Top: John S. Allen's
Up: Table of Contents
Previous: Collection of data
Next: Specific variables
ANALYSIS OF DATA ON AGGREGATE BASIS
This chapter analyzes data from all respondents on an aggregate basis. That is, data from all returned questionnaires are used to arrive at figures for average annual miles traveled; average trip length, overall accident experience, and similar items. The following chapter analyzes the data by examining similar variables but characterized by the respondent's bicycling experience, geographic location of residence, age, sex, and similar categories.
GEOGRAPHIC AND POPULATION GROUP DISTRIBUTION
A total of 3,270 usable questionnaires were received with responses from all 50 states and the District of Columbia. It should be noted that all subjects did not provide responses to every inquiry. For this reason, the total .response in the following analyses usually will not be equal to 3,270. A question was asked of respondents as to the population size of the metropolitan area where they lived. Five choices were provided, ranging from greater than one million to less than 5,000, which was classified as rural. Table 2 shows the distribution that resulted from this question.
DISTRIBUTION OF RESPONDENTS BY METROPOLITAN AREA SIZE
Although not intended, the returns showed an almost even split between areas greater than 250,000 population and those less than 250,000. This particular split corresponds closely to that of the United States as a whole (41). However, a goodness of fit test (p=.05) showed that a significant difference exists between the overall distributions.
Bicycling ease is dependent on both the gearing of the bicycle and the topography of the land. Table 3 shows the response to the questions regarding the topography of the area where the respondent lives, and the topography of the area where the respondent does most of his bicycling.
DISTRIBUTION OF RESPONDENTS WITH RESPECT TO TOPOGRAPHY
Upon examining the responses, two interesting items are noticed. First, it appears that rolling terrain does not inhibit bicycling. While 49.7 percent of the respondents stated that they lived in areas that they considered rolling, 60.5 percent of the respondents reported that they bicycled in rolling terrain. It may be presumed that with the :tremendous increase in sales of the 10-geared bicycle many cyclists actually seek out rolling topography in order to use the gears. The second item that is seen in the Table is that there are almost 20 percent less respondents who live in steep or mountainous terrain reporting that they also bicycle in that type terrain. This may show that even bicycles with 10 gears cannot be used very easily in very steep areas. However, the small sample responding in this category (less than 7 percent of the total usable data) may not be representative of the larger population.
AGE AND SEX DISTRIBUTION
Of the respondents respondents 5 Who provided information on their age and sex, 2,873 were males (88 percent) while 391 were females (12 percent). This large male sample was expected, and due, in part, to the fact that the majority of members in L.A.W. were males. The researcher still felt that comparisons between the sexes could be made with a fair degree of reliability, because almost 400 responses from women were received.
The mean age of all respondents to the questionnaire was 37.7 years, with only a minor difference between the sexes. Figure 2 shows the age distribution of the respondents. It is somewhat surprising to see the large number of persons older than 46 (28.7 percent) that were
FIGURE 2 - AGE DISTRIBUTION OF L.A.W. RESPONDENTS
Age Distribution of L.A.W. Respondents
On this figure and all those following,
N is equal to the number of respondents
to that particular question.
|Age Groups||Percent of all
|active cyclists. The oldest respondent was an 82 year old man, and
there were 54 persons above the age of 66 who were considered regular riders for the
purposes of the survey. It is also interesting to note that the latter age group-above 66
years of age--averaged more miles riding per year than any other age group. (This is
discussed in more detail later.)
Respondents were asked how many automobiles were available for their use. Table 4 shows the distribution of the respondents with respect to the number of automobiles available.
DISTRIBUTION OF RESPONDENTS WITH RESPECT TO AUTOMOBILE AVAILABILITY
It can be seen that only 5.4 percent of the respondents do not have any available autos. The age distribution showed that there are an appreciable number of respondents who are 60 years of age or older who may not need a car. Also, some of the younger respondents are probably college students who do not own a car. There were more respondents with two cars (43.1 percent) than one car (42.4) available. Multiple car respondents account for 52.2 percent of the total and there are on the average 1.6 automobiles available for each respondent. It is difficult to relate these figures to any national data because the L.A.W. survey was addressed to individuals whereas national sampling is usually made on a "per dwelling unit" basis.
