether or not to make a trip and decisions about an optimal route to a destination should be based upon a careful consideration of the weather and lighting conditions that will be encountered during the trip. Judicious decisions can be made only if bicyclists have a clear knowledge of the effect of inclement weather on accident likelihood, the effect of darkness on accident likelihood, and the necessity for effective lighting equipment when riding at night. According to the evidence presently available, the incidence of bicycle riding is reduced drastically at night and during periods of inclement weather. Even so, night accidents account for about one-third of all fatalities and ten percent of all injuries resulting from bicycle/motor-vehicle accidents. About three percent of the bicycle/motor-vehicle accidents in the Cross and Fisher study (1977) sample were found to occur during inclement weather, but the number of inclement-weather accidents may be considerably higher in some geographical areas. Moreover, it is probable that inclement weather contributes to a substantial number of NMV accidents.

For the above reasons, educating bicyclists to carefully evaluate weather and lighting conditions is considered a worthwhile objective for a bicycle-safety education program. The education should induce bicyclists to either refrain from riding at night and during periods of inclement weather or, at least, select routes that are safest when lighting and/ or weather conditions are suboptimal. Education about route selection is discussed in detail below.

Route Selection

Teaching bicyclists to select the "safest" route is among the most common objectives of existing educational programs. However, programs differ greatly in what bicyclists are taught about route selection. Some programs accomplish little more than making an emotional appeal to always select the safest route. These programs make no reference whatsoever to the criteria to be used in evaluating the relative safety of alternate routes. Some safety-education programs developed for school-age children provide explicit instruction on safe routes to school. The safe routes to school are usually defined by an employee of the local traffic engineering department. Although the criteria used to select safe routes to school are seldom defined explicitly, it appears that heavy weighting is placed on: roadway width, traffic volume, traffic speed, parking density, number of intersecting roadways (including driveways and alleys), number of traffic signs and signals, complexity of intersection configuration, roadway-surface condition, and number of left-hand turns required to travel the route. No program has been found that identifies both the types and relative importance of route selection criteria.

There are three problems that must be addressed before it will be possible to increase bicyclists' ability and inclination to select the safest route to their destination. First, it will be necessary to conduct research to clearly establish the relationship between accident likelihood and various route characteristics. A careful search of the literature has revealed no such data, so it must be concluded that existing materials on route selection criteria are based upon logical considerations rather than on empirical data. Although it is probable that accident likelihood is far greater on some types of roadways than on others, it is also probable that the most dangerous routes are not the ones that would be judged most dangerous by a panel of experts. As a whole, bicyclists and motorists are capable of recognizing dangerous situations and countering the hazard by exercising more caution than normal. Because of this fact, accident likelihood may very well be inversely related to the apparent hazardousness of the situation.

If route selection criteria can be established empirically, the next problem that must be dealt with is that of developing techniques that will enable young bicyclists to evaluate alternate routes in terms of these criteria. Unless all criteria are of equal importance, a comparison of alternate routes will require a bicyclist to formulate a composite safety index for each route based upon different combinations of differently weighted variables. Such a task would be difficult for adults and quite impossible for children if the number of criteria is large and their weights highly variable.

A third problem is that of inducing bicyclists to select the safest route when an alternate route is-faster, shorter, flatter, or otherwise more desirable to the bicyclist. Bicyclists are notorious for their reluctance to: deviate significantly from the most direct route, climb long or steep hills when they can be avoided, ride on rough roads when a smoother route is available, and ride on a roadway with many stop signs/signals when a more continuous route is available. Therefore, it will be difficult to develop educational methods that will modify bicyclists' values to such an extent that safety considerations .will always have priority over riding ease, efficiency, and enjoyment.

As is shown in Table 22, the uncertainties about educating bicyclists to select safe routes are so great that it is not presently possible to define Level II objectives; the above discussion shows that the possibility and desirability of accomplishing this educational objective remains in serious doubt. However, one fact seems certain: it would be extremely difficult to teach young children the complex computational skills needed to make an objective assessment of the safety of alternate routes. This suggests that it may be necessary for an expert to make an objective assessment of all or most roadways in a community and to use the results to develop a special map that classifies each roadway in terms of its safety for bicycle travel. Such a map would be a valuable training aid for young bicyclists and a valuable route-selection aid for bicyclists of all ages.

Bicyclists' Evaluation of Their Own Physical/Mental Capabilities

It would be highly desirable if bicyclists could be taught to make a rational assessment of their own capability to complete a trip safely -- given the bicycle they intend to ride, the conditions under which they intend to ride, and the characteristics of the route they plan to take to the destination. In order to provide such education, it is first necessary to define, in some reasonably exacting terms, the types of bicyclists that are incapable of safely completing certain types of trips. In principle, a bicyclist may be incapable of completing a certain type of trip safely because of specific knowledge deficiencies, physical impairments, sensory impairments, or mental impairments. The impairments may be temporary or permanent. In practice, it is very difficult to establish a firm relationship between accident likelihood and any of the types of deficiencies or impairments that may contribute to accidents. Furthermore, it would be necessary to measure the capabilities of individual riders before it would be possible to instruct them about their ability to complete a trip with reasonable safety. For these reasons, it is necessary to identify specific groups with easily measurable characteristics that are known to be related to accidents.

Since there is a high correlation between age on the one hand arid knowledge and skill level on the other, it seems safe to assume that age can be used to identify bicyclists who are incapable of making certain types of trips with reasonable safety. A second group of bicyclists who are incapable of bicycling with reasonable safety are those who are temporarily impaired by alcohol or drugs. The study of bicycle/motor-vehicle accidents showed that about two percent of the accidents involved a bicyclist who was under the influence of alcohol or drugs. A third high-risk group is bicyclists who are retarded. Surprisingly, it was found that about one percent of the bicycle/motor-vehicle accidents involved a retarded bicyclist, and it is altogether possible that the actual proportion is somewhat greater. Clearly, even moderately retarded bicyclists must be considered incapable of bicycling safely. A fourth high-risk group includes bicyclists with sensory or motor impairments. Impaired vision and impaired limbs, together, were found to contribute to about one percent of all bicycle/motor-vehicle accidents.

Is it reasonable to believe that education would be effective in inducing young bicyclists, permanently impaired bicyclists, or temporarily impaired bicyclists to refrain from riding? With one exception, it is believed that this question must be answered negatively. It is believed that education has some potential for inducing young bicyclists to refrain from taking trips [7] that are clearly beyond their capability to complete with a reasonable degree of safety. It is believed that this education would be particularly effective if combined with a program to educate parents about the types of trips that young children should not be permitted to take unless accompanied by an adult. Some additional research will be required to identify the types of trips that are excessively hazardous for bicyclists in various age groups.

