Top: John S. Allen's Home Page
Up: Table of Contents
Previous: Introduction
Next: Classifications


II. RESULTS OF STUDY -
DESCRIPTION OF SAMPLE

A. Year of Accident (n=516) (9).

Of the 516 accidents, 45 percent occurred in 1979 and 55 percent in 1980 (10).

B. Month of Year (n=513)

The majority of accidents occurred during the summer months, from June through August (54 percent). This is consistent with statistics provided by the Massachusetts Department of Public Works for the MAPC region as a whole (see Figure 2). Although no information on comparative ridership exists for the study area, the report Bicycle-Traffic Volumes in Metropolitan Boston, covering primarily Boston and its immediate neighbors, shows a less steeply peaked distribution (11). The difference may be due to a higher proportion of children in the study area relative to the area in which bicycle counts were undertaken. In this case, it is assumed that children are more likely to ride in summer and to have accidents. Additional work is necessary to determine the relationship between accident courts and ridership.

The accidents in the MAPC study showed a greater tendency to cluster during the summer months than in the Cross study (p. 117), which included two cities with year-long moderate weather in the sample.


Figure 2: Accident Frequency by Month

fig2.gif (11257 bytes)


C. Day of Week (n=512)

Accidents were more likely to occur on weekdays, with Friday being the highest frequency day (17 percent) and Sunday the lowest frequency day (10 percent). Figure 3 displays these results. This is consistent both with the MAPC region and the Cross study (p.112). This variable was not studied in Bicycle Traffic Volumes in Metropolitan Boston.


Figure 3: Accident Frequency by Day

fig3.gif (10583 bytes)


D. Time of Day (n=479)

Accidents occurred during different time periods on weekdays and weekends. Weekday accidents were concentrated during the afternoon peak hours, between 3 - 7 p.m. (42 percent). Weekend accidents were more likely to occur during the mid-day period, between 10 a.m. - 3 p.m. (46 percent). These and the percentages for the other periods are shown in Figure 4.


Figure 4: Time of Day

fig4.gif (27623 bytes)


E. Light Conditions (N=488)

Over 82 percent of accidents occurred during daylight (see Figure 5). The Cross study found a similar percentage of daylight accidents (85 percent), and noted this was consistent with several other studies of bicycle-motor vehicle accidents (p. 116).


Figure 5: Light Conditions

fig5.gif (5112 bytes)


F. Weather Conditions (N=481)

Most accidents occurred on clear days (88 percent). Cloudy (5 percent) and rainy (5 percent) weather were the next most likely conditions under which accidents occurred. Snow was reported in less than 2 percent of the cases. These findings are consistent with the Cross study (p. 118).

G. Road Surface (n=472)

Not surprisingly, given the above weather conditions, most of the accidents occurred on dry surfaces (91 percent). Wet surfaces accounted for 8 percent of the accidents, and snowy surfaces for less than 1 percent. Cross does not report on this variable separate from weather.

H. Road Condition (n=454)

Almost all of the accidents - 97 percent - occurred on roads with no defects. Another 3 percent occurred on roads with holes, ruts, foreign matter, or other non-ideal conditions. Over twelve percent of the accidents did not report on this item. These findings are different from those in the Cross study (P.135). They also reflect the judgement primarily of operators and police, who filed most of the reports studied. As so few bicyclists completed reports, it is not possible to determine whether their greater sensitivity to the condition of the road would result in a more critical judgment.

I. Age of Cyclist (n=382)

Table 1 shows the distribution of the ages of bicyclists involved in accidents, using the same categories chosen for the Cross study. Unfortunately, the cyclist's age was not given on 26 percent of the accident reports. Percentages both including and excluding these unlisted ages are shown.

As can be seen from Table 1, cyclists between the ages of 6 and 19 accounted for over 65 percent of the accidents in the MAPC study. Although this is high, it is still less than that accounted for in the Cross study (p. 83).

Age was not recorded in Bicycle Traffic Volumes in Metropolitan Boston. However, the large number of universities in the area may suggest a somewhat higher number of riders in the 17-25 age group (many of these colleges have graduate schools) than other areas with fewer universities.

Over 30 percent of the accidents on which age was given occurred to cyclists over the age of 20 years.

J. Cyclist Wearing Helmet (n=516)

In over 97 percent of the cases, the reporter did not check off that the bicyclist was wearing a helmet. In 3 percent of the cases, it was checked. However, the form of the question - a box with the instruction "Check if wearing helmet" and its obscure placement raise the possibility that many persons did not see the question.

Table 1: Age of Bicyclist

Age # of
Accidents
(n=516)
% (including
"Not Listed")
(n=516)
% excluding
"Not Listed")
(n=382)
Cross Study
(Non-Fatal)
(n=753)
<6 10 2 3 2
6-11 81 16 21 28
12-15 89 17 23 37
16-19 80 16 21 14
20-29 66 13 17 12
30-44 37 7 10 4
45-59 15 3 4 2
60 4 1 1 2
NA 134 26 - -
  516 100* 100 100*

*Actual total exceeds 100% due to rounding.

K. Cyclist Injury (n=516)

In almost three-quarters of the accidents, the cyclist was reported as being injured or killed (73 percent). One fatality was obtained from our sampling procedure. However, eight fatalities occurred in the study area during the study period, and all were included in our sample, resulting in an overrepresentation of fatalities.