A study completed in 1971 in Reston, Virginia, a "new town" type suburb of Washington, D.C., showed a very similar distribution of auto availability, based on dwelling units (30). It is likely that the respondent in the bicycle user survey was, in fact, providing information on automobile availability for the household, because the respondent appears to be the household head. If that is the case, the socio-economic status of the L.A.W. membership would be comparable to that of Reston, that is, of higher middle class suburban characteristics.
BICYCLE TYPE AND EQUIPMENT
There was an overwhelming majority of respondents (96.8 percent) who stated that most of their riding was on bicycles that had five speeds or more. Table 5 shows the distribution with regard to bicycle type.
DISTRIBUTION OF RESPONDENTS WITH RESPECT TO BICYCLE TYPE
Although all types of bicycles are being sold today, the multi-speed bicycle (3, 5, or 10 speed) has taken over as today's popular model, It is estimated that some 85 percent of bicycle production today is of the multi-speed variety (14). This trend is evident in the responses received, for less than 1 percent reported that they commonly used a one-speed bicycle.
Respondents were asked whether or not their bicycles (and themselves) used certain types of equipment. The equipment ranged from rear view mirrors, lights, and helmets (safety equipment), to registration of the bicycle. Whether or not the bicycle was equipped with an odometer or other device to measure distances was also noted. This latter item played an important role in determining the accuracy of the trip lengths and overall mileage figures reported in other sections of the questionnaire. Results of these inquiries as to equipment usage are displayed in Figure 3.
FIGURE 3 -RESPONSES TO EQUIPMENT USED -L.A.W. SAMPLE
Reponses to Equipment Used - L.A. W Sample
|Does your bicycle have a rear view mirror||67%||33%||3262|
|Do you wear a helmet||59%||31%||3263|
|Do you use lights||39%||61%||3263|
|Does your bicycle have an odometer||65%||35%||3260|
|Is your bicycle registered||54%||46%||3261|
|Over one-third (35 percent) of the sampled members reported that they did have
odometers or used some other measuring device. In order to discover if any significant
differences existed in reporting mileage between those who did have odometers and those
that did not, a statistical T-test was performed on the data. Respondents who stated that
they did not have a measuring device actually reported mileage figures 4.3 percent higher
than those who did have an odometer. However, there was no statistical difference between
the two groups at the .01 level of significance. (The T-test analysis is shown in Appendix D.)
The length of time that the respondent had been bicycling prior to filling out the survey form was important to know in order to discover any differences that might exist in accident or riding characteristics between different groups. Table 6 shows the distribution of respondents with respect to their years of continuous bicycling experience.
DISTRIBUTION OF RESPONDENTS WITH RESPECT TO BICYCLING EXPERIENCE
The word "continuous" was used in the bicycling experience question in order to prevent persons from counting childhood bicycling in the total years they use as past experience. The researcher felt that many older adults who might include these years, bicycled under different conditions than those of today, and therefore it should not be considered comparable experience.
The small response received to the choice "less than one year" was probably because the questionnaire was directed at those members who could provide trip figures and mileage data for the year 1974. Since the form was mailed in March, 1975, some potential respondents could not provide 1974 information if they had begun to bicycle within the last 11 months.
Over half (51.7 percent) of the respondents stated that they had been bicycling regularly for four years or less. This time period corresponds very closely to the bicycle boom that started in the early 70's. Apparently, much of the League of American Wheelmen's membership growth in the last two or three years is made up of persons just beginning to bicycle.
TOTAL MILES RIDDEN IN 1974
The average miles ridden during the year 1974 for all respondents was 2,332 miles. This is based on an average of 8.9 months of the year that respondents rode, or about 260 miles in every month that the respondents stated were suitable for cycling. The distribution of respondents with regard to the number of miles traveled is shown in Figure 4. There are 315 respondents (9.8 percent of total) stating that they rode over 5,000 miles during 1974, and of these, 44 returns showed mileages in excess of 10,000 miles. Many of these were from persons who had done extensive touring during the year.
