200611091452Crashes in Taiwan: Causes, Patterns, and Relationships

Crashes in Taiwan: Causes, Patterns, and Relationships



By



Kuang Yang Kou, Associate Professor
Central Police University
56 Shu Jen Road, Ta Kang Chun Kuei San, Taoyuan Taiwan, ROC
Phone: 1-866 3 328 2321 ext 4644
Fax: 1 866 3 396 3022
E-mail:peterkou@sun4.cpu.edu.tw



Seth Adu Asante
Central Transportation Planning Staff
State Transportation Building
10 Park Plaza, Suite 2150
Boston, MA 02116
Phone: 617 973 7098
Fax: 617 973 8855
E-mail:seth.asante@ctps.org



Submission date: June 2005








Abstract: 260
Body of paper: 3880
Acknowledgment: 20
References: 100
3 Tables: 750
Overall 5010















TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 2


ABSTRACT

This paper provides information about traffic safety in Taiwan and programs implemented to improve traffic safety. A year of reportable traffic crash data covering different classes of roads
in different geographical areas in Taiwan under various weather, environmental, and roadway conditions were examined. The elements that were investigated in this study were fatalities and injuries, distribution of the main causes and contributing factors as well as the association between them. Further analysis involved modeling the effect of the contributing factors on injury rates.
The analysis indicated that speeding, use of alcohol or drugs, and defective traffic control devices or safety devices produces more injury related crashes than normal. Poor sight distance
and environmental conditions such, as bad weather and poor light conditions were found associated with higher injury rates than normal as is night time and Saturdays. It was also determined that the main cause and contributing factors that produce significantly higher injury
rates are not necessary those that occur most frequently. Over 28% of the crashes were caused by drivers' disregard of traffic control devices or regulations making it the most frequent cause of crashes. Failure to yield right-of-way and inattentiveness each individually, accounted for 16%
of the crashes. Together, four factors contributed nearly 75% of the total crashes.
Vehicle-vehicle crashes were the predominant crash type; they were attributed to 81% of the crashes. Vehicular-pedestrian crashes and vehicular-highway object crashes were
respectively attributed to 8.6% and 9.5% of the crashes. The paper provides insight into the areas where more emphasis is needed for effective safety programs.








































TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 3


INTRODUCTION

This paper provides information about traffic crashes in Taiwan in an attempt to improve safety. The objectives of many safety programs are to reduce severity of crashes and its occurrence. Understanding these issues would lead to the development of effective corrective measures and improved highway safety programs. This information includes distribution of crashes and injury rates among the main cause and contributing factors. The paper discusses actions that lead to the occurrence of crashes and the relationships among the main causes and contributing factors. The study is based on analysis of crash records in Taiwan.

COLLISION DATA

Taiwan has a tropical climate with rainy season during southwest monsoon (June-August). On average, Taiwan receives 76 in of rainfall, 102 rainy days and no snow annually. The average temperature varies between 60o F in December/January and 85o F in July/August, and because it is a small
island, there is not much variation in temperature across the country.
Crash data were obtained from a national crash database maintained by the Department
of Transportation's Safety Division. The data was extracted form reportable crashes that occurred within the period of January 2000 to December 2000. The crash data covers the different classes
of roads in different geographical areas in Taiwan under various weather, environmental, and roadway conditions. The elements in this study were the number of persons killed or injured, dates and time of the crash, and environmental and roadway conditions at the time of the crash.
The crash data was reclassified for the purpose of analysis to remove redundancies. Summary statistics that characterize both main cause and contributing factors were calculated. Further analysis involved modeling the effect of the contributing factors on injury rates. Also analyzed were the effects of time of day and day of week, weather, light conditions, pavement conditions, and sight distance on crashes. The relationship between crash type and main cause of crash were investigated. The Statistical Analysis System (SAS) was used in this study (1).

