2
Quantitative Analyses: Data and Methods

There are two basic categories of safety data—exposure data and outcome data (BTS 1999). Exposure data, which measure the susceptibility of an individual or class of individuals to the undesired outcome, include number of trips taken, distance traveled (e.g., vehicle-miles or passenger-miles), and number of travelers; outcome data, which measure untoward or undesirable events, such as crashes, include numbers of accidents, fatalities, and injuries. The committee selected exposure and outcome datasets that could be used to shed light on the issues of concern. Because of the relatively infrequent occurrence of fatalities and severe injuries among children on trips during normal school travel hours, it was necessary to combine multiple years of statistics to obtain reasonable sample sizes. In addition, some otherwise promising datasets were not usable because they lacked the specificity needed to identify incidences or trips related to students going to and from school, or because they could not be paired with the corresponding exposure or outcome data. For example, as noted by Stutts and Hunter (1999, 505):

Traditionally, the U.S. Department of Transportation has relied on state motor vehicle crash data, based on reports completed by police and other law enforcement officers, as their primary source of information on events causing injury to pedestrians and bicyclists. While these data provide considerable information to help guide safety program and countermeasure development, they have often been referred to as ‘the tip of the iceberg’ because they are limited almost entirely to motor vehicle-related events that occur on public roadways. Specifically, they exclude (1) many bicycle-motor vehicle and pedestrian-motor vehicle crashes that occur in non-roadway locations such as parking lots, driveways, and sidewalks; and (2) bicyclist and pedestrian falls or other non-collision events that do not involve a motor vehicle, regardless of whether they occur on a roadway or in a non-roadway location. Even using emergency-room data will not fill in the gaps because many of the injuries may not result in visits to the emergency room, and if they do, the forms that are filled out will not include information on the purpose of the transportation (e.g., pleasure, to/from work, to/from school, etc.).

The various datasets that were examined for possible use in the committee’s analyses are briefly described in this chapter. A detailed description of the three datasets selected, the analyses conducted with each, and the limitations of each for the purposes of this study is then provided. Finally, conclusions are presented.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment 2 Quantitative Analyses: Data and Methods There are two basic categories of safety data—exposure data and outcome data (BTS 1999). Exposure data, which measure the susceptibility of an individual or class of individuals to the undesired outcome, include number of trips taken, distance traveled (e.g., vehicle-miles or passenger-miles), and number of travelers; outcome data, which measure untoward or undesirable events, such as crashes, include numbers of accidents, fatalities, and injuries. The committee selected exposure and outcome datasets that could be used to shed light on the issues of concern. Because of the relatively infrequent occurrence of fatalities and severe injuries among children on trips during normal school travel hours, it was necessary to combine multiple years of statistics to obtain reasonable sample sizes. In addition, some otherwise promising datasets were not usable because they lacked the specificity needed to identify incidences or trips related to students going to and from school, or because they could not be paired with the corresponding exposure or outcome data. For example, as noted by Stutts and Hunter (1999, 505): Traditionally, the U.S. Department of Transportation has relied on state motor vehicle crash data, based on reports completed by police and other law enforcement officers, as their primary source of information on events causing injury to pedestrians and bicyclists. While these data provide considerable information to help guide safety program and countermeasure development, they have often been referred to as ‘the tip of the iceberg’ because they are limited almost entirely to motor vehicle-related events that occur on public roadways. Specifically, they exclude (1) many bicycle-motor vehicle and pedestrian-motor vehicle crashes that occur in non-roadway locations such as parking lots, driveways, and sidewalks; and (2) bicyclist and pedestrian falls or other non-collision events that do not involve a motor vehicle, regardless of whether they occur on a roadway or in a non-roadway location. Even using emergency-room data will not fill in the gaps because many of the injuries may not result in visits to the emergency room, and if they do, the forms that are filled out will not include information on the purpose of the transportation (e.g., pleasure, to/from work, to/from school, etc.). The various datasets that were examined for possible use in the committee’s analyses are briefly described in this chapter. A detailed description of the three datasets selected, the analyses conducted with each, and the limitations of each for the purposes of this study is then provided. Finally, conclusions are presented.