FIGURE 4- DISTRIBUTION OF RESPONDENTS
WITH RESPECT TO ANNUAL MILES TRAVELED
Distribution of Respondents with Respect to Annual miles Traveled
|Miles traveled in 1974||Number of respondents|
|It should not be considered common for mileages greater than 4,000 or 5,000 miles a year to be traveled on a bicycle. However, these latter annual mileages are definitely within reason, because a regular bicycle commuter traveling 15 to 20 miles a day, or a recreation bicyclist riding 50 to 75 miles each weekend (not too uncommon), will come close to totaling this many miles. Still, over a third of the respondents (34.8 percent) reported that they traveled less than 1,000 miles during 1974. For comparison purposes, it should be noted that the average passenger car traveled 9,992 miles in 1973,. and the average motorcycle traveled 4,498 miles (23). Map 2 shows the average miles traveled in each state. Map 3 shows the average months ridden in each state. As expected, those states with milder winters experienced both more miles traveled.and more average months in which riding took place.|
MAP 2 - AVERAGE MILES RIDDEN DURING 1974
(Same data in table form in Appendix B)
MAP 3 - AVERAGE MONTHS
RIDDEN DURING 1974
(Same data in table form in Appendix B)
|As mentioned previously, over one-third of all respondents stated that their bicycles
were equipped with odometers. The student's T-test, at the 99 percent level of confidence,
showed no significant difference in mileages reported by those who had odometers and those
who didn't (Appendix D). The test is based on over 3,000
respondents with combined bicycle riding of more than 7 million miles.
Table 7 shows the distribution of responses to the question that asked "How much bicycling do you think you will do in the current year as compared to the past year?" Only 176 responses (5.4 percent) stated that they would ride less or much less. Over 1,800 respondents (almost 60 percent) reported that they will ride more or much more. There were many comments on this question with a number of persons stating that they planned to begin riding to work by bicycle. Also, a few respondents mentioned plans for riding on the Bikecentennial route this summer from coast to coast.
DISTRIBUTION OF RESPONDENTS
Respondents were also asked where most-of their riding occurred. They were provided with four facility types to choose from and asked to provide the percentage of miles they rode in each facility. The results are shown in Figure 5.
FIGURE 5 - PERCENT OF TOTAL MILES RIDDEN ON EACH TYPE FACILITY
Percent of Total Miles Ridden On Each Type Facility
|Facility Type||Percent of Total||N=|
|On-Street Facility (Lane, Route)||3.2%||705|
|Off-Street Facility (Path)||3.5%||663|
|It should be noted that less than 25 percent of the respondents (N=705, N=663)
reported any travel at all on special bicycle facilities including bicycle routes, lanes,
or separate pathways. This travel amounted to less than 7 percent of the total miles
traveled by all respondents. This may be due to the relative scarcity of special
facilities (although almost every city has made an attempt at installing some type of
bikeway), or the fact that most respondents did not find any advantage to using a bikeway,
and therefore did not.
TRIP PURPOSE AND LENGTH
A section was provided on the survey form that requested each respondent to provide information on the total average number of trips taken and the average mites traveled per month for each of five distinct trip purposes. Figure 6 shows the distribution of respondents who rode for each trip purpose.
FIGURE 6: DISTRIBUTION OF RESPONDENTS
WITH RESPECT TO TRIP PURPOSE - L. A.W
Distribution of Respondents with Respect to
Trip Purpose - L.A.W.
|Trip Purpose||Percent of Respondents
|Almost 85 percent of all respondents reported using a bicycle for recreation riding
and/or touring, and almost half reported using a bicycle for either work/school commute
trips or utility trips. More than a third of the respondents stated they made trips for
the sole purpose of exercising. Less than 10 percent reported that they raced. The League
of American Wheelmen had acquired a reputation as being primarily a touring organization.
Apparently, many new members are using bicycles for commute-type trips and other utility
purposes in addition to recreation and touring.
Table 8 provides details on the characteristics of each trip type including the average round trip length. It is seen that although the recreation and touring trip accounts for over 50 percent of the total miles traveled each month, the work or school commute trip accounts for almost a third of the total trips reported each month. This, of course, is because the commute trip is made much more frequently than other types. The utility trip category accounted for a surprising 17.6 percent of all trips but only 6.4 percent of the total miles. The low round trip distance of 4.5 miles accounts for this difference. The overall average number of trips and average number of miles traveled by all respondents each month are 9.8 and 132.6, respectively. This reduces to an average round trip length of 13.5 miles for all trips.