SEVERITY OF CRASHES

Approximately 6.5% of the crashes involve fatalities while 93.5% involves injuries or property damage. A detail examination the crash data indicated that of the nonfatal crashes, only 4.5% were without injury. The injury rates by main cause and contributing factors are presented in
Table 1. The overall injury rate including fatalities was 1.33 injuries per crash; shading denotes
rates higher than normal. The summary statistics show that speeding, use of alcohol or drugs, and defective traffic control devices or safety device produces more injury related crashes than
normal. Poor sight distance and environmental conditions such, as bad weather and poor light conditions were found associated with higher injury rates than normal. Furthermore, night time
(9:00 PM to 4:00 AM) and Saturdays had higher injury rates than normal. The following section gives a brief discussion of why these found to produce more injuries than normal.
Analysis of variance (ANOVA) was used to examine the data for significant differences
(α =0.05) in injury rates among the contributing factors. The results of the ANOVA using the general linear model (GLM) procedure of the SAS statistical package are presented in Table 2. The analysis indicates significant differences in injury rates among the main cause and contributing factors that are described below. Because F-value was significant at α = 0.05,






TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 4


multiple comparison of means was carried out using Tukey method to identify which groupings
of main cause and contributing factors are significantly different (2). The multiple comparisons indicated that each significant factor has one or more levels that have significantly higher injury rates than normal confirming the pattern shown in Table 1.

Speeding
Although the exact relationship is not known, it is a fact that crash rates and injury severity rates
increase directly with increasing speed of travel. Speeding defined as exceeding the posted speed limit or driving too fast for conditions is one of the most prevalent factors contributing to traffic crashes and a major highway safety issue. Speeding reduces a driver's ability in several ways, for example, the ability to steer safely around curves or objects in the roadway. In addition, speeding extends the distance necessary to stop a vehicle, and increases the distance a vehicle travels
while the driver reacts to a dangerous situation. Faster driving speed also provide less time to process information and to act on it.

Drug and Alcohol
Drug and alcohol impair human motor skill and ability and as such are associated with more
injuries and fatalities than normal. During the past decade, there have been significant advances in
the USA and Europe in reducing fatalities and injuries caused by impaired driving such as under the influence of alcohol or drugs. Like speeding, alcohol and drugs contribute to traffic crashes
and is a major highway safety issue facing highway officials and lawmakers. Alcohol and drugs reduces a driver's ability in several ways, for example, the ability to steer safely around in curves
or objects in the roadway. Alcohol induces drivers to speeding as National Highway Traffic Safety Administration (NHTSA) data shows that alcohol involvement is prevalent for drivers involved in speeding-related crashes and that the two are clearly a deadly combination (3).

Defective Vehicle
Vehicle safety is public safety because we are all at risk when we share the road with vehicles
whose safety systems are not working properly. Vehicle defects are not limited to defective braking systems only. They include vehicular visual systems such as scored windscreen, inoperative wipers, mirrors, and headlamps. Defective vehicular warning light systems such as
turn, signal, and brake lights are also part of vehicle defects. Another category of vehicle defects are security and safety systems such as broken seatbelts, insecure seats, defective airbags, and safety systems such as leaky and insecure fuel tank and excessive smoke.
Vehicle defects impair the driver's view of the road; they also impair the visibility of the vehicle to other road users or prevent the driver from indicating his or her intentions
(conspicuity) e.g. inoperative lights and inoperative horn. Steering and tire/brake defects and insecure driver's seat impair the driver's control of direction and/or speed of the vehicle while missing or broken seat belts, insecure seats, weakened body structure impair the built-in occupant protection afforded by the vehicle in the event of a crash. Insecure fuel tank and inoperative emergency exits increase the risk of further injury after a crash such as fire.

Sight Distance
Substandard highway designs that lead to inconsistent transitions to and from horizontal curves
and inadequate sight distance increase the risk of crashes and severity. Poor sight distance simply does not give drivers adequate time and distance needed to react to events on the roadway. At






TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 5


roadway intersections, poor sight distance causes angle and right-angle collisions and in horizontal and vertical curves, they are often the cause of head-on collisions during passing
maneuvers and rear-end collisions at obscured traffic lights. On urban highways and streets with traffic queues, poor sight distance sometimes causes rear-end collisions when queues are hidden
by curvatures in the roadway.