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment DATA SOURCES The committee identified a number of databases that could be used in assessing the safety of the various student transportation modes; unfortunately, most of the datasets are highly limited, contain data that cannot be used to generalize to the population of interest (i.e., all students in kindergarten through grade 12), or do not include the data categories needed to conduct the analyses of interest (e.g., purpose of trip or time of day). Only one dataset, the Nationwide Personal Transportation Survey (NPTS), provides any usable exposure data at the national level. The database of the National Association of State Directors for Pupil Transportation Services (NASDPTS), recorded annually in industry journals and School Bus Fleet and School Transportation News magazines, contains the most accurate data available on school bus ridership on a state-by-state basis.1 However, the underlying data collected by the states, in many cases, is used to reimburse school districts for school bus services resulting in an overestimation of the number of student riders. In addition, the database does not contain data on the other modes used by children traveling to and from school. In addition, as with the other datasets, there is a lack of consistency in terminology. For example, some states report ridership based on actual head counts, others base their counts on the number of students “authorized” to use the school bus in their daily commute to and from school, and some also include private school ridership. In this dataset, it should be noted, ridership numbers are attached to trip purpose; however, this dataset could not be linked to others because it does not use the data selection criterion of normal school travel hours that is applied for the other modes. Therefore, the NASDPTS data could not be used for cross-mode comparisons. Nine national outcome (or crash) datasets were considered by the committee: (a) the National Automotive Sampling System (NASS) General Estimates System (GES), (b) the NASS Crashworthiness Data System (CDS), (c) the NASS Pedestrian Crash Data System (PCDS), (d) the Crash Outcome Data Evaluation System (CODES), (e) the Highway Safety Information System (HSIS), (f) the Kansas Department of Education school bus loading/unloading fatality dataset, (g) the National Transit Database (NTD), (h) the National Electronic Injury Surveillance System (NEISS), and (i) the Fatality Analysis Reporting System (FARS).2 1 On the basis of this dataset, School Transportation News estimates annual school bus ridership to be approximately 10.5 billion. Using the Annual Student Ride Formula, this calculation incorporates estimates for student rides on regular route to-and-from K-12 school service, school-related activity trips, some private and parochial school transportation service, summer school transportation service, and Head Start transportation service. Unlike the data used in the committee’s analyses, this estimate is not restricted to children 5 to 18 years of age, specific time periods during the day, and specific months of the year. 2 The committee was interested in examining state fatality and injury data, as well as insurance industry data, but did not have time or resources available to do so for all 50 states. The committee did select two states (Texas and California) for which available fatality and injury data were obtained and analyzed; however, the data could not be compared directly with the national data because of the lack of data for nonbus modes and the small counts. In addition, the committee could not easily access crash and injury data from the Department of Education. These data and data from local school districts could also prove valuable.