_{1}

Due to the increasing trend of population growth and urbanization, pedestrians form one of the largest single road user groups. However, they are the most neglected group among all road users. Pedestrian safety is now a growing concern in the USA. Identifying the factors associated with fatal pedestrian crashes plays a key role in developing efficient and effective strategies to enhance pedestrian safety. This study addresses safety issues by identifying contributory factors associated with fatal pedestrian crashes in Kansas and the USA. For Kansas, the study uses KARS (Kansas Accident Reporting System) database while for the USA FARS (Fatality Analysis Reporting System) database ha s been used. Different variables considered in this study are human variables (age, and gender), environmental variables (atmospheric condition and light condition), time (time of day, day of week, and crash month), location (intersection vs . mid-block), and roadway variables (speed limit). Different factors that are found to have an association with fatal pedestrian crashes are male pedestrians, older pedestrians, weekend, off peak hours, winter months, dark hours, non-intersection, clear atmospheric conditions, higher speed limit. The findings from Kansas have been compared with that from the USA. This study helps to implement potential countermeasures by identifying the factors that have an association with fatal pedestrian crashes.

Due to the increasing trend of population growth and urbanization, pedestrians form one of the largest single road user groups. However, they are the most neglected group among all road users. The Federal Highway Administration (FWHA) conducted a survey in 2005, which is called Traveler Opinion and Perception (TOP) survey. According to the survey [

Pedestrian fatalities are 36 times higher than car occupant fatalities per km traveled [

The objectives of the study are:

• To compare the trend of pedestrian fatalities in Kansas to that of the USA,

• To identify the causative factors of fatal pedestrian crashes in Kansas,

• To compare the findings with USA,

• To determine if any relationship exists between different variables considered with the help of Chi-Squared test.

The remainder of the paper is organized as follows: Section 2 describes the relevant studies on pedestrian crashes, Section 3 describes the data and the methodology used in this study, Section 4 analyzes the results and the statistical analysis used in this study, Section 5 concludes the study.

A study done in Florida by Spainhour et al. [

The largest risk factors for fatal injuries of pedestrians hit by a vehicle are identified by Siddiqui et al. [

Another study conducted by Carter et al. [

There was a research at Monash University Accident Research center carried out by Corben et al. [

Despite significant research efforts made in past studies, to the best of author’s knowledge, there was no study done which compared the factors associated with fatal pedestrian crashes between Kansas and USA. The factors that have been found to have strong association with fatal pedestrian crashes will be helpful to implement countermeasures which will ultimately reduce the number of fatal pedestrian crashes.

The study uses KARS (Kansas Accident Reporting System) database [

To analyze the data, chi-squared statistics have been used. Degree of freedom has been referred to as the number of independent variables which make up the statistics [^{2} are n − 1 [^{2}-test of goodness of fit. Therefore, if s is the number of populations parameter estimated from the sample observation of n number, then the degree of freedom for χ^{2 }would be (n-s-1) [^{2}-test. So, in a (r * s) contingency table, the required number of degrees of freedom can be expressed as follows [

Degreeoffreedom = r s − ( r + s − 1 ) = ( r − 1 ) ( s − 1 ) (1)

Chi-squared test has been formulated by Karl Pearson in 1900 and it is a very powerful test for testing significance of the discrepancy between theory and experiment [_{0} is true, then the observed and expected values are likely to be close to each other. To define the test statistic, let k be the number of outcomes and let O_{i} and E_{i} be the observed and expected number of trials, respectively, that result in outcomes i. So, the chi-squared statistic is [

χ 2 = ∑ i = 1 k ( O i − E i ) 2 E i (2)

It is called chi-square distribution with (k − 1) degrees of freedom. The larger the value of χ^{2} is, the stronger the evidence against H_{0}. It is necessary to know the null distribution of this test statistic in order to determine the p-value for the test.

For trend analysis, ten years’ crash data from 1999 to 2008 have been considered both for USA and Kansas. However, when different factors are considered which have association with fatal pedestrian crashes, data for 2004-2008 have been considered. For statistical analysis also, data for USA from 2004-2008 have been considered. For Kansas, the number of observations is low and that is why, statistical test was not considered.

or in USA have been improved. On the contrary, a study conducted by the National Household Travel Survey (NHTS) in 2001 explains the fact that in the United States, travel behavior is accompanied mainly by auto and in recent decades, auto ridership has been increased tremendously (from 66.9% in 1960 to 87.9% in 2000) [

Therefore, as level of exposure for pedestrians has been decreased drastically, number of pedestrian fatalities has been less too.

Gender and age of persons who involved in fatal pedestrian crashes are analyzed in this section.

Mode of Transportation | Census Year | ||||
---|---|---|---|---|---|

1960 | 1970 | 1980 | 1990 | 2000 | |

Auto | 66.9 | 77.7 | 84.1 | 86.5 | 87.9 |

Public Transit | 12.6 | 8.9 | 6.4 | 5.3 | 4.7 |

Walk | 10.3 | 7.4 | 5.6 | 3.9 | 2.9 |

Bicycle | Na | Na | 0.5 | 0.4 | 0.4 |

All | 100 | 100 | 100 | 100 | 100 |

**The modal share doesn’t add to 100 because there is a small percentage of other modes (work at home etc.) which is not included here.