Environmental Conditions
Environmental conditions such as weather and light affects driving and subsequently crashes.
Foggy, smoggy, and windy weather impairs driver's view of the road and visibility of the vehicle
to other road users. In foggy or smoggy weather, headlights, warning, and signal lights are hardly visible. Strong wind usually affects steering of vehicle and therefore impairs the driver's control
of direction and/or speed of the vehicle.
Lighting conditions affects injury rates as nighttime (with/without lights) have significantly higher rates than daylight. Three primary reasons behind the high nighttime injury
rates are reduced visibility and sight distance of drivers' especially older drivers who have vision problems. Another reason is the entertainment activities involving alcohol and drugs that takes
place at night; many drunken driving arrests occur in the night. Fatigue and stress after driving for a long time usually take effect during nighttime driving.
The time of day and day of week also affects injury rates as nighttime crashes have significant injury rates than daytime crashes. Saturdays also have higher injury rates due to the entertainment and recreational activities that take place at weekends that might involve alcohol.


DISTRIBUTION OF CRASHES

This section describes the distributions of crashes among the main cause and contributing factors
for designing effective crash reduction programs. The main cause and contributing factors that produce significantly higher injury rates are not necessary those that occur most frequently.

Disregard of Traffic Control Devices or Regulations
Over 28% of the crashes were caused by drivers' disregard of traffic control devices or
regulations making it the most frequent cause of crashes. Disregard of traffic control devices or regulation causes all kinds of crashes, its injury rates are closer to the average (1.31 versus 1.33). Even though the motoring public usually thinks about traffic control in terms of traffic signals, because the signal is the most visible form of control and the one which most obviously affects travel, traffic control is exercised through a combination of road markings, traffic signs, and
signals. Even the absence of these devices is a conscious decision, with the public guided by the
"rules of the road" in such cases (4).
The purpose of traffic control devices and warrants for their use is to help insure highway safety by providing for the orderly and predictable movement of all traffic, motorized and non- motorized, throughout the highway system, and to provide such guidance and warnings as are needed to insure the safe and informed operation of individual elements of the traffic stream. For
a traffic-control device to be effective, it must fulfill a need, command attention, convey a clear and simple meaning, command respect of road users, and give adequate time for proper
response. To insure that these requirements are met, consideration should be given design, placement, operation, maintenance, and uniformity.







TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 6


Failure to Yield Right-Of-Way
Failure to yield right-of-way accounted for 16% of the crashes. Failure to yield the right of way
results in all types of crashes, angle, sideswipes, and rear-ends; its injury rate is close to the average (1.30 versus 1.33). Even though it was considered separately because of its high
frequency, it is part of disregard of traffic regulations. Crashes involving failure to yield right-of- way usually occurs at locations where right-of-way is assigned by priority rules or by signs such
as at intersections, at ramp-freeway junctions, and at ramp-arterial street junctions. With increasing traffic congestion in Taiwan, it is not surprising that some of these crashes are deliberate act of aggressiveness or road rage, while others are willful miscalculation of gaps in traffic by drivers. Enforcement by police and driver education has been implemented by law enforcement officials in an attempt to get drivers to obey rules of the road.

Inattentiveness
Inattentiveness accounted for nearly 16% of the crashes. Rear-end and sideswipes were the most
common forms of crashes caused by inattentiveness, however, they can lead to any kinds of crashes including runoff and head-on collisions. Its injury rate is close the average (1.30 versus
1.33). They occur when the driver is distracted from keeping his or her eyes on the roadway through conversations, talking on the phone, fatigue and sleeping, and distraction from roadside events such as incidents, crashes, and scenery. Crashes involving inattentiveness can not be reduced by forgiving designs but by public awareness programs. Recent advances in collision avoidance systems is yet to play a significant role in reducing such crashes as most average Taiwanese vehicle owners can not afford such vehicles.