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment GES, which became operational in 1988, contains data from a nationally representative sample of police-reported motor vehicle crashes of all types. The reports are chosen from 60 areas that reflect the geography, roadway mileage, population, and traffic density of the United States. Data are collected weekly from approximately 400 police jurisdictions, from which about 50,000 police accident reports (PARs) are randomly sampled each year. PARs are completed by police officers investigating crashes that result in personal injury and/or property damage above the state’s reporting threshold. For example, in the state of Texas, if a vehicle involved in a crash is towed, a PAR must be completed, while in the state of California, one must be completed if property damage is estimated to be $500 or more. Each state, the District of Columbia, Puerto Rico, and the Virgin Islands has a unique crash report form. The reader is referred to the State Crash Report Forms Catalog 1999 Update (NHTSA 1999), which contains a copy of each state’s form, as well as the state crash reporting threshold. The National Highway Traffic Safety Administration (NHTSA) encourages uniformity across states in the data elements contained on the crash report form, and for this purpose encourages the use of American National Standards Institute (ANSI) D-16 (Manual on Classification of Motor Vehicle Traffic Accidents) and D-20 (Data Element Dictionary for Traffic Records Systems). NHTSA, in collaboration with the Federal Highway Administration (FHWA) and the National Association of Governors’ Highway Safety Representatives, has also developed a guideline for collection of crash data, referred to as the Model Minimum Uniform Crash Criteria (NHTSA 1998). Although states do not generally record data from crashes (either on- or off-highway) in which a motor vehicle is not involved, GES is the most complete injury database available. CDS, which covers about 5,000 accidents in depth annually, contains data on vehicle occupants and is used to study injury mechanisms; precrash events are not examined in detail. This shortcoming is moot for purposes of the committee’s analysis, however, as there are not enough cases involving school and transit vehicles for this dataset to be useful. PCDS contains data on pedestrian crashes. It is essentially an update (or continuation) of earlier pedestrian data files, such as the Pedestrian Injury Causation Study of 1977 and the special pedestrian data collected for NASS from 1979 through 1987. PCDS data for 1995–1997 relate to only 280 cases, too few for the purposes of this study. Although there is a code for school buses and other buses, none of the 280 cases in the file involved these types of vehicles. CODES contains linked statewide crash and injury data for 20 states that match vehicle, crash, and human characteristics with final medical and financial outcomes. The purpose of the dataset is to improve decision making related to highway safety and injury control. CODES studies often tend to be in-depth examinations of special problems; however, they usually cover highly limited geographical areas. HSIS, which contains selected data for eight states (California, Illinois, Maine, Michigan, Minnesota, North Carolina, Utah, and Washington), is a roadway-based system that includes data on a large number of accident, roadway, and traffic variables. The data are collected annually from the participating states,

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment processed, documented, and prepared for analysis. The dataset contains little information specifically about travel to and from school, and thus could not be used to address the issues of concern to the committee. The Kansas Department of Education conducts an annual national survey of school bus loading and unloading fatalities. As with the NASDPTS dataset, however, it does not contain data for the other modes, making comparison with other studies difficult, if not statistically invalid. For example, there are no comparable loading and unloading data for the other modes. NTD, which is maintained by the Federal Transit Administration (FTA), contains performance, operational (such as miles traveled, passenger-miles, and passengers carried), and financial information, as well as injury data stratified by mode [motor bus, trolley bus, light rail (passenger rail operating on exclusive or separated guideways)] reported by all transit agencies receiving federal assistance.3 This database, which is used for management and planning by transit systems, as well as for policy analysis and investment decision making, also contains data on numbers of fatalities and injuries by transit system; however, data on age of victim and purpose of trip are not included. The data collected make it possible to compute injury rates for each transit mode for each agency, as well as national rates by mode. However, the data are not reported by passenger age, and this greatly restricts use of the data in school transportation analyses. NEISS, which has been maintained for approximately 30 years by the Consumer Product Safety Commission (CPSC), is a representative sampling of U.S. hospital emergency departments (Stutts and Hunter 1999). Its primary purpose