If age group is considered, it is found that age groups of 45 - 54 and 65+ years are the most vulnerable groups. For Kansas, the maximum number of pedestrian fatalities was found for the 65+ years older pedestrians in 2004 (

Age groups of 25 - 34 and 55 - 64 comprise of the lesser number of pedestrian fatalities too.

For time of occurrence, three variables are considered: days of week, time of day and crash month. These three variables are analyzed in this section.

The reason might be the increased level of interaction between motorized vehicles and pedestrians during these two days. Also, alcohol could be another reason for which there is a larger number of fatal pedestrian crashes during weekends. One exception to this is that in 2005 the maximum number of fatal crashes occurs on Tuesday in Kansas. The finding is compared with USA and the same information is found. This is illustrated by

If the time of the day is considered, then it is found that off-peak hours (6 pm to 6 am) are the critical time for fatal pedestrian crashes to occur (

The maximum number of fatal crashes occurs between midnight and 6 am in the morning (24%) followed by between 9 pm and 12 pm (23%). The least number of pedestrian fatalities occur between 6 am and 9 am in the morning (12%). For USA this time interval accounts only for 9% of the total fatalities which is

the lowest number of pedestrian fatalities among all time intervals (

Fatal pedestrian crashes occur in all months, but some months or some particular seasons are critical for crashes to take place.

When atmospheric condition is considered, it is found that the majority of fatal crashes occur on a clear day with no adverse condition (

When light condition is considered, it is found that most of the fatal pedestrian crashes occur in dark, with or without light (

If location of fatal pedestrian crashes is considered, it is found that most of the fatal pedestrian crashes occur in non-intersection or in other words, in mid-block. The same trend is found for both Kansas and the USA. 79% of the crashes occur in Kansas is in non-intersection and for the USA the percentage is about 74% (

When speed limit is considered, different trends are found for Kansas and the USA (

USA speed limit between 35 mph and 40 mph represent the maximum number of crashes (27%). For Kansas, speed limits between 30 mph and 40 mph account for 52% of total crashes (26% crashes for 30 mph and 26% for 35 mph or 40 mph), whereas the same speed limit represent 20% and 27% of total crashes. Speed limit between 45 mph and 55 mph represent a very small percentage of total crashes (7%) in Kansas. This can be explained by the fact that Kansas has lot of rural roads, where the speed limit is high and in rural roads, laws are not strictly enforced, all of which might lead to a larger number of fatal pedestrian crashes.

Statistical analysis is performed to determine if there is any dependency or relationship between different variables considered in this study. The test of independence analyzes the independency of two variables using Chi-square distribution. The test is also called contingency table test. The dependency between age group and gender, light condition, and days of week were analyzed and is shown in

Dependency between age groups and days of week is tested. From the cross tabulation of age groups and days of week, it is found that most of the fatal crashes occur on Friday and Saturday for both young and middle-aged pedestrian, but older pedestrians are more vulnerable during weekdays. The reason might be those older pedestrians do not go out during weekend, which contributes to a smaller number of crashes during the weekend than weekdays. As younger pedestrians are found to be more vulnerable during weekends, it could be assumed that they were under the influence of alcohol. This variable (alcohol) was not analyzed in this study. The calculated chi-square value for this group (age groups and days of week) is 246.51. Comparing this value with the tabulated chi-square value, which is 21.03 for 12 degrees of freedom, leads to reject the null hypothesis and dependency between age groups and days of week is established.

Contingency table analysis between gender and speed limit (

Description | Young | Middle Aged | Older | Total | |||
---|---|---|---|---|---|---|---|