A program to induce retarded bicyclists to refrain from riding would be difficult to develop and costly to administer. It would be equally difficult to develop education that would induce bicyclists to avoid riding a bicycle when they are under the influence of alcohol or drugs. Historically, education has not proved highly effective in reducing the number of motor-vehicle operators who drive while under the influence of alcohol or drugs, and there is no reason to expect that such education would be more effective for bicyclists than for motorists. In short, education to increase retarded and impaired bicyclists' ability and inclination to consider their own capabilities for completing the contemplated trip safely must be considered of secondary importance.

EDUCATION TO ENHANCE ANTICIPATORY-PHASE FUNCTIONS

Once a bicyclist has decided to travel a given segment of roadway, he must decide upon the specific path he will follow and the speed he will travel as he traverses that segment of roadway. The term course selection refers to the selection of a path and a speed to be traveled along a roadway segment and should not be confused with the term route selection. By definition, the Anticipatory-Phase functions are those that a bicyclist must perform to select the safest course through an area. The safest course through an area is the lawful and reasonable course for which accident likelihood is smallest. A course is not considered reasonable if it is so inconvenient or uncomfortable that even a skilled, safety-conscious bicyclist would never select that course. All courses other than the safest one are referred to here as "suboptimal" courses; only the safest course is referred to as "optimal." As is discussed later, it is often difficult to define the single course that is optimal in some traffic contexts.

It was found that about 75% of all bicycle/motor-vehicle accidents were either the direct or indirect result of the bicyclist's selection of ,a suboptimal course. In about 15% of the cases, the bicyclist's suboptimal course led directly and immediately to the crash. That is, because of the bicyclist's suboptimal course, neither operator had sufficient time to initiate successful evasive action once the other vehicle could first be seen. These accidents became inevitable at the moment the bicyclist decided on the suboptimal course. In another 60% of the cases, it was judged that the bicyclist's suboptimal course was not the most immediate cause of the accident, but contributed to the accident by (a) decreasing the time and space available for evasive action and/or (b) increasing the level of skill required for successful evasive action. That is, while there was sufficient time and space for successful evasive action once the other vehicle first became observable, successful evasive action required far more skill or a higher level of alertness than would have been required If the bicyclist had not selected a suboptimal course.

The contribution of bicyclists' suboptimal pre-crash courses to bicycle/motor-vehicle accidents has important implications for bicycle-safety education. These findings emphasize the. fact that there are some accidents that simply cannot be avoided by educating bicyclists to respond to potentially threatening-motor vehicles in the environment. Rather, the education must concentrate on decisions that are made before a potentially threatening motor vehicle can be observed. This is not to suggest that educating vehicle operators about course selection is new. Indeed, many of the laws, ordinances, and safety rules that have been developed for both motor-vehicle and bicycle operators have been designed to induce vehicle operators to select the safest possible path and speed through an area.

The educational objectives for enhancing the performance of the Anticipatory-Phase functions are listed in Table 23 and are discussed below. Level 1 and Level II objectives are discussed in turn.

Education on Course Selection -- Level I Objective

It is unlikely that there would be sufficient time in even the most extensive education program to teach bicyclists the optimal course to follow in every traffic context they might encounter. And yet, with only a few exceptions, it is impossible to formulate general rules about course selection that are valid for all traffic contexts and effective in prescribing the exact course that minimizes accident likelihood. It is therefore necessary to focus on the course-selection behavior that has the greatest accident reduction potential. The course-selection behavior that is most critical for bicycle/motor-vehicle accidents can be defined from a study of the accident illustrations presented in Section V. Other important course-selection behavior undoubtedly would be revealed by the study of NMV accidents.

The general objectives of education to enhance the performance of Anticipatory-Phase functions is to increase the bicyclists' ability and inclination to select the optimal course through an area. The specific behavior the education must induce is outlined in the left-hand column of Table 23 and is discussed in more detail below. it is important that the reader keep in mind that the objectives included here do not include teaching bicyclists the evasive actions that are required when a potentially threatening motor vehicle is observed. Evasive action is one of the Reactive-Phase functions discussed later.

• Always ride with traffic.

• Select optimal course when entering the roadway.

• Stop at signed intersections.

• Avoid entering signalized intersections after the onset of the amber phase.

• Select optimal course at uncontrolled intersections.

• Select optimal course when making left-hand turns.

• Select optimal course when visual obstructions are encountered.

• Ride an optimal distance from right-hand edge of roadway.

• Select optimal speed when riding downhill, when riding during darkness, and when riding on wet or debris-covered roadway.

Increase the validity of bicyclists' assessment of the relative degree of risk associated with optimal and suboptimal courses.

Increase knowledge of needs that are in competition with the need for safety, and decrease the perceived

need satisfaction associated with suboptimal courses.

Increase ability and inclination to search for, recognize, and cope with visual obstructions.

Increase validity of expectations that may influence course selection.

Increase knowledge of the time and space required to respond to a threat (as a function of bicycle handling skill and bicycle speed).

Increase knowledge of amber-signal phase.

Entering roadway. About 15% of all fatal and 14% of all non-fatal bicycle/motor-vehicle accidents occurred as a bicyclist was entering the roadway at a mid-block location; the bicyclist's course was judged suboptimal in nearly every case. It is tempting to propose that bicyclists be educated to always stop and walk their bicycle into the roadway. However, it is unlikely that education could ever induce bicyclists to adopt such an inconvenient behavior pattern, especially when there are no laws and ordinances to motivate them to do so. It seems more reasonable to develop educational methods to accomplish the following objectives:

• Teach bicyclists to never enter a roadway by riding over a curb or any other discontinuity at the roadway edge that is so large/rough that they must scan downward at the curb/discontinuity rather than searching for approaching traffic.

• Teach bicyclists to slow their roadway-entry speed to the extent needed to provide sufficient time to search for approaching traffic and to initiate successful evasive action.

• Teach bicyclists to select an entry path that minimizes the time they are exposed to traffic.

These educational objectives cannot be accomplished by requiring bicyclists to learn a few general rules, because the optimal course for entering the roadway varies greatly with the physical characteristics of the traffic context and the intended direction of travel by the bicyclist. Rather, bicyclists must be taught the exact course to follow in a wide range of specific traffic contexts, including those in which visual obstructions are present. For each traffic context, the bicyclist must be taught the best course to follow when turning right, turning left, and proceeding straight across the roadway.