has been to provide data on consumer product–related injuries occurring in the United States. In 2000, CPSC initiated an expansion of the system to collect data on all injuries. Thus, although not useful to the committee for the present study, NEISS may provide highly useful information in the future, depending on the data elements that are added. FARS, established in 1975, provides a census of all highway fatalities (including deaths of school-age children by time of day and vehicle type) in all 50 states, the District of Columbia, and Puerto Rico. The data in FARS are obtained from existing state data: PARs, state vehicle registration files, state driver licensing files, state highway department data, vital statistics, death certificates, coroner/medical examiner reports, hospital medical reports, emergency medical service reports, and other state records. From these documents, data for more than 100 FARS data elements are obtained. Each year, specific data elements are modified in response to changing user needs, vehicle characteristics, 3 According to the 2002 Public Transportation Fact Book (APTA 2002) there were approximately 9.363 billion total transit trips in 2000; however, this number includes all trips by all transit modes (bus, commuter rail, demand response, light rail, heavy rail, trolley bus, and other). In the committee’s analyses, only data for other buses were included. According to the Fact Book,of these 9.363 billion trips, 5.678 billion were taken by bus, and only about 10 percent of transit trips were provided to riders under 18 years of age. Thus, using NTD data, approximately 568 million trips were taken by children under 18 years of age. This number includes trips taken at all hours during the entire year and also includes children under 5 years of age. School-related trips relevant to this study would be a portion of these trips.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment and highway safety emphasis areas. As specific as the data are, however, no personal identifying information (such as names or addresses) is recorded in FARS. Similar to GES, FARS was developed to provide an overall measure of highway safety, to identify traffic safety problems, and to provide a single objective basis for evaluating the effectiveness of motor vehicle safety standards and highway safety incentives. NHTSA annually provides descriptive statistics for traffic crashes of all degrees of severity in its Traffic Safety Facts series, based on GES and FARS data (see NHTSA 2000 for a recent report). (FARS data may also be accessed on the Internet at www.fars.nhtsa.dot.gov.) DATASETS USED FOR THIS STUDY Given the study objectives and the data available in the various datasets, the committee elected to use the following three datasets for its analyses: NPTS, the only national dataset that contains exposure/travel information, used to estimate the number of trips taken and miles traveled by school-age children; FARS, used to analyze student fatalities; and GES, used to analyze student injuries. Information from all three datasets had to be filtered and grouped to extract relevant inputs for the committee’s risk analyses. Three main generalizations had to be made: (a) time of day was used as a surrogate for school travel trips; (b) age groupings were created from individual age groups; and (c) general transportation mode categories were delineated. The FARS and GES datasets do not reliably record purpose of trip for each incident. As a surrogate for purpose of trip, the committee defined school travel hours (i.e., hours during which most school-related trips would occur), as follows:4 Months of year: September 1 through June 155, 6 Days of week: Monday through Friday Hours of day: 6:00 a.m. to 8:59 a.m. and 2:00 p.m. to 4:59 p.m.7 4 All child fatalities not recorded during the months, days, and hours shown for school travel are referred to as occurring during non–school travel hours. 5 The GES data analysis included the entire month of June because, while it is possible to select days of the week, it is not possible to break out or select part of a month (i.e., the first 2 weeks but not the last 2 weeks of the month). 6 In this study, more than 180 days are included in the defined “school year” to capture school trips for children attending schools with different schedules. In addition, the GES and NPTS datasets do not allow one to identify holidays or other special days in order to remove data for these days from the analyses. The rates per school day are rough estimates that assume different travel behavior for children during vacations (i.e., significantly fewer trips during normal school travel hours while on vacation and minimal school bus use). 