L. Seriousness of Cyclist Injury (n=382)

The accident report form's injury categories and the, proportions in each category are shown below. Only accidents involving an injury or fatality are included in calculating percentages.

  • Killed (see K, above)
2%
  • Visible signs of injury, as bleeding wound, or distorted member; or had to be carried from scene
31%
  • Other visible injury, as bruises, abrasions, swelling, limping, etc.
45%
  • No visible injury but complaints of pain or momentary unconsciousness
22%

M. Other Persons Injured

In only 12 cases - 2 percent - was a person other than the cyclist injured. In 10 of these cases, it was another cyclist. In one case, it was a cyclist passenger, and, in another, a driver passenger. In three other cases, the identity of the person injured was not shown. These results were similar to the findings in the Cross study.

N. Severity of Other Person's Injuries (n=12)

The severity of other person's injuries was reported as follows:

  • Killed
0%
  • Visible signs of injury, as bleeding wound, or distorted member; or had to be carried from scene.
33%
  • Other visible injury, as bruises, abrasions, swelling, limping, etc.
42%
  • No visible injury, but complaints of pain or momentary unconsciousness
25%

O. Accident's Roadway Location (n=491)

The majority of the accidents occurred at intersections (52 percent). After intersections, mid-block locations accounted for the largest portion (30 percent), followed by driveways (16 percent). Alleys, rotaries, offramps, parking lots and other locations accounted for only a negligible proportion of accidents (2 percents).

The Cross study (p. 128) reported a lower proportion of accidents at intersections (44 percent) and a slightly higher proportion of mid-block locations (34 percent). This is probably due to the greater number of rural roads included in the study.

P. Traffic Controls Present (n=241) (12).

For the most part, traffic controls were only indicated on reports where accidents occurred at intersections. These were distributed as shown in Table 2.

Table 2

Traffic Controls Present
(n=241)

Type of Control Percentage Cross Percentage
Stop Sign

27%

59%

Signal Light

35%

30%

No Controls

36%

11%

Other

2%

-
Totals

100%

100%

 

Traffic control information was not available for six percent of the intersections.

The Cross study thus showed a much higher percentage of the Intersections with stop signs, and a much lower percentage with no controls. The proportion with signal lights was approximately the same. it is likely that the differences are due in part to a higher proportion of uncontrolled intersections in the MAPC region. However, in the absence or additional information on this subject, =he extent to which other factors account for the-difference (e.g. failure of cyclists or motorists to yield at these intersections) is unknown.

Q. Situation For Motorist (Prior to Accident) (n=476)

Motorists proceeding 'straight accounted for the highest proportion of accidents (48 percent). Fight turns (16 percent) and left turns (15 percent) were the next most likely maneuver prior to the accident. Parked cars (6 percent) accounted for a significant number of accidents. Figure 6 displays these results.

 


Figure 6: Situation for Motorist

fig6.gif (12223 bytes)


R. Situation For Cyclist (Prior to Accident) (n=208)

Cyclists proceeding straight accounted for 63 percent of the accidents for which this information was available, followed by 13 percent making left turns. Right turns, passing and other movements accounted for the remainder (24 percent). Unfortunately, the situation for the cyclist was only reported on 40 percent of the accident reports, making it difficult to assess the accuracy of these statistics for the overall sample. Figure 7 displays these results.


Figure 7: Situation for Cyclist

fig7.gif (10301 bytes)


R. Situation For Cyclist (Prior to Accident) (n=208)

Cyclists proceeding straight accounted for 63 percent of the accidents for which this information was available, followed by 13 percent making left turns. Right turns, passing and other movements accounted for the remainder (24 percent). Unfortunately, the situation for the cyclist was only reported on 40 percent of the accident reports, making it difficult to assess the accuracy of these statistics for the overall sample. Figure 7 displays these results.

S. Cyclist Violations

Three types of cyclist violations were reported: wrong-way riding, riding through a red light and running a stop sign.

1) Wrong-way cyclists were reported in 24 percent of the accidents (n=442). Cross reported that 19 percent of the non-fatal sample were traveling against the flow of traffic. These proportions must be considered in light of the fact that most cyclists observe directional rules.

2) Cyclists entering an intersection on a red light were involved in 6 percent of the accidents (N=465). The Cross study noted no accidents in this situation. However, Cross' standards were somewhat higher in assigning an accident to this type (i.e. that the cyclist entered after the light turned red).

3) Cyclists entering an intersection without observing a stop sign accounted for only 2 percent of the accidents (n=477). On the other hand, 8 percent of the accidents in the Cross non-fatal sample were considered to have violated a stop sign. The difference here may be due to the much higher percentage of signed intersections included in the Cross study, and the greater difficulty of our coder, in the absence of an interview, in determing whether the stop sign was obeyed.

T. Motorist Violations (Red Light: n=482; Stop Sign: n-470)

In fewer than 2 percent of the cases did the motorist run a red light or a stop sign. This is consistent with the Cross study (p. 160).

U. Accident Distribution by City or Town (n=514)

Figure 8 shows the distribution of accidents. Statistical tests show this distribution to be similar to that of all bicycle-motor vehicles accidents reported during the study years. (Pearson's Chi Square, 34 df, p 4 .05).


Figure 8: Distribution of Accidents by Community

Larger scan of this image -- city and town names legible (133 kb).

fig8.gif (31050 bytes)


Top: John S. Allen's Home Page
Up: Table of Contents
Previous: Introduction
Next: Classifications
No copyright, government document
Last modified January 29, 2001