Improper Maneuver
Improper maneuver represented nearly 14% of the crashes. Improper maneuvers results in all
kinds of collisions, however, its injury rates are lower than the average (1.27 versus 1.33). As with failure to yield right-of-way, it is part of disregard of traffic control devices or regulations. Improper maneuvers include making prohibited turns, exiting and turns from the wrong lane, improper lane changes, stopping in moving traffic, and stopping in areas designated as "no stopping". Improper maneuvers occur on all classes of roads but they are prevalent on urban
streets where traffic controls and regulations, and restrictions are more frequent. Enforcement by police and driver education are some of the methods used to get drivers to obey rules of the road.

Alcohol and Drugs
Driving under the influence of alcohol/drugs and speeding contributed 8% each of the total
crashes; however, its injury rates are much higher than the average (1.43 versus 1.33). Alcohol/drugs influence speeding, as it has been established that alcohol involvement is prevalent for drivers involved in speeding-related crashes. In 2003, 41% of the drivers with blood alcohol content (BAC) of 0.08 grams per deciliter or higher involved with fatal crashes
compared with only 14% for drivers with BAC of 0.00 involved with fatal crashes. Even though alcohol and speed contributed 8% each as the main cause, studies have found that alcohol/drugs and speeding are prevalent contributing factors in 31% of fatal crashes (3). Enforcement,
effective deterrent measures backed by legislation, and public education programs are some of the approaches that have been employed to stop driving under the influence of alcohol or drugs.









TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 7


Sight Distance
Poor sight distance was the main cause of nearly 7% of the total crashes. However, it was not
clear whether these were caused by environmental factors such fog and smog or sight distance deficiencies in the roadway. On the other hand, inadequate signs and malfunction traffic controls devices were each attributed to less than 1% of the crashes indicating the high safety standards
that highway and traffic officials strive to move people and goods efficiently. Forgiving designs and upgrading of substandard designs to improve sight distances are the most effective ways of preventing sight distance related crashes.

Crash Type
Vehicle-vehicle crashes were the predominant crash type; they were attributed to 81% of the
crashes. Vehicle-pedestrian crashes and vehicle-highway object crashes were respectively attributed to 8.6% and 9.5% of the crashes. Crash occurrence was evenly distributed across the days of the week, with each day accounting for about 14% of the total crashes. Time of day was significant, for example, the lowest crash occurrences were between midnight and 6:00 AM and the highest between 3:00 and 6:00 PM. These patterns are influence by exposure; low traffic volumes in the early morning hours and the high traffic volumes in the afternoon rush hours.

ASSOCIATION OF MAIN CAUSE AND CONTRIBUTING FACTORS

In many crash situations, there are other factors contributing to the occurrence of the crash in addition to the main cause. The contributing factors usually exist in the form of a defect or impairment and are not known until at the time of the crash when such functions or skill are found lacking. Association between the main cause of crashes and contributing factors are the primary focus of analysis in this section. The Chi-Square test of independence was used to
compare the frequencies of observed crashes and expected number of crashes for each cell group
(5). The results of the test statistics are summarized in Table 3.

Main Cause versus Crashes Type:
The test indicated strong association between main cause and type of the crash with a p-value of
0.001. Speeding, inattentiveness, and alcohol/drugs-related crashes were found to have higher fractions of vehicle-highway object. Crashes involving failure-to-yield, follow-too-closely, and improper maneuver were found to have higher fractions of vehicle-vehicle crashes. Crashes involving inattentiveness and disregard of traffic control or regulations were found to have
higher fractions of vehicle-pedestrian crashes.

Main Cause versus Severity of Crashes:
The test indicated strong association between main cause and severity of crash with a p-value of
0.001. Crashes involving speeding, alcohol/drugs, and inattentiveness were found to have higher fractions of fatal crashes. Failure-to-yield, follow-too-closely, and improper maneuver were
found associated with higher fractions of injury.

Main Cause versus Light Conditions:
The test indicated that there was a strong association between main cause and light conditions
with a p-value of 0.001. This test confirmed that speeding, inattentiveness, and alcohol/drugs related crashes occur more than expected during the night or dark conditions. Crashes involving






TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 8


failure-to-yield, follow-too-closely, and improper maneuver occur more than expected during daylight conditions.