Signed intersections. The importance of education that will induce bicyclists to select an optimal course at signed intersections cannot be emphasized enough; a suboptimal course at signed intersections was a prime contributor to about eight percent of all fatal and 10% of all non-fatal bicycle/motor-vehicle accidents. There is no question that young bicyclists should be taught to come to a complete stop at all signed intersections. Whether older bicyclists should be taught to come to a complete stop or to merely slow to a very low speed remains open to question.

One fact is certain: this educational objective will not be accomplished by teaching bicyclists the law. Bicyclists know full well that the law requires bicycles to stop at signed intersections, even very young bicyclists. To be effective, education must convince bicyclists of the necessity for stopping (or at least slowing significantly) at all signed intersections, including those that carry light traffic and are not perceived as hazardous by the bicyclist.

Signalized intersections. Inducing bicyclists to avoid running red lights probably should be included among the objectives of an education program, but few accidents occur because bicyclists blatantly ride through a red light. Instead, the accidents usually occur because a bicyclist enters the intersection after or shortly before the onset of the amber phase. The problem is sometimes compounded by a multiple-threat situation in which the bicyclist is struck after passing in front of one or more lanes of standing motor vehicles whose operators have observed the bicyclist and are waiting for him to pass.

Accordingly, two important objectives are:

Teach bicyclists to avoid entering a signalized intersection after the onset of the amber phase.

When bicyclists see they cannot clear the intersection before the onset of the red light, teach them to stop at a central island or, if none is available, continue at a slow speed and search the traffic lanes beyond any motor vehicles that are stopped -- apparently waiting for the bicyclist to clear the intersection -- before proceeding.

Uncontrolled intersections. Less than three percent of all bicycle/motor-vehicle accidents occur at uncontrolled intersections. Even so, education on course selection must be considered an important objective. Except for the different traffic context, these accidents occur in much the same way as those occurring when bicyclists enter the roadway from a mid-block location. Speed control is of primary importance but path selection may be important when visual obstructions are present.

Left-hand turns. Most accidents that occur when a bicyclist turns left into the path of a motor vehicle are the direct result of the bicyclist's failure to search.

However, it is possible that the bicyclist's inclination to search may be influenced by the specific course he adopts for making the left-hand turn. Analytical considerations and casual observations have led the author to conclude that the course many bicyclists select for left turns imposes excessive demands on their information-processing system. For instance, a sharp left-hand turn from the right-hand curb requires the bicyclist to search simultaneously both the overtaking and the opposing lanes of traffic. The difficulty of this task is a direct function of the number of traffic lanes in each direction and the volume of motor-vehicle traffic at the time. The information-processing load on the bicyclist would be less if he executed a two-phase turn. He would first scan behind for overtaking traffic and proceed to the center of the roadway when it was safe to do so. He would then ride along the center of the roadway until he had scanned ahead and determined that it was safe to turn left across the opposing traffic lane(s).

Education on left turns requires that bicyclists be taught to evaluate the traffic context in terms of its general complexity and select a course that does not place excessive demands on the bicyclist's information-processing system. The greatest benefit would result from explicit demonstrations of the optimal course for a left-hand turn in a wide variety of traffic contexts, including those that clearly overload the bicyclist's information-processing system.

Visual obstructions. The importance of visual obstructions for course selection has been mentioned in the above discussion of course selection when entering the roadway at a mid-block location and when entering signed, signalized, and uncontrolled intersections. Visual obstructions are also important when riding on sidewalks that intersect alleys and driveways and when riding on uncontrolled roadways that intersect controlled streets and controlled or uncontrolled driveways/alleys. It is in these situations that motorists inadvertently drive into the path of the bicyclist because the bicyclist is obscured from view. Since visual obstructions are a contributor to such a large number of different types of accidents, it seems worthwhile to establish as a separate objective the education of bicyclists to recognize and cope with visual obstructions. Obviously, there is a great deal of overlap between this objective and those discussed earlier in this section.

Proximity to right-hand edge of roadway. Most communities have a law or ordinance stating that "bicyclists must ride as close to the right-hand edge of the roadway as is practicable." Such a law is difficult to enforce because what is "practicable" depends on such a wide variety of factors. Simple rules about how close to the edge of the roadway bicyclists should ride are more likely to be counterproductive than productive. What is needed is highly specific instruction on the best path to follow (relative to the edge of the roadway) on each of a wide variety of traffic contexts and for bicyclists with various skill levels.

Unfortunately, there is considerable disagreement, even among bicycling experts, about how close to the edge of the roadway bicyclists should travel in order to minimize accident likelihood. The problem stems from the fact that riding too far to the right increases the likelihood of some types of accidents and riding too far to the left increases the likelihood of other types of accidents. For instance, when riding along a row of parallel-parked motor vehicles, riding too far to the right increases the chances of colliding with an opening car door, and riding too far to the left increases the chances of being struck by an overtaking motor vehicle. The path that is optimal in this situation depends on such factors as: the width of the roadway, the volume and speed of overtaking motor vehicles, the bicyclist's ability to see whether the parked vehicles are occupied, the bicyclist's ability to maintain accurate lateral control, and perhaps others as well. The bicyclist is faced with a similar dilemma when there are other obstacles or roadway-surface debris in the area where he would ordinarily choose to ride.

The author is not yet prepared to make specific recommendations about how close to the right bicyclists should be taught to ride in various traffic contexts. It is believed that analytical study by a group of experts and perhaps additional field research will be required to formulate specific recommendations about where bicyclists should be taught to ride in various traffic contexts.

Other course control. In the above paragraphs, the educational requirements for

course selection were defined in terms of hazardous traffic contexts, hazardous maneuvering, or both. There are additional educational requirements for course selection that must be

defined in terms of general conditions or situations rather than specific traffic contexts.

All of the important requirements of this type are for speed control, including:

• Speed control when riding downhill.
• Speed control when riding on a wet roadway.
• Speed control when riding with wet caliper brakes.
• Speed control when riding on a roadway covered by sand, gravel, or other debris.
• Speed control during darkness.

There was no one of the above conditions in which suboptimal speed control contributed to large numbers of accidents; but together they easily constitute a significant enough problem to warrant attention in a safety-education program. What must be accomplished is to teach bicyclists the fastest speed that is safe under each of these conditions.

Education on Course Selection -- Level II Objectives

At Level II, educational objectives are defined in terms of the knowledge that must be imparted, the skills that must be developed, and the values that must be modified in order to achieve the behavioral changes specified by the Level I objectives. Accordingly, the Level II objectives described below were formulated through a study of the reasons why bicyclists in the accident sample selected a suboptimal course. The Level II objectives for enhancing the performance of the Anticipatory-Phase functions are listed in the right-hand column of Table 23 and are discussed below.