7 The selection of these hours for school travel omits the transporting of kindergarten children to and from school during midday (for those school districts in which kindergarten is a half-day). A fatality or injury of a kindergartner or bus driver during such a midday trip would not be included in the reported risk calculations. Also omitted are injuries and fatalities that occur on school activity trips that take place during the school day between 9:00 a.m. and 1:59 p.m. or after 5:00 p.m. Thus, trip purpose and time of day are not perfectly correlated.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment Because of the limited amount of data in the two output datasets (FARS and GES), an analysis for each age group was not possible. Therefore, the committee created the following four age groupings: 5–10 years of age, 11–13, 14–15, and 16–18. This categorization is associated with the ages at which children change schools (the break around 10–11 years is when children generally move from elementary to middle school; 11–13 years of age maps to middle-school ages, and 14–18 is the rough equivalent of high school ages). These groupings also correspond roughly to transition points in the developmental and behavioral characteristics of children. Further, while the last two age groupings both include high school students, those aged 16–18 are more likely to have a driver’s license (or friends that have one), resulting in a different distribution of travel modes used. In addition, those aged 16–18 are more likely to hold after-school jobs or to participate in after-school activities that require transportation after regularly scheduled school bus service times. Therefore, the committee believed this last age group would be significantly different in both modes of travel used and number/length of trips during normal school travel hours, and thus should be examined separately. There are many ways to classify injury and fatality data. Table 2-1 shows a classification consisting of 16 categories. FARS and GES provide data for 14 of these categories; categories 6 and 8 do not include motor vehicles. For much of the analysis, these 16 categories were combined into six more general categories: Categories 1 and 2—school bus–related crashes; Categories 3, 4, and 5—passenger vehicle crashes; Categories 6 and 7—pedestrian crashes; Categories 8, 9, and 10—bicyclist crashes; Categories 11 and 12—other bus crashes; and Categories 13, 14, and 15—motorcycle crashes. Nationwide Personal Transportation Survey Description NPTS serves as the nation’s inventory of daily personal travel. It is a computer-assisted telephone interview survey of households in the United States. For each trip made during a preselected 24-hour time period, data are gathered regarding purpose of trip, mode of travel, length of trip, day of travel, vehicle occupancy, driver characteristics (e.g., age, gender, worker status, education level), and vehicle attributes (e.g., make, model, model year, annual miles driven, odometer readings). These data are gathered for all areas of the country, all days of the week, and all months of the year. Data are collected only for the civilian, noninstitutionalized population in the United States aged 5 years and older. NPTS does not include responses from military personnel living on base or overseas, or from residents of nursing homes, assisted-living facilities, long-term medical institutions, college dormitories, or prisons. NPTS has been conducted five times—in 1969, 1977, 1983, 1990, and 1995; the sixth survey was initiated in late 2001. Since the survey contents and method-

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment TABLE 2-1 Categories Used to Classify Injury and Fatality Incident Data Category Description 1 Child school bus passenger in a school bus–related crasha 2 Child pedestrian in a school bus–related crash 3 Child passenger in passenger vehicle driven by an adult 4 Driver younger than 19, passenger vehicle 5 Child passenger in passenger vehicle, driver younger than 19 6 Child pedestrian not involved in a motor vehicle crashb 7 Child pedestrian incident, not school bus–relatedc 8 Child bicyclist not involved in a motor vehicle crashb 9 Child bicyclist in a school bus–related crash 10 Child bicyclist not involved in a school bus–related crash 11 Other bus, driver younger than 19 12 Child passenger in other bus 13 Child passenger on motorcycle operated by an adult 14 Motorcycle, driver (operator) younger than 19 15 Child passenger on a motorcycle operated by a driver younger than 19 16 Unknown a Includes any child riding in a vehicle being used as a school bus. b Not included in FARS or GES. c This category includes child pedestrian injuries and fatalities occurring in other bus-related crashes as well as in incidents with passenger vehicles driven by teens. ology have been modified each time, data from multiple years cannot be used in one analysis without