Main Cause versus Sight Distance:
The test indicated that there was a strong association between main cause and sight distance with
a p-value of 0.001. This test confirmed that speeding and inattentiveness occur more than expected at locations with poor sight distance.

Main Cause versus Weather Conditions:
The test indicated that there was a strong association between main cause and weather conditions
with a p-value of 0.001. This test confirmed that crashes involving alcohol/drug and inattentiveness occur more than expected under rainy conditions.

APPROACHES TO HIGHWAY SAFETY


Improvements to highway traffic safety can be handled in several different ways and requires consideration of three elements-the driver, the vehicle, and the roadway. Policy level safety
efforts are effective in addressing speeding and alcohol/drug related crashes that cause more fatal and severe injuries. Such programs should be supported by legislation at governmental level and
be based on well founded research that such programs potentially improve safety. The listing below is a sampling of some of the more notable programs influencing traffic safety in the Taiwan:

Policy Level Safety Efforts

Vehicle-Inspection Programs
In Taiwan private automobiles of less than five years from the year of manufacture are exempt
from routine inspections while those exceeding five years but less than ten years are subject to at least one annual auto inspection. Those over ten years are subject to at least twice yearly auto inspections. However, automobiles that operate on liquefied petroleum or compressed natural
gas and commercial vehicles with less than five years are subject to at least one yearly auto inspection. Those over five years are subject to at least twice yearly auto inspection. Heavy-duty motorcycles of less than five years are exempt from routine auto inspection while those over five years and less than ten years are mandated to undergo at least one inspection every year. Heavy- duty motorcycles over ten years are mandated to undergo at least two inspections per year.
Trailers are mandated to undergo routine inspection at least one every year.

Driver Licensing Programs
In Taiwan, those who apply for automobile driver license must be at least 18 years of age and
there is no restriction to the maximum age cap. Those who apply for commercial driver's license must be at least be 20 years of age and there is no restriction to the maximum age cap. In
addition, they must also have the following experience:

Be free from any restriction of prior driving experience.
Have over three months of driving experience.







TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 9


Have over six months of driving experience if applying for general commercial sedan driver's license.
Have one year of driving experience if applying for heavy-duty cargo freighter or commercial heavy-duty freighter driver's license.
Hold a valid general heavy-duty cargo freighter driver's license with over one year of driving experience or hold a valid general sedan driver's license with over two years of driving experience, and have a driving training for advancing from driving a sedan to heavy-duty passenger vehicle at a registered driving training institution if applying for a general heavy-duty
passenger or commercial vehicle driver's license.
Hold a valid general heavy-duty vehicle driver's license with over two years of driving experience, if applying for general trailer vehicle or commercial trailer vehicle driver's license.

National Minimum Drinking Age
The Child and Adolescent Welfare Act of Taiwan mandates that children and adolescents under the age 18 may not engage in drinking, administering narcotics, or illegally administering controlled drugs or other substances that are harmful to one's health.

National Driving While Intoxicated Programs
The Roadway Traffic Administration Penal Code of Taiwan specifies punitive measures for driving while intoxicated. Drivers who are found with blood alcohol concentration exceeding certain specified criteria or found to be on narcotics, hallucination drugs, anesthetics, or similar controlled drugs are to face the following:

A penalty fine of over NT$15,000 (US$485) and up to NT$60,000 (US$1,935) and
subject to impoundment and safekeeping of ones vehicle and suspension of ones driver's license for one year.
Those who are involved in moving violations in which human injuries are reported may also face a suspension of ones driver's license for two years.
Those who are involved in moving violation in which severe injuries or deaths are
reported may also face suspension of driver's license for good, and be barred from ever retaking the driver's road test.

Driver Training and Testing Programs
Improvement of driver skills through training and testing programs and removal of drivers with
bad crash and/or violation records from the highways has been found effective to reduce crashes. Driver training and removal of "bad" drivers from the highways are complex issues involving li- censing and enforcement procedures.