Increase ability to identify optimal course. Although bicyclists often select a course they know is less safe than another, there are many cases in which bicyclists lack the knowledge and skill needed to differentiate the optimal course from the many suboptimal courses that are available. Thus, a primary educational objective is to teach bicyclists to identify the optimal course for a wide variety of traffic contexts, maneuvers, and conditions. First priority should be given to the high-hazard traffic contexts, maneuvers, and conditions that were identified in the discussion of level I objectives.

Education to increase bicyclists' ability to identify optimal courses must commence with instruction that will (a) increase bicyclists' inclination to search their immediate surroundings and (b) increase their ability to recognize the physical and operational attributes of the traffic context that influences the relative safety of alternate courses through that area. The acquisition of this skill involves discrimination learning. The bicyclist must learn to scan a highly complex visual field and discriminate the relatively small number of stimuli that are relevant for course selection. Although the acquisition of this skill sounds difficult, humans can often acquire such a skill in less time than it takes to describe it.

The second task is to teach bicyclists to recognize the high-hazard locations, maneuvers, and conditions. This education must establish a powerful association between specific sets of cues and a bicyclist's expectation that a hazardous situation will arise. If the associations are powerful enough, it would be difficult for bicyclists to avoid becoming more alert and attentive when such cues are encountered in the traffic environment. The sole purpose of this type of training -- often referred to as hazard-recognition training -- is to increase a vehicle operator's level of alertness and attentiveness under selected circumstances. In some instances, hazard-recognition training is all that is required. That is, once an operator is alerted to the fact that he is in a potentially hazardous situation, he has both the motivation and capability to cope with the situation.

A third task is to eliminate any uncertainties and misconceptions about the exact path that is safest. As was suggested earlier, this task cannot be accomplished by teaching bicyclists a few generalized rules. Rather, it will be necessary to demonstrate the exact course that is optimal for a large and representative sample of high-hazard traffic contexts, maneuvers, and conditions. Specifying the optimal course Is simple and straightforward for some situations; in other cases, it may be difficult or impossible with the information presently available. Although there is still a considerable amount of controversy about the course that is optimal in some situations, this fact in no way changes the need to provide highly explicit instructions on course selection.

The final task is to reinforce the education described above through exercises in course selection. Such exercises would expose bicyclists to a variety of situations and require them to identify the optimal course for each situation. The exercises must cover the full range of traffic contexts, maneuvers, and conditions. An essential part of such exercises is immediate feedback on the correctness of the bicyclist's choice and reiteration of the reasons why one course is safer than others.

Risk assessment. There are countless cases in which bicyclists select a course that they know is less safe than another they could have chosen. Riding through a stop sign is a good example; even very young bicyclists know it is safer to stop than to proceed without stopping. Such acts, when committed knowingly, are often assumed to reflect an abnormally high willingness to take risks; persons who commit such acts are often called "risk takers." However, there is no evidence that more than a minute fraction of the so-called risk takers are any more willing to accept risk than the bicycling population at large. That is, the thought of an accident, with its attendant pain and suffering, is no less repulsive to the so-called "risk taker" than to persons who ride more safely. The results of the bicycle/motor-vehicle accident study showed that most bicyclists who knowingly select a suboptimal course do so because of the fallacious belief that the added risk associated with the suboptimal course is inconsequential. In short, the problem is that of risk assessment rather than risk acceptance.

Conventionally, bicyclists acquire their notions about the relative risk of alternate courses through long-term observation of near accidents or through analytical considerations. This is an inefficient and often unreliable way to acquire knowledge about how much more risky one course is than another. Young children are at a particular disadvantage because both their experience and their analytical skills are more limited than an adult's. Consequently, an important objective of a safety-education program is to increase the validity of bicyclists' assessment of the relative degree of risk associated with optimal and non-optimal courses that may be chosen. It is presumed that this education would be administered at the same time bicyclists are taught to identify the optimal course for the various high-hazard situations.

It is unlikely that it would be possible to obtain the data needed to develop an objective, numerical index of risk for each course that could be selected in the many traffic contexts that bicyclists encounter. However, it is believed that sufficient information is available (research data, analytical findings, and expert opinion) to convince bicyclists that (a) the optimal course is significantly safer than any other and (b) the absolute risk associated with suboptimal courses is great enough to justify avoiding them.

Competing needs. Bicyclists sometimes have momentary needs that are best served by a suboptimal course. Such needs are referred to here as "competing needs" because they are in direct competition with the need for safety. A need to conserve time is an example of a competing need. Every bicyclist knows that stopping for stop signs serves to frustrate a need to conserve time; the need is better served by failing to stop. Even though a bicyclist is fully aware of the risk associated with both courses and has a normal need for safety, he will always choose to ride through a stop sign when his need to conserve time becomes very strong. If the bicyclist is rushing to the aid of a sick relative, his decision to ride through a stop sign may be altogether rational.

How does one induce bicyclists to select the optimal course when competing needs are present? In principle, this can be done in two ways: increase the composite need satisfaction associated with the optimal course, or decrease the composite need satisfaction associated with suboptimal courses. There is much uncertainty about how to achieve either of these results through education. One potential technique is to educate bicyclists about undesirable consequences of selecting a suboptimal course other than accidents. At present, the number of other undesirable consequences are few. However, with effective law enforcement, parental guidance, and school programs, the relative need satisfaction associated with optimal and suboptimal courses might be modified by informing bicyclists of the likelihood and consequences of receiving a traffic citation and/or getting caught and punished by parents, school authorities, or both. In principle, this technique would serve to increase the positive value of the optimal course and increase the negative value of the suboptimal courses.

Another approach is to reduce the perceived need satisfaction associated with a suboptimal course. For instance, bicyclists might be presented with objective data that demonstrate the small amount of time saved and energy conserved by failing to stop for stop signs, taking a shortcut that requires riding against traffic, and so on.

There is no reason to expect that it will be easy to develop educational methods that will be effective in offsetting the influence of competing needs on course selection. Hopefully, one or more readers will be able to offer suggestions about how best to deal with this difficult problem.

Visual obstructions. A course that is optimal under ordinary circumstances may be highly hazardous when an object is present that obstructs the bicyclist's view, the motorist's view, or both. The failure to adopt a course that best offsets the effects of a visual obstruction may be due to the bicyclist's failure to observe the obstructing object, his failure to recognize that the object obstructs his view of a critical part of the traffic environment, or his lack of knowledge about the course that minimizes accident likelihood in such situations. Therefore, education is needed to accomplish the following:

• Increase bicyclists' ability and inclination to search for and recognize objects that obstruct their view, the motorist's view, or both.