some type of data manipulation. Therefore, the committee used data from the most recent survey available at the time (the 1995 survey) for its analyses. In 1995, NPTS consisted of a national sample of 21,020 households and an additional 21,013 households in five add-on areas (New York; Massachusetts; Oklahoma City and Tulsa, Oklahoma; and Seattle, Washington). The data were collected from May 1995 through July 1996 (Chen et al. 2000). [See RTI and FHWA (1997) and Hu and Young (1999) for an in-depth description of the 1995 NPTS survey procedures and methodology.] NPTS is a random sample of the nine census regions stratified by population size and other factors. The 1995 survey began with 160,000 telephone numbers, approximately 45 percent of which were dropped when there was no response. Because the stratification was done using telephone numbers, all households without a telephone—a group in which households at the poverty level are over-represented—were excluded. It was determined that approximately 72 percent of households that received the travel log filled it out. The overall response rates were 55.3 percent for household-level data and 34.3 percent for person-level data.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment Analyses Given the data fields available in NPTS, it is possible to extract data on trips made and miles traveled by transportation mode, age, date, time of day, purpose of trip, and type of geographic region (i.e., rural versus urban).8 On the basis of this dataset, it was estimated that approximately 50 million students (Census Bureau 2000) made 9.7 billion trips (5.9 billion urban and 3.8 billion rural) to school during normal morning school travel hours and traveled approximately 44.0 billion student-miles (20.0 billion urban and 24.0 billion rural). Table 2-2 provides data on number of trips, and Table 2-3 shows student-miles traveled during normal morning and afternoon school travel hours. Tables containing detailed trip and student-mile data are provided in Annex 2-1. The data suggest that, on average, each student made approximately 194 trips during normal morning school travel hours throughout the school year—an average of 1.1 school trips per student each school morning.9 The definition of a trip is one-way travel from one address to another. Thus, if a person travels from home to school to drop off a student, and then goes to work, the driver has made two trips and the student one. If a child walks to the bus stop and rides a bus to school, this is considered one trip. It was estimated that during normal afternoon school travel hours, these same students made 13.8 billion trips (8.6 billion urban and 5.2 billion rural) and traveled approximately 69.3 billion student-miles (34.7 billion urban and 34.6 billion rural). The data suggest that, on average, each student made approximately 276 trips during normal afternoon school travel hours throughout the school year, which translates to approximately 1.5 trips per student each afternoon. Given that it is known, at least anecdotally, that more trips are taken in the afternoon than in the morning travel period (e.g., to run errands, go to the library, go to lessons, meet friends), one would expect there to be more afternoon than morning trips on average. For each analysis, the sample size (N), the estimated statistic (trips or student-miles) based on the sample size, and the standard error are reported in the tables in Annex 2-1. These data are the basis for the analyses in Chapter 3 that compare the risk to students of traveling to and from school during normal school travel hours using the various modes. Limitations The committee’s analyses were hampered by problems inherent in the quantity and quality of the data, especially difficulties in obtaining accurate and reliable enrollment and ridership data. This was especially the case when the committee tried to gather data at other than the aggregate level. As noted, the NPTS data are based on a sampling of some 42,000 households in the United States, and 8 In the 1995 NPTS dataset, urbanized areas were defined as areas having population densities of at least 1,000 persons per square mile (RTI and FHWA 1997). 9 As noted earlier, time of day rather than stated purpose of trip was used to determine school-related travel trips because the outcome datasets (i.e., FARS and GES) do not report purpose-of-trip information reliably.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment TABLE 2-2 Population Estimates for Number of Student Trips Made During Normal Morning and Afternoon School Travel Hours by Mode Mode Morning Afternoon Total Urban Rural Total Urban Rural Total Passenger vehicle 3,279,726,554 2,092,628,516 5,372,355,070 4,831,405,530 3,049,681,460 7,881,086,990 13,253,442,060 School bus 1,309,995,348 1,441,861,776 2,751,857,124 1,309,959,884 1,416,289,818 2,726,249,702 5,478,106,826 Other bus 163,558,718 36,486,530 200,045,248 198,811,400 40,970,533 239,781,933 439,827,181 Bicycle 58,188,176 16,409,185 74,597,361 220,961,475 156,812,750 377,774,225 452,371,586 Walk 777,336,383 133,346,747 910,683,130 1,428,155,719 358,272,221 1,786,427,940 2,697,111,070 Other 28,403,226 1,488,789 29,892,015 34,962,929 6,140,836 41,103,765 70,995,780 Unknown 263,551,965 68,015,175 331,567,140 567,138,229 189,249,284 756,387,513 1,087,954,653 Total 5,880,760,371 3,790,236,718 9,670,997,089 8,591,395,167 5,217,416,902 13,808,812,069 23,479,809,158