Engineering Design Safety Efforts

Reducing Crash Occurrence
Reducing crash occurrence by preventing crashes from occurring is an effective means to improve highway traffic safety but very difficult to accomplish, because the causes of traffic crashes are
many and complex and on many occasions involve some form of driver error. Proper highway
and traffic designs and proper signing and marking reduce driver confusion and the risk of driver error. Policy level safety efforts addressing speeding and alcohol/drug related crashes that cause





TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 10


more fatal and severe injuries are also crash reduction efforts. Improving driver skills through training and testing programs have also been found to reduce crashes in Taiwan.

Reducing Crash Severity
Reducing the severity of crashes is another approach to improving highway traffic safety. This approach involves highway designs that recognize those locations where crashes are most likely
to occur and design them in such a way as to give drivers time and space to "recover" from errors and to minimize the severity of the crash when it occurs. Proper use of highway facilities such as guardrail, median barriers, impact attenuators, and breakaway signpost, and light standards, and
gore areas on freeway off-ramps kept free of objects, can reduce the damage when a vehicle leaves the travel lane or give drivers opportunity to recover from their error.

Improving Crash Survivability
Improving crashes survivability is another approach to improving highway traffic safety. Crashes survivability primarily involves vehicle designs that absorb most of the impact of crashes without transferring it to the occupants. Such design features as energy-absorbing
bumper systems, padded dashboards, seat belts, air bags, and similar measures are all attempts to improve crash survivability.

CONCLUSIONS

This paper provides information on highway traffic crashes in Taiwan. Both the main cause and contributing factors of crashes were analyzed and found to affect injuries differently; they were
also found to be distributed differently. Crashes involving speeding and alcohol/drug were found
to be significantly associated with more fatalities and injuries. Disregard of traffic control
devices and regulations were the most frequent cause of crashes in Taiwan. The paper provides insight into the association between main cause and contributing factors. Certain contributing factors were found to be closely associated with the main causes. In spite of the cultural,
climatic, and lifestyle differences, the paper suggests that some of the highway traffic safety programs that have been successfully implemented in the USA, Europe, Japan, and Australia should be expected to work in Taiwan too.

ACKNOWLEDGEMENTS

The author would like to thank the Department of Transportation's Safety Division for providing the data for this paper.




















TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 11


REFERENCES

1. SAS/STAT User's Guide, Version 6, Volume I and II, SAS Institute Inc., Cary, North
Carolina, USA, 1990, pp.

2. Montgomery, D. C., Design and Analysis of Experiments, Third Edition, John Wiley and
Sons, Inc., 1991

3. Traffic Safety Facts, 2003 Data, National Highway Traffic Safety Administration, National
Center for Statistic and Analysis, Washington, D.C., 2003.

4. McShane, W. R., and Roess, R. P., Traffic Engineering, Prentice Hall, Englewood Cliffs, New Jersey, USA, 1990.

5. Schlotzhauer, S. D., and Littell, R. C., SAS System for Elementary Statistical Analysis, SAS Institute Inc., Cary, North Carolina, USA, 1987.















































TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 12




LIST OF TABLES

TABLE 1: Injury Rates by Contributing Factors and Crash Types

TABLE 2 Tests for Significant Differences among the Groupings of the Contributing Factors