• Increase bicyclists' understanding of the degree to which visual obstructions may reduce response time and, thereby, decrease the possibility of successful evasive action.

• Teach bicyclists the best course to follow to compensate for visual obstructions.

Teaching bicyclists to recognize and cope with visual obstructions is a special case of the education to increase their ability to identify the optimal course through an area (discussed above). In order to accomplish the above objectives, bicyclists must be taught the types of objects that frequently obstruct vehicle operators' views, the types of locations where critical visual obstructions are frequently encountered, the types of accidents that most frequently result (wholly or in part) from visual obstructions, the relationship between the size of the obscured field and the size and distance of the obstructing object, and the exact course (speed and path) that should be followed in each of a wide range of the traffic contexts where visual obstructions often contribute to accidents. Young bicyclists have the most urgent need for such education, but the need is by no means limited to juveniles.

Invalid expectations. All vehicle operators develop a set of expectations about the physical characteristics of the traffic system and about the behavior of those who use the traffic system. This set of expectations has an important influence on both the path and speed a vehicle operator chooses to travel. Some expectations are developed from a knowledge of the laws and ordinances that govern the behavior of the various users of the traffic system. Other expectations are based upon direct observations of the physical characteristics of the traffic system and the behavior of persons who use it. Vehicle operators frequently develop expectations that do not correspond with reality. These invalid expectations usually stem from the assumption that the physical characteristics of the traffic system and the behavior patterns of roadway users are more predictable and uniform than they are in fact.

The results of the study of bicycle/motor-vehicle accidents showed that invalid expectations were a frequent contributing factor to bicyclists' selection of a suboptimal course. [8] Invalid expectations most often led bicyclists to travel at an excessive speed, but path selection was adversely affected in a significant number of cases. Invalid expectations that frequently had an adverse influence on course selection include: expectations that an area will be void of motor-vehicle traffic, expectations that motor-vehicle operators can and will observe bicyclists, and expectations that motor-vehicle operators will always behave in a lawful manner. There were a small, but important, number of bicyclists whose suboptimal course resulted from invalid expectations about the behavior of another bicyclist. The most important expectations that must be corrected through education include:

• The expectation that all traffic on intersecting roadways will stop/yield in accordance with the law.

• The expectation that all traffic in opposing lanes will yield before turning left -- in accordance with the law.

• The expectation that a riding companion will select a safe course.

• The expectation that a specific roadway will be void of all traffic at a specific time.

• The expectation that bicyclists will always be observed by motorists when visibility conditions are good.

• The expectation that lawful lighting equipment on a bicycle ensures that it will be observed at night by all motorists.

• The expectation that parallel-parked vehicles will not be occupied.

Probably the best way to eliminate invalid expectations is to illustrate and discuss the types of accidents that result from invalid expectations. Invalid expectations cannot be eliminated by instructing the bicyclist to "expect the unexpected." Such worthless bits of advice are worse than no education at all.

Time/space required to respond to threat. Much of the instruction discussed above presumes that bicyclists will be taught to estimate, with reasonable accuracy, the time and space required to stop and to change directions as a function of such factors as: bicycle speed, roadway-surface condition, the direction and magnitude of the roadway slope, the bicyclist's reaction time and vehicle-handling skill, the type of bicycle, and the type of brakes. It would be quite impossible to define a safe course without knowing the amount of time and space required to reduce the bicycle's speed and/or change directions in response to an actual or potential threat.

Unfortunately, no data are available to use in estimating the frequency with which bicyclists select a suboptimal course because they underestimate the time/space required to stop or to turn. However, there is evidence that bicyclists sometimes delay the initialization  of evasive action unnecessarily because they misjudge their ability to stop or turn under unusual conditions. The study of bicycle/motor-vehicle accidents revealed a. small, but significant, number of accidents that were caused partly by the bicyclist's misjudgment of the time/space required to stop/turn when riding on a wet roadway, when riding with wet caliper brakes, when riding down a steep slope, and when riding on a roadway surface covered by sand or gravel. There were also a few cases in which the bicyclist misjudged the amount of time required to grasp and manipulate the caliper-brake levers. If these misjudgments adversely affect evasive actions, it seems reasonable to assume they also would adversely affect course selection.

Some attention has been given to educating bicyclists about the stopping distance of both bicycles and motor vehicles. Several films have been produced for this purpose. Additionally, tables and graphs have been developed for use in demonstrating the relationship between stopping distance and vehicle velocity. In the author's view, classroom instruction must be supplemented with outdoor training and demonstrations. To be maximally effective, the training must cover a wide range of speeds and a wide range of conditions (wet roadway, wet caliper brakes, traveling downhill, and sand-covered roadway surface). Moreover, the training must address both stopping distance and turning radius.

Length of amber-signal phase. In order to counter the bicycle/motor-vehicle accidents that occur at signalized intersections, bicyclists must be taught to avoid entering the intersection after the onset of the amber phase. An important part of this education is to inform bicyclists of the length of the amber phase and the distance a bicyclist can travel during this brief period. Since the length of the amber phase is variable, bicyclists should be taught to base their decisions on the shortest amber phase that may be used for roadways with two, four, and more than four lanes. Bicyclists must also be taught exactly what to do if they are unable to clear the intersection before the light turns red. The primary objective, however, is to teach bicyclists to avoid this situation.

EDUCATION TO ENHANCE REACTIVE-PHASE FUNCTIONS

The Reactive-Phase functions are those required for a bicyclist to observe a motor vehicle that poses a threat and to perform the actions necessary to avoid a collision with that motor vehicle. Specifically, the bicyclist must: (a) search the relevant portions of the environment for threatening vehicles, (b) detect the presence of vehicles that constitute a threat, (c) assess the velocity vector of the other vehicle with respect to his own and judge whether the vehicles are on a collision course, (d) if the vehicles are not on a collision course, determine whether a probable action by the motor-vehicle operator could place the vehicles on a collision course, (e) identify the action that is most likely to result in accident avoidance, and (e) perform the motor behavior required to implement the action decided upon. When defining accident causation, a function failure during the Reactive Phase can occur only if the threatening motor vehicle could have been observed soon enough for the bicyclist to have successfully completed all of the Reactive-Phase functions. Thus, when critical visual obstructions are present, it must be said that the critical function failure occurred during the Anticipatory rather than the Reactive Phase.

The types of Reactive-Phase function failures that most often contribute to bicycle/ motor-vehicle accidents are reflected in the educational objectives that were formulated to enhance the performance of the Reactive-Phase functions. In the following discussion of these objectives, the description of each Level I objective is followed immediately by a description of the associated Level II objectives. The Level I and Level II objectives are summarized in Table 24.