Number | Percent | Number | Percent | Number | Percent | ||

Gender | |||||||

Female | 1758 | 33% | 4109 | 29% | 1906 | 59% | 7773 |

Male | 3652 | 67% | 10441 | 71% | 2789 | 41% | 16,862 |

Total | 5390 | 100% | 14,550 | 100% | 4695 | 100% | 24,635 |

Chi-squared value = 254.56, df = 2, Significant | |||||||

Light Condition | |||||||

Daylight | 1301 | 31% | 3274 | 22% | 2196 | 47% | 6771 |

Dark | 1527 | 36% | 5485 | 37% | 921 | 20% | 7933 |

Dark but lighted | 1184 | 28% | 5602 | 38% | 1326 | 28% | 8112 |

Dawn and Dusk | 191 | 5% | 505 | 3% | 227 | 5% | 923 |

Total | 4203 | 100% | 14,866 | 100% | 4670 | 100% | 23,739 |

Chi-squared value = 1284, df = 6, Significant | |||||||

Day of Week | |||||||

Sunday | 868 | 16% | 2318 | 16% | 432 | 9% | 3618 |

Monday | 633 | 12% | 1806 | 12% | 699 | 15% | 3138 |

Tuesday | 599 | 11% | 1819 | 12% | 739 | 16% | 3157 |

Wednesday | 652 | 12% | 1968 | 13% | 750 | 16% | 3370 |

Thursday | 655 | 13% | 1907 | 13% | 665 | 14% | 3227 |

Friday | 837 | 16% | 2359 | 16% | 785 | 17% | 3981 |

Saturday | 1039 | 20% | 2780 | 18% | 613 | 13% | 4432 |

Total | 5283 | 100% | 14,957 | 100% | 4683 | 100% | 24,923 |

Chi-squared value = 246.51, df = 12, Significant |

Description | Male | Female | Total | ||
---|---|---|---|---|---|

Number | Percentage | Number | Percentage | ||

Speed Limit | |||||

30 mph or less | 2889 | 18% | 1845 | 27% | 4735 |

35 or 40 mph | 4395 | 28% | 2124 | 31% | 6519 |

45 or 50 mph | 3404 | 22% | 1317 | 19% | 4721 |

55 mph | 2497 | 16% | 811 | 12% | 3308 |

60 mph or higher | 2587 | 16% | 730 | 11% | 3317 |

Total | 15,772 | 100% | 6827 | 100% | 22,599 |

Chi-squared value = 358.32, df = 4, Significant |

chi-square value of 9.488. As the calculated value is much higher than the tabulated value, it can be said that there is a strong relationship between gender and speed limit. From the contingency table analysis, it can be concluded that male pedestrians tend to have strong association with higher speed limit than female pedestrians. Also, it was found that the lower the speed limit is, the stronger the association is with female pedestrians.

Based on the findings, some of the countermeasures have been suggested which can minimize or lessen the number of pedestrian crashes.

The countermeasures have been categorized into three following sections [

• Geometric countermeasures: Some of the geometric countermeasures that can be considered are converting unsignalized intersections into roundabout, installing pedestrian underpass/overpass, installing raised median at unsignalized intersections, installing Refuge Island, installing raised pedestrian crossing, providing paved shoulders etc.

• Signalized countermeasures: Signalized countermeasures can include installing exclusive pedestrian phasing, improving signal timing, replacing existing WALK/DON’T WALK signals with pedestrian countdown signal heads, removing unwarranted signals (one-way street), increasing pedestrian crossing time, implementing leading pedestrian interval etc.

• Signs, marking and operational countermeasures: Adding intersection lighting, adding segment lighting, improving pavement friction, prohibiting right turn on red, prohibiting left turns, advanced stop/yield sign, converting parallel lane into high visibility crosswalk, restricting parking near intersections etc. could be some of the measures that can be taken as operational countermeasures.

An attempt has been made in this study to identify the contributing factors of fatal pedestrian crashes both in Kansas and the USA and at the same time, comparison on the basis of contributing factors has been made between Kansas and the USA. Initially, the trends of fatal pedestrian crashes both in Kansas and the USA are identified. Conclusion is made by remarking that the percentage of pedestrian fatalities to that of total fatalities both for Kansas and USA does not reveal that the overall safety situation has been improved. It is found in the study that exposure level of walking has been decreased which might contribute to the lower number of fatal pedestrian crashes. The findings of the study can be summarized as follows:

• When involved persons in fatal pedestrian crashes are analyzed, it is found that male pedestrians contribute to the significantly larger number of fatal pedestrian crashes than female pedestrians both for Kansas and USA.

• Analysis of age groups reveals the fact that pedestrians aged between 45 and 54, and 65+ years old are the most vulnerable groups among all age groups both for Kansas and USA. This finding is well consistent with a study done by Siddiqui et al. [

• When days of week of fatal pedestrian crashes are considered, it is found that Friday and Saturday have the largest number of fatal pedestrian crashes than any other days of the week.

• When time of occurrence is considered, it is found that fatal pedestrian crashes occur mostly during off peak hours of the day and especially, after 6 pm.

• Among different months analyzed, it is observed that November and December are the prime months for fatal pedestrian crashes to occur. For Kansas, March and August are also critical for fatal pedestrian crashes.

• When location is concerned, it is examined that most of the fatal pedestrian crashes occur in non-intersection, which is mid-block. Intersections have a very few pedestrian crashes both for Kansas and USA. The same result has also been obtained by a study done by Siddiqui et al. [

• Among different atmospheric conditions observed, it was found that clear atmospheric conditions lead to the larger number of fatal pedestrian crashes compared to adverse atmospheric condition. When light condition is studied, it is found that majority of fatal crashes occur in dark conditions, even if lighted. This result resembles the study done by Carter et al. [

• Speed limits have the different results for Kansas and the USA. Speed limits higher than 60 mph result in a larger number of fatal pedestrian crashes in Kansas (similar result has been obtained by a study done by Corben et al. [

The identified factors related to fatal pedestrian crashes are helpful to implement potential countermeasures which will ultimately reduce the number of fatal pedestrian crashes.

The author would like to acknowledge the Kansas Department of Transportation for providing necessary data for carrying out this research.

The author declares no conflicts of interest regarding the publication of this paper.

Roy, U. (2019) Comparative Analysis of Fatal Pedestrian Crashes between Kansas and USA. Journal of Transportation Technologies, 9, 381-396. https://doi.org/10.4236/jtts.2019.93024