TABLE 3 Association of Main Cause of Crash with Weather and Highway Conditions


























































TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 13


TABLE 1 Injury Rates by Contributing Factors and Crash Types


Main Cause Injuries per Crash Standard
Deviation Number of Cases Distribution of Crashes
Speeding 1.47 0.98 4202 8.5
Failed to yield Right-of_Way 1.30 0.61 7701 15.5
Disregard of traffic control or regulation 1.31 0.65 14367 28.9
Inattentive 1.30 0.63 7793 15.7
Alcohol and drugs 1.43 0.82 4078 8.2
Follow too closely 1.34 0.72 3570 7.2
Out of control vehicle 1.32 0.80 267 0.6
Improper maneuver 1.27 0.66 6795 13.7
No violation 1.20 0.48 321 0.6
Defective traffic control or safety device 1.40 1.33 346 0.6
Driver or passenger moving violations 1.35 0.53 305 0.6
Weather
Rainy 1.32 0.73 6533 13.3
Cloudy/Foggy/Smoggy/Windy 1.38 0.80 6100 12.1
Sunny 1.32 0.66 37112 74.5
Light Conditions
Daylight 1.29 0.67 28152 56.6
Dawn/Dusk 1.33 0.72 2826 6.7
Night with light 1.37 0.69 16274 32.7
Night without light 1.45 1.02 2479 5.0
Sight Distance
Poor 1.43 1.04 3285 6.6
Adequate 1.32 0.68 46460 93.4
Crash Type
Vehicle-Pedestrian 1.31 0.55 4298 8.5
Vehicle-Vehicle 1.33 0.69 40691 81.8
Vehicle-Roadway object 1.30 0.95 4719 9.5
Vehicle-Rail 1.46 1.50 37 0.1
Day of Week
Sunday 1.30 0.66 7036 14.1
Monday 1.32 0.75 6987 14.1
Tuesday 1.29 0.63 7067 14.2
Wednesday 1.30 0.62 7194 14.5
Thursday 1.32 0.67 7243 14.5
Friday 1.35 0.73 7537 15.2
Saturday 1.43 0.87 6681 13.4
Time of Day
000 - 300 (Night) 1.42 0.80 2873 5.8
300 - 600 (Dawn) 1.37 0.82 2790 5.6
600 - 900 (Morning) 1.27 0.63 7951 15.9
900 - 1200 (Morning) 1.29 0.63 6875 13.8
1200 - 1500 (Afternoon) 1.32 0.68 7201 14.4
1500 - 1800 (Afternoon) 1.32 0.62 8947 17.9
1800 - 2100 (Evening) 1.38 0.70 7231 14.5
2100 - 2400 (Night) 1.39 0.73 6404 12.9





TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 14


TABLE 2 Tests for Significant Differences among the Groupings of the Contributing
Factors



Degree of Sum Mean
Source Freedom Squares Square F Value Prob.


Model 49 400.63 8.17 20.31 <0.0001
Error 49695 20003.74 0.40



Contributing Degrees of
Factors (Source) Freedom Type I SS MS F Value Prob.

Main Cause 10 161.18 16.11 40.04 0.0001
Weather 2 9.26 4.63 11.50 0.0001
Light 3 64.49 21.50 53.41 0.0001
Sight Distance 1 16.49 16.49 40.98 0.0001
Crash Type 3 53.21 17.74 44.07 0.0001
Pavement Surface 1 1.04 0.04 2.60 0.1068
Time of Day 23 31.57 1.37 3.41 0.0001
Day of Week 6 63.35 10.56 26.23 0.0001





Contributing Degrees of
Factors (Source) Freedom Type III SS MS F Value Prob.

Main Cause 10 144.13 14.41 35.80 0.0001
Weather 2 4.26 2.13 5.30 0.0050
Light 3 8.43 2.81 6.98 0.0001
Sight Distance 1 18.89 18.89 46.92 0.0001
Crash Type 3 57.59 19.19 47.70 0.0001
Pavement Surface 1 0.97 0.97 2.42 0.1200
Time of Day 23 27.08 1.18 2.93 0.0001
Day of Week 6 63.35 10.56 26.23 0.0001




















TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.

K. Y. Kou and S.A. Asante 15


TABLE 3 Association of Main Cause of Crash with Weather and Highway Conditions

Contributing Degrees of
Factors Statistic Freedom Value Prob.

Main Cause vs. Crash Type Chi-Square 30 12057.50.001
Main Cause vs. Light Conditions Chi-Square 30 3427.7 0.001
Main Cause vs. Sight Distance Chi-Square 10 605.2 0.001
Main Cause vs. Weather Chi-Square 20 264.4 0.001


























































TRB 2006 Annual Meeting CD-ROM Paper revised from original submittal.
沒有上一則|日誌首頁|沒有下一則
回應
    沒有新回應!
關鍵字





Powered by Xuite