Increase inclination and ability to search selectively and to recognize cues signaling the presence of a threat.

Increase validity of expectations that may influence bicyclists' assessment of the need to search.

Increase knowledge of stimuli that may distract attention, and increase ability to cope with distractions.

Increase ability to cope in situations where information-processing capacity is overloaded.

Increase the validity of bicyclists' assessment of the degree of risk associated with failures to search.

Increase bicyclists' ability to make critical spatial judgments.

Increase bicyclists' ability to execute emergency braking, turning, and controlled slides.

Education to Enhance Search Behavior

The first Level I objective listed in Table 24 is to "increase bicyclists' ability and inclination to search effectively for motor vehicles that pose a threat." This is the most important single objective discussed in this report. The data on bicycle/motor-vehicle accidents showed that a search failure by the bicyclist contributed to 50% of the fatal and 41% of the non-fatal accidents. In all of these cases, the motor vehicle with which the bicyclist collided could have been observed early enough for the bicyclist to have avoided the accident; the accident occurred because the bicyclist failed to search in the motor vehicle's direction until it was too late to avoid the collision. Unquestionably, education to enhance bicyclists' search behavior has great potential for reducing bicycle/ motor-vehicle accidents; it is probable that such education would effect a reduction in NMV accidents as well.

Significant improvements in bicyclists' search behavior cannot be achieved by merely informing them of the importance of visual search and advising them to increase their search activity. Such instruction has no more effect on bicyclists' behavior than telling them to "ride safely." Rather, what is needed is highly specific instruction on where bicyclists must search in various traffic contexts, the types of factors and events that may momentarily disrupt search behavior, and the types of situations in which effective search is difficult or impossible without a substantial reduction in speed. The Level II objectives described below reflect the author's views on the instruction needed to enhance bicyclists' search behavior. These objectives were formulated from a careful study of the various factors that contributed to the search failures that, in turn, led to a bicycle/ motor-vehicle accident.

Limitations of the visual system. Vision is such a highly developed skill that it is difficult to keep in mind that the visual system has some highly important limitations. Because the eyes usually function without conscious effort, children and some adults tend to think of the visual system as an autonomous mechanism that automatically supplies them with all the visual information required to perform the task at hand. Because the eyes perform so many functions with such a high level of efficiency, it is difficult to avoid behaving as if the visual system is a perfectly functioning mechanism with no limitations whatsoever. These fallacious notions are not conducive to the development of effective search behavior. It is difficult to induce bicyclists to deliberately and systematically search the traffic environment if they believe that their eyes will automatically detect hazards, and it is difficult to teach bicyclists how to scan effectively if they have no understanding of the limitations of the visual system and the reasons for these limitations Therefore, it was reasoned that a basic understanding of the limitations of the visual system is a prerequisite for the development of effective search behavior.

Listed below are the educational objectives that are considered most essential. These objectives were derived analytically from a consideration of what bicyclists in the primary target group (fourth graders) must know about the visual system in order to be fully receptive to education about the necessity for visual search and the techniques required to search effectively.

• Teach bicyclists the concepts of central and peripheral vision and demonstrate differences in visual acuity for central and peripheral vision. Teach bicyclists the functions served by central and peripheral vision and why both are essential for the safe operation of a bicycle in traffic.

• Teach bicyclists the size of the central and peripheral fields of view and the extent to which these fields of view are increased by eye, head, and torso rotation.

• Teach bicyclists the concepts of scanning and fixation and demonstrate (a) the amount of time required for the eyes to search for and fixate on an object and (b) the limited number of objects the eye can locate and fixate upon per unit of time.

• Introduce the concept of information overload and explain how and why the visual system may become overloaded when riding in traffic.

• Teach bicyclists that increasing bicycle speed increases the information-processing load on the visual system; also, demonstrate this fact by exposing bicyclists to conditions that overload the visual system (high speeds, complex traffic environments, and a combination of the two).

• Teach bicyclists how to determine when their visual system is overloaded and how to compensate when this occurs.

Education on the above topics should provide the background knowledge needed for bicyclists to understand why they must learn to search selectively, why they must avoid overloading their visual system, and how visual-system overload can be avoided. Education about the limitations of the visual system is also needed to teach bicyclists why they may not be observed by motor-vehicle operators.

Selective search and threat detection. In discussing the Anticipatory-Phase functions, it was stated that searching the environment is necessary to select an optimal course through an area. The purpose of search during the Reactive Phase is to detect visual cues that signal the presence of an actual or potential threat. Because of the complexity of the visual environment and the limitations of the visual system, the likelihood of threat detection would be low if the bicyclist searched the visual scene in a random or unsystematic manner; instead, the bicyclist must learn to search selectively. Selective search means the maximum allocation of available search time to the areas where cues signaling the presence of a threat are most likely to appear.

In some instances, the cue to a threat is a motor vehicle traveling in the bicyclist's direction that obviously is on a collision course with the bicyclist. The threat is so obvious in such instances that it is unnecessary to teach bicyclists that an approaching motor vehicle is a cue to threat. However, there are many valid cues to threat that are less obvious. That is, a significant portion of the bicycling population has not learned to associate a cue with the occurrence of a threatening event. Some cues to threat are recognized by most adults but few children; others are recognized by only the most experienced bicyclist. For instance, most adult bicyclists recognize that a standing motor vehicle in the opposing traffic lane constitutes a potential threat because that motor vehicle may turn left into the bicyclist's path. This cue to threat is less apparent to young bicyclists who have not yet learned that motorists sometimes fail to observe bicyclists. Highly experienced bicyclists report that they attend to such subtle cues as:

• The scan patterns of motorists, including the direction of head movements and length of pause. The direction and length of the bicyclist's shadow in daytime (a long shadow pointing in a motorist's direction indicates that the motorist may be blinded by sun glare).

• The movement of the bicyclist's shadow at night (an overtaking motor vehicle is on a collision course with the bicyclist if the bicyclist's shadow, cast by the motor vehicle's headlights, fails to move in a right-hand direction).

• The presence of movement in the side mirror of a parallel-parked motor vehicle (signals the presence of an occupant that may open the vehicle's door).

• The presence of activated stop or backup lights or the movement of the front wheels of a parallel-parked vehicle (signaling that the vehicle may emerge from the parking space).