TRIP CHARACTERISTICS BY PURPOSE - L.A.W.
Total trip percentages by purpose were tested against those of total miles to note if a significant overall difference exists between the variables. The goodness of fit test (p=.05) showed the results were significantly different; the one variable could not be used as an indicator of the other.
When asked "What percentage of the total miles you traveled in 1974 was on weekdays?", respondents reported that 46.4 percent of their riding occurred on Monday through Friday, and the remainder (53.6 percent) on weekends. When these figures are studied in conjunction with the trip characteristics table, it appears that some recreation and exercise trips are being made during the week, because utility and commute trips (assumed to be made on Monday through Friday only) account for less than 25. percent of the total miles.
An attempt was made to obtain a "snapshot" of bicycling activity during a one-week period. The questionnaire asked the respondent the number of trips taken in the week immediately prior to the time the form was completed. In most cases, the seven-day period fell during the last weeks. of winter or the early weeks of spring, not ideal bicycling weather. Throughout the country, the weather was cold, snowy, or rainy with the possible exception of a few mild winter states. Even under these conditions, over 67 percent (2,188 of the 3,256 responding) stated that they had made at least one trip by bicycle during the week. Of these, almost one-third (31.2 percent) reported making 5 or more trips which suggests commuting to work or school. Many of the respondents living in states that have harsh winters with season-long snow cover reported that their bicycles were undergoing winter "tune-ups". One man from Alaska who reported no trips made in the week commented, "There's four feet of snow on the bike path:"
There were 3,249 responses to the question, "Did you have a collision or serious fall on your bicycle last year?" Of these, 694 (or 21.4 percent) answered "yes" and 2,555 (78.6 percent) responded "no". Upon examination, the actual number of collisions or serious falls experienced by these 694 individuals showed that 126 persons, or 18.2 percent, were involved in more than one incident. Table 9 shows this distribution.
DISTRIBUTION OF RESPONDENTS REPORTING ACCIDENTS
The terminology "collision or serious fall" was used in place of "accident" in order to allow some judgment on the part of the respondent as to what he considered a serious mishap. Based on the grand total of 7,546,287 miles reported by all respondents in 1974, the accident rate is 113 per million bicycle miles traveled. Appendix E contains a complete state-by-state breakdown of accident rates based
on the total mileage and accidents reported by all respondents living in that state. It should be noted that this rate reflects collisions and falls that resulted in both bicycle damage and minor scrapes and bruises, as well as those accidents that required professional medical treatment.
In order to stratify the accident data, four categories, each describing a certain seriousness of injury, were provided for the respondent to select. Table 10 shows this distribution of injury severity for all accidents reported.
DISTRIBUTION OF RESPONDENTS WITH
As explained earlier, some respondents provided information in this and the following sections on incidents that occurred in other years in addition to 1974. However, in the determination of all collision rates on a per mile basis, adjusted values that reflect 1974 mileage traveled are shown in the tables and figures.
Less than 30 percent of the collisions and falls resulted in injuries where professional medical treatment was needed. A total of 193 respondents (1974 adjusted figure), or 27.8 percent of the total persons reporting at least one fall or collision, required at least an emergency room or doctor visit. This total expands to 237 actual incidents when adjusted for those reporting more than one collision or fall, or a rate of 31.4 serious accidents per million bicycle miles. This incident rate (31.4) is equivalent to an injury requiring some type of medical treatment every 31,800 miles, or one injury approximately every 14 years of riding for a L.A.W. member. When only the very serious injuries are examined (those requiring extended medical treatment), the rate is 7.6 per one million miles, or one such accident every 132,000 miles. This means that an L.A.W. member might experience this type of injury once in 57 years. Of course, other factors, such as the sex of the rider, years experience, city size, and other variables will enter into this estimate. They are examined in more detail in the next chapter.
It is also important that one does not attempt to apply these accident rate values to the general bicycling public. As shown later, cycling experience tends to play an important role in accident involvement (along with age and other factors). L.A.W. members are probably more experienced cyclists than the average rider, and probably travel 5 to 10 times as many miles as the average cyclist does.