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment TABLE 2-3 Population Estimates for Student-Miles Traveled During Normal Morning and Afternoon School Travel Hours by Mode Mode Morning Afternoon Total Urban Rural Total Urban Rural Total Passenger vehicle 11,878,985,126 13,133,834,292 25,012,819,418 27,063,682,981 23,290,793,916 50,354,476,897 75,367,296,315 School bus 5,512,961,165 10,175,511,847 15,688,473,012 5,142,320,970 10,130,758,328 15,273,079,298 30,961,552,310 Other bus 1,504,235,227 574,239,365 2,078,474,592 1,004,595,164 651,981,293 1,656,576,457 3,735,051,049 Bicycle 70,908,770 9,919,170 80,827,940 181,293,770 109,111,080 290,404,850 371,232,790 Walk 404,887,916 70,401,933 475,289,849 791,461,811 32,911,699 824,373,510 1,299,663,359 Other 256,003,012 1,544,134 257,547,146 288,301,481 43,662,333 331,963,814 589,510,960 Unknown 390,248,076 65,991,805 456,239,881 256,879,036 182,406,298 439,285,334 895,525,215 Total 20,018,229,293 24,031,442,548 44,049,671,841 34,728,535,213 34,621,458,107 69,349,993,320 113,399,665,161

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment they cannot be used reliably for analyses at the state and community levels with any degree of confidence. Thus the committee’s ability to gauge the effects of state-by-state or local differences was constrained. While it was recognized that substantial variation exists in the modes of student travel according to location and such factors as climate, infrastructure, and local economic and demographic characteristics, the committee was unable to measure this variation directly. Without such data, conclusions concerning the direct effects of local conditions on the numbers of fatalities and injuries could not be drawn since rates and risk ratios could not be computed. In addition, the data in the NPTS database are self-reported, a feature associated with many well-known limitations. For example, there may be an undercount of trips because, in the case of pupil transportation, the bus trip to school may be documented, but not the trip between home and the bus stop. All travel data for children aged 5–13 are reported by their parents, while travel data for teens aged 14–17 may be either self-reported or reported for them by household adults. Teenagers may not report all afternoon trips they made, or parents responding for their older children may not be aware of all the trips the children made. The NPTS survey is intended to gather information on personal travel of U.S. households, including why, how, when, where from, where to, how frequently, how long, and with whom trips were made. The survey is not limited to schoolage children and not focused on school-related transportation—small facets of daily travel that are captured by the dataset only to a limited extent. This can be seen by the extremely small sample size (N) shown in the tables in Annex 2-1. Fatality Analysis Reporting System Description Number of fatalities was one of two outcome measures examined by the committee. These data are contained in FARS, a database first developed in 1975 that contains only data on all fatal traffic crashes that occur on public roadways in the United States.10 Data in this national database are extracted from medical examiner, coroner, emergency medical, and police accident reports, as well as from driver, vehicle, and roadway classification records. There is detailed information in the database on crash, vehicle, driver, and occupant characteristics. Analyses Analyses were performed using 9 years of FARS data (1991–1999). In these 9 years, a total of 51,350 children between the ages of 5 and 18 were killed in all traffic crashes in the United States (Table 2-4). Of these, 7,470 were killed during normal school travel hours,11 2,719 were killed during school session 10 To be classified as a fatal crash, an incident must involve a vehicle occupant or nonmotorist who dies within 30 days of the crash from injuries caused by the crash. 11 These tables include data for the previously defined 205 school travel days. Extra days include holidays and other weekday nonschool days. Because of the extra days included in the tables, fatalities for non–school bus modes may be overrepresented. See note 4.