It is believed that the method used to enhance bicyclists' selective search and threat detection skills must meet four important requirements. First, it must provide for the teaching of selective search and threat detection in concert. It would be difficult to teach bicyclists when they should search without telling them what they are searching for. Secondly, the education must be highly specific in both its content and context. Bicyclists must be instructed on exactly where to search and what to search for, rather than being taught abstract rules and principles about selective search and threat detection; this specific instruction must be administered in a real-world context or within the context of a high-fidelity simulation of real-world imagery. Thirdly, the instruction should be aimed specifically at the types of traffic contexts in which accidents most frequently occur. The relevant traffic contexts for bicycle/motor-vehicle accidents are illustrated and described in Section V. Finally, the method should enable bicyclists to actively practice the selective search and threat detection tasks, with provision for immediate feedback after each practice trial. Lectures are useful to a point, but active practice with immediate and detailed feedback probably will be required to refine the skills to an acceptable level.

Invalid expectations. The invalid expectations that lead bicyclists to select a suboptimal course also lead them to fail to search when it is appropriate to do so. Because of invalid assumptions, bicyclists incorrectly conclude that it is unnecessary to search in a particular direction. As was stated above, most of the invalid expectations are that an area or location will be void of motor-vehicle traffic, that bicyclists will be observed by motorists, that motorists will adhere to the law, and that a riding companion will search for and detect threats.

Many of the bicyclists' invalid expectations will be eliminated by the education on selective search and threat detection. An important part of learning to search effectively and to detect threats consistently is the recognition that behavior must be guided by a consideration of both the typical and the atypical events that occur In the traffic environment.

Coping with distractions. It was found that about one-half of the search failures by bicyclists were due partly to the presence of a momentary distraction. In the vast majority of cases, the bicyclist was distracted by a riding companion. Other distractions include: another vehicle considered an accident threat, non-traffic-related mental activity, abnormal street-surface condition, unfamiliar vehicle, carrying object in hands, malfunctioning vehicle, improper size bicycle, scenic attractions, hostile animal, and inclement weather. [9]

One can only guess how many of the bicyclists would have searched effectively if the distraction had not been present, but it is likely that many would have searched. If so, considerable benefit would be realized from education that would serve to offset the effects of momentary distractions. In considering methods of accomplishing this educational objective, it must be kept in mind that few of the distractors were of the type that cause a reflexive or involuntary shift of attention, such as a gunshot, an elephant in the roadway, and so on. Rather, the distractors were common persons or things that the bicyclist voluntarily attended to because, at the moment, the distracting person or thing was considered of greater importance than traffic-related stimuli. In other words, the bicyclist was voluntarily directing his attention to environmental stimuli in accordance with his system of priorities at the moment. It follows that the only way to offset the effect of such distractors is to modify the bicyclist's system of priorities; the perceived importance of traffic-related stimuli must be increased or the perceived importance of distractors must be decreased.

Some benefit may result from a straightforward explanation of this problem, including: a description of the meaning of the word distractor, the types of distractors that are most important, the manner in which distractors may influence search behavior, and the consequences of being momentarily distracted from the search task. However, it is believed that some more active form of education and practice is needed to produce the desired behavioral changes. Unfortunately, no definitive ideas about an effective educational approach can be offered at this time.

Information overload. Some search failures occurred because bicyclists simply had insufficient time to search for and detect all of the relevant stimuli in the environment; the search requirements exceeded the bicyclist's information-processing capacity. Information overload is a joint function of the complexity of the traffic environment and the bicyclist's speed. An educational solution to this problem requires that bicyclists be taught to recognize when their information-processing capacity is becoming overloaded. That is, bicyclists must be taught to recognize when they have insufficient time to accomplish all the search tasks that are necessary in order to ride safely. If this difficult objective can be accomplished, it then becomes necessary to teach bicyclists that speed reduction is usually the best way to decrease the information-processing load to a manageable level.

The author knows of no educational techniques that have been developed to teach bicyclists, or any other vehicle operator, to recognize when their information-processing capacity has become overloaded. However, it should not be too difficult to develop a useful technique. One promising approach would involve the use of a simple cinematic simulator. A fairly low-cost system could be developed whereby bicyclists would be shown 16-mm films of typical street scenes on a variable-speed projector. With this system, the information-processing load could be increased systematically by increasing the speed of the projector from one frame per second to 24 frames per second. The information-processing load would also vary as a function of the visual complexity of the street scenes that are filmed. This system would provide the capability for increasing the information-processing load in a systematic manner, and could be used to demonstrate an information-overload condition for any bicyclist, regardless of his individual information-processing capacity. A critical requirement of such a system is an objective performance measure that would provide a valid and precise indication of when the bicyclist's information-processing capacity was approaching an overload condition.

Risk assessment. Invalid risk assessment is another factor that sometimes contributes to bicyclists' search failures. Risk assessment is a particularly important factor for, accidents that occur in quiet residential areas that appear very safe. As a consequence, education must focus on risk assessment in safe-appearing traffic contexts. The education must somehow convince bicyclists that the likelihood of an accident in such areas is great enough to warrant effective search behavior on every occasion, even though potentially threatening motor vehicles are present only rarely.

Education to Enhance Evaluation Behavior

An evaluation failure occurs when the bicyclist performs the search function and observes the threatening motor vehicle, but fails to recognize the need for evasive action. An evaluation failure was the precipitating cause of about seven percent of the fatal and 36$ of the non-fatal bicycle/motor-vehicle accidents. In each of these cases, the bicyclist observed the motor vehicle early enough to have avoided the accident; the bicyclist failed to initiate evasive action soon enough because of a misjudgment or an invalid expectation concerning the motorist's behavior. These findings make it clear that there is an important need for education to increase bicyclists' ability to evaluate situations and to recognize the need for evasive action. This is the second of the three Level I objectives for enhancing the performance of the Reactive-Phase functions. The Level II objectives are discussed below.

• Invalid expectations. In the preceding paragraphs, it was explained that invalid expectations may adversely influence bicyclists' selection of a course and their assessment of the need to search. The same types of invalid expectations often have an adverse influence on bicyclists' assessment of the need for evasive action. That is, invalid expectations lead bicyclists to the conclusion that there is no need for evasive action when, in fact, an accident is about to happen. The invalid expectations most often reported by bicyclists involved in bicycle/motor-vehicle accidents of this kind are:

• The expectation that the motorist had or would observe the bicyclist.

• The expectation that a stationary motor vehicle would remain stationary until the bicyclist had passed.

• The expectation that a turning vehicle would proceed straight ahead.

• The expectation that a stopping/slowing vehicle would proceed at a constant velocity.

• The expectation that occupants in parallel-parked motor vehicles would not open the vehicle's door until the bicyclist had passed.

• The expectation that a motor vehicle was going to turn when, in fact, it proceeded straight ahead.

• The expectation that a motor vehicle was going to turn in a direction opposite to that of its actual turn.