The cause of the collision or fall is, of course, important to determine. However, the researcher felt that since official accident records were not studied for the incidents that the respondents discussed, certain information was not available to make a fair determination of cause. Also, many incidents were not reported. Therefore, information on only the other vehicle, or object collided with, was requested along with where the accident occurred and what trip purpose was underway at the time of the crash. Table 11 shows the 7 distinct classifications of crash occurrences and the total number of all incidents reported by collision type. Also shown are those incidents that required professional medical treatment, classified as "serious" collisions.
RATIO OF SERIOUS ACCIDENTS TO ALL ACCIDENTS BY CRASH TYPE
a) Railroad crossing is part of "Fall" category but is shown separately for closer study.
b) This table is for all accidents reported, not 1974 only.
NOTE: The goodness of fit test (p = .05) showed the percent of "serious collisions" by crash type was not significantly different than those percentages for "all collisions." Therefore, if data of only one type is known, the other may be reasonably estimated.
The "seriousness ratio" shown is calculated by comparing the percent of serious crashes to the percent of all crashes reported for each type of collision. The higher the ratio, the more serious that type of collision. As seen in the Table, collisions with moving motor vehicles had the highest ratio and collisions with dogs or other animals also ranked high. The "all other" category included many incidents of being forced off the road by a motor vehicle and falling or striking an object. This probably accounts for the high seriousness ratio of 1.4 in this category.
The most frequent crash type, listed simply as "fall", contains incidents where the bicyclist did not actually collide with another object other than the pavement or riding surface. The comments received from respondents indicated that about 60 percent of the accidents in this category can be classified as bicyclist error. A bicyclist turning too sharply going around a corner and subsequently falling would be an example of this. In the remaining 40 percent of the incidents classified as a fall, a poorly-maintained road surface containing large pot holes or longitudinal cracks, rocks, gravel, or other debris caused tie bicyclist to fall.
All other reported collisions or falls which did not fit in one of the first seven categories were grouped together. These included a few cases of the bicyclist being forced off the road by an auto or truck and perhaps then hitting a tree or other object. Also included were a few cases of mechanical failure of a bicycle component that led to an accident. Only 26 respondents reported involvement in this latter accident type which seems very low when it is realized that over 7 million miles of bicycle travel were studied. It may be that the respondents maintain their bicycles very well and, if a component does break, they may handle the situation well enough to prevent an accident from occurring.
It is surprising to note from Table 11 that crashes classified as falls and including those involving railroad crossings accounted for over 43 percent of all accidents and over 38 percent of serious injuries. These crash types could be considered as single-vehicle incidents. On the other hand, collisions with motor vehicles showed a lower than expected percentage (28 percent) of the total accidents requiring medical treatment. There were a very small number of pedestrians struck by bicycles, but dogs and other animals represented over 10 percent of the serious accident cases reported. In addition, there were numerous comments from respondents on the many "near misses" due to dogs.
A large number of bicycle-bicycle collisions also were reported (149 out of the total 872 accidents). The separation of motor vehicle and bicycle traffic is usually suggested as the inherent solution to reduce the number of bicyclist injuries and deaths. This is probably quite true for fatality cases. However, the results as displayed in Table 11 show that almost 60 percent of all serious injuries reported could just as well have occurred on a bicycle path completely separated from automobile traffic. This is a fact worth considering when evaluating the safety benefits of a separate bikeway.
Additional information requested regarding accident experience included the location of the collision or fall, and the trip purpose engaged in when it happened. These results are shown in Table 12. A comparison of accidents by location and activity as related to miles traveled in each category is discussed in Chapter V.
DISTRIBUTION OF RESPONDENTS WITH RESPECT TO LOCATION AND TRIP PURPOSE AT TIME OF ACCIDENT
It should be remembered in looking at Table 12 that although over 88 percent of the collisions occurred on major and minor streets having no bicycle facilities, the vast majority of the miles traveled by respondents were on these facilities. This relationship is discussed further in the following chapter.
A last item directly involving accident experience on the questionnaire asked those respondents who reported having one or more accidents in 1974 how many of these were officially reported to police authorities. Only 124 falls or collisions (less than 18 percent of the total experienced) were reported. This number represents a little over 64 percent of all the incidents requiring professional treatment.