• The expectation that lawful lighting equipment on the bicycle would ensure that the bicyclist would be observed by the motorist.

The most important of the invalid expectations is the expectation that the bicyclist had been or would be perceived by the motorist. In recognition of this fact, some existing educational materials instruct bicyclists to establish eye contact with the motorist before assuming that they have been observed. This education implies that the bicyclist can safely assume that he has been observed by the motorist if he can see that the motorist has scanned in his direction. This education is counterproductive; many instances were found in which bicyclists reported that they decided to proceed only because they observed that the motorist had looked directly at them. Most motorists involved in accidents of this kind verified the bicyclist's claim that they had searched in the bicyclist's direction, but still insisted that they had not observed the bicyclist. In short, the direction of a motorist's search is not a valid indication of what he has observed.

The education on the limitations of the human visual system should prove useful in increasing the validity of bicyclists' expectations about being observed by motorists. That is, bicyclists who clearly understand that a motorist's visual system is subject to the same limitations as their own will be less likely to assume that they have been or will be observed by motorists. However, education on the limitations of the visual system is not enough. Bicyclists must also be given highly explicit instructions on how to behave when the actions of motorists cannot be predicted with a high degree of reliability. For instance, when a bicyclist encounters a motor vehicle waiting to enter the roadway from a driveway or alley, the bicyclist must be taught that the motor vehicle may proceed into the roadway without having observed the bicyclist. But, what is a bicyclist supposed to do when confronted with this situation? The bicyclist must be taught to modify his speed and/or path in a manner that provides sufficient time for evasive action in the event that the motor vehicle does, in fact, proceed into the bicyclist's path. Decreasing speed, in turn, decreases stopping distance; moving left increases the buffer zone between the bicyclist and motorist and thereby provides additional time and space for evasive action. Since it is not possible to formulate any generalized principles or rules about the best way to respond in situations of this kind, bicyclists must be instructed on how to respond in each of a wide range of specific situations and traffic contexts.

It has been suggested that bicyclists should be taught methods for attracting motorists' attention in situations where accidents could occur because the motorists fail to observe the bicyclists. Hand signals, voice warnings, and the use of auditory-warning devices (bells or horns) have been suggested. Few expert bicyclists are enthusiastic about this approach to the problem. Although voice warnings are effective in some situations, they cannot be relied upon in all situations because motorists often drive with their windows closed and with their radio playing at a high volume. Hand signals require the bicyclist to remove one hand from the handlebars at a time when control of the bicycle may be highly critical. Although hand signals may be used effectively in some situations, there are other situations in which the hands would be more effectively used in steering the bicycle, braking, or both. Auditory-warning devices have the disadvantages of both voice warnings and hand signals; they require that the bicyclist remove one hand from the handlebars in order to operate them, and one cannot depend upon the motorist hearing the warning device under all conditions.

One expert bicyclist reported that he installed a powerful air horn on his bicycle for use in alerting motorists of his presence. His first opportunity to use his air horn came when a motor vehicle in the opposing traffic lane turned left in front of him. The blast of the air horn startled the motorist to such an extent that he came to a complete stop in the bicyclist's path, making it impossible for the bicyclist to avoid an accident. Anecdotal evidence such as this suggests that extreme caution should be used before deciding to teach bicyclists to use some form of signal to attract motorists' attention.

Critical spatial judgments. There were relatively few bicycle/motor-vehicle accidents that resulted from faulty spatial judgments. A small number of bicyclists misjudged the space required to clear an opening door of a parallel-parked motor vehicle. Additionally, there were a few instances in which a bicyclist was clearly riding too far to the left because he misjudged the space required for a motor vehicle to overtake and pass him. This type of misjudgment is most often a factor in night accidents on narrow roads; bicyclists are inclined to ride as far left as possible to avoid roadside debris that would be difficult to see at night, so they sometimes ride so far left that they are well within the path of overtaking motor vehicles.

Although it is believed that some attention should be devoted to increasing bicyclists' ability to make such spatial judgments, this educational objective appears to be among the least important objectives discussed in this report.

Education to Enhance Selection/Performance of Evasive Actions

The evidence on bicycle/motor-vehicle accidents indicated that only about three percent of the accidents were clearly the result of an incorrect choice of evasive actions or an inability to execute the evasive action decided upon. In most of these cases, some type of unusual condition contributed to the bicyclist's failure to select the correct evasive action or his failure to initiate the evasive action soon enough to avoid the accident.

It must be admitted that it is extremely difficult to evaluate the appropriateness and effectiveness of a bicyclist's evasive actions from post-accident interview data. It is particularly difficult to judge whether or not a high level of proficiency in performing emergency turns and stops would have enabled a bicyclist to avoid the accident. As a consequence, the potential benefits of increasing bicyclists' ability to select and execute optimal evasive actions remains somewhat uncertain. This Level I objective was included because much of the education needed to accomplish this objective is also required to educate bicyclists on course selection. The Level II objectives are described below.

Stopping distance and turning radius. Judicial decisions about the optimal evasive action must be based upon an ability to estimate accurately the stopping distance and maximum turning radius of the bicycle. Since stopping distance and turning radius are influenced by a variety of factors, the education should teach bicyclists to estimate stopping distance and turning radius as a function of such factors as: bicycle speed, roadway gradient, type of bicycle, type of brakes, condition of brakes, and roadway-surface condition.

There appears to be no type of classroom instruction that would be effective in enhancing bicyclists' ability to judge stopping distance and turning radius. Outdoor training with repetitive trials under a range of conditions appears to be the most effective way to teach bicyclists the necessary judgmental skills, but this approach would be highly costly and time consuming. If it is impossible to develop a less costly educational approach, there is serious reason to question whether or not this objective should be included in an educational program.

Emergency stops, turns, and slides. Expert bicyclists report that the ability to execute emergency stops, turns, and slides has enabled them to avoid accidents that could not have been avoided by bicyclists who do not possess these skills. Although no empirical data are available to verify these opinions, the anecdotal evidence is sufficiently impelling to consider training in emergency evasive action as a possible objective. Based upon the author's limited knowledge, it appears that training in emergency evasive action requires that bicyclists be instructed on the appropriate vehicle-handling procedures and be given the opportunity for supervised practice until the skills have been refined. No information is available on the amount of time that would be required to acquire the necessary skills, but it is probable that a considerable amount of practice would be required to achieve a high level of proficiency. Clearly, it will be necessary to make a more objective assessment of the accident-reduction potential and the training time/costs before it will be possible to evaluate the cost-effectiveness of training in emergency evasive actions.

Contents copyright © 1978,
AAA Safety and Educational Foundation
Republished with permission
Internet edition prepared by John S. Allen