This would appear to substantiate many claims that bicycle accidents are underreported. There was evidence in the comments of attempts at reporting some of the incidents mentioned but, in most cases, the respondent stated that the police refused to file an official written report.
Recipients of the questionnaire were also asked a question relating to the minimum air temperature at which. they usually would still ride a bicycle: Interestingly, although the average minimum temperature of the 3,073 respondents who answered this question was 29.6 degrees Fahrenheit, there was quite a range noted when each individual state was studied. Persons in colder states, such as Wyoming and Montana, stated that they usually rode until the temperature was below 15 degrees. Warmer state residents, such as Florida and California, showed a minimum over 35 degrees. The several respondents from Alaska stated a temperature of 36 degrees was necessary for them to ride a bicycle, possibly due to long periods of time when snow is on the ground. Map 4 shows these minimum temperatures for each state. Of interest is the fact that there were 199 respondents (6.7 percent of the total) that stated that they still rode their bicycle when the temperature was zero degrees or below.
MAP 4: MINIMUM TEMPERATURE
BELOW WHICH RESPONDENT
WOULD NOT USUALLY RIDE A BICYCLE (DEGREES RAHRENHEIT)
(Same data in table form in Appendix B)
|Other attitudinal questions asked the frequency of riding done after dark and in the rain. The distribution of the responses to these queries along with those to a question asked concerning obeying the vehicle laws is shown on Figure 7.|
FIGURE 7 - RESPONSES TO ATTITUDINAL QUESTIONS - L.A.W.
|How often do you ride after dark?||-||-||15.3%||49.5%||35.1%||3245|
|How often do you ride in the rain?||-||-||9.6%||66.2%||24.2%||3217|
|Do you obey vehicle laws that
apply to you as a bicyclist?
|It is seen that only 64.8 percent of the respondents ride after dark either
occasionally or frequently. This is surprising since the League of American Wheelmen
members probably represent the most active riders in the country. A close relationship
exists between the number of respondents who use lights (60.5 percent) and those that do
ride at night. It appears from this that about 4 percent of the respondents ride at least
occasionally at night with no lights.
Over 75 percent of the respondents stated that they rode in the rain; many commented that they also rode in snow. Apparently, many bicyclists feel that darkness is more of an inhibitor to riding than weather and temperature conditions. Accident rates for riders who do ride in the rain and after dark are discussed in the next chapter.
With regard to the question of obeying vehicle laws, only 50.2 percent stated that they "always" obeyed the laws that applied to them as a bicyclist. However, there were many comments that the applicable laws were unknown or vague in the respondent's mind. This problem is presently being studied by the U.S. Department of Transportation (32). A special report on bicycle laws in the United States was published in September 1974 specifically discussing many of the contradictory and confusing laws now in existence (6). The National Committee in Uniform Traffic Laws and Ordinances(NCUTLO) will meet in 1975 to discuss updating the bicycle section of the Uniform Vehicle Code. This latter document is a guide that most states adopt as their vehicle laws with minor alterations (40).
Returning to the question concerning law obeyance on the questionnaire, a large number of the 47.3 percent who checked that they "usually" obeyed laws stated that they obeyed all laws, except they usually "slid by" STOP signs if no traffic was coming. This treatment of the STOP sign as if it were a yield symbol is commonly practiced by many cyclists, including the author, at deserted intersections. Many times, a bicyclist can come to an almost complete stop while still on the bicycle and continue on without actually having his foot hit the ground. The definition of "stop" for bicyclists has undergone considerable discussion in law enforcement circles and was considered as a possible change to the Uniform Vehicle Code. A special panel on bicycle laws (as part of NCUTLO) recommended that no changes be made, and that bicyclists continue to come to a full stop at all STOP signs. They did note that the bicycle is most unstable while being started and stopped, and increases the bicyclist's danger, but they felt that it would be confusing to have different meanings for a STOP sign (35). It may be that many L.A.W. members realize this danger and therefore commonly do not make a complete stop at all STOP signs. Accident experience analysis categorized by the different responses to the question of law obeyance is discussed in Chapter V.
Top of page