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 7 Child Fatalities During Normal School Travel Hours by Fatality Category and Age Group (N = 7,470) Category Description Age Group (years) 5–10 11–13 14 –15 16 –18 Total N % N % N % N % N % 1 Child school bus passenger fatality in a school bus–related crash 15 0.9 12 1.5 10 10 4 0.1 41 0.5 2 Child pedestrian fatality in a school bus–related crash 105 6.1 19 2.3 9 0.9 3 0.1 136 1.8 3 Child passenger fatality in all other vehicles driven by an adult 747 43.2 238 28.8 161 15.7 371 9.5 1,517 20.3 4 Child driver fatality, all other vehicles 3 0.2 22 2.7 139 13.6 2,381 61.2 2,545 34.1 5 Child passenger fatality in all other vehicles driven by a child 61 3.5 119 14.4 459 44.8 844 21.7 1,483 19.9 7 Child pedestrian fatality, not school bus–related 623 36.0 264 32.0 158 15.4 132 3.4 1,177 15.8 9 Child bicyclist fatality in a school bus–related crash 5 0.3 5 0.6 0 0 2 0.1 12 0.2 10 Child bicyclist fatality not in a school bus–related crash 165 9.5 139 16.8 50 4.9 48 1.2 402 5.4 12 Child passenger fatality in other buses 1 0.1 2 0.2 1 0.1 1 0.0 5 0.1 13 Child passenger fatality on motorcycle operated by an adult 0 0 0 0 1 0.1 6 0.2 7 0.1 14 Child driver (operator) fatality, motorcycle 3 0.2 5 0.6 36 3.5 91 2.3 135 1.8 15 Child passenger fatality on motorcycle operated by a child 1 0.1 1 0.1 1 0.1 7 0.2 10 0.1 Total   1,729 100.0 826 100.0 1,025 100.0 3,890 100.0 7,470 100.0

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 8 Child Fatalities During Normal School Travel Hours by Fatality Category and Age (N = 7,470) Category Description Age (years) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Total 1 Child school bus passenger fatality in a school bus-related crash 0 0 3 5 7 0 1 4 7 3 7 3 0 1 41 2 Child pedestrian fatality in a school bus-related crash 35 25 17 16 7 5 9 4 6 6 3 0 1 2 136 3 Child passenger fatality in all other vehicles driven by an adult 170 136 126 128 87 100 81 78 79 80 81 90 113 168 1,517 4 Child driver fatality, all other vehicles 0 0 0 0 2 1 2 9 11 36 103 793 766 822 2,545 5 Child passenger fatality in all other vehicles driven by a child 10 11 4 9 17 10 19 36 64 168 291 357 311 176 1,483 7 Child pedestrian fatality, not school bus-related 88 128 113 91 99 104 92 101 71 91 67 55 49 28 1,177

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment Category Description Age (years) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Total 9 Child bicyclist fatality in a school bus-related crash 1 1 0 1 2 0 2 2 1 0 0 1 0 1 12 10 Child bicyclist fatality not in a school bus-related crash 13 23 30 31 27 41 51 45 43 24 26 15 20 13 402 12 Child passenger fatality in other buses 0 0 1 0 0 0 1 1 0 0 1 1 0 0 5 13 Child passenger fatality on motorcycle operated by an adult 0 0 0 0 0 0 0 0 0 0 1 0 3 3 7 14 Child driver (operator) fatality, motorcycle 0 0 1 1 0 1 0 1 4 11 25 15 29 47 135 15 Child passenger fatality on motorcycle operated by a child 0 1 0 0 0 0 0 1 0 1 0 3 2 2 10 Total   317 325 295 282 248 262 258 282 286 420 605 1,333 1,294 1,263 7,470

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 9 Child Fatalities During Non–School Travel Hours by Fatality Category and Population (N = 43,374) Category Description Population Rural Urban Total N % N % N % 1 Child school bus passenger fatality in a school bus–related crash 7 0.0 7 0.0 14 0.0 2 Child pedestrian fatality in a school bus–related crash 5 0.0 19 0.1 24 0.1 3 Child passenger fatality in all other vehicles driven by an adult 7,727 28.0 3,708 23.6 11,435 26.4 4 Child driver fatality, all other vehicles 10,266 37.2 4,083 25.9 14,349 33.1 5 Child passenger fatality in all other vehicles driven by a child 6,095 22.1 3,008 19.1 9,103 21.0 7 Child pedestrian fatality, not school bus–related 1,766 6.4 3,083 19.6 4,849 11.2 9 Child bicyclist fatality in a school bus–related crash 1 0.0 1 0.0 2 0.0 10 Child bicyclist fatality not in a school bus–related crash 1,056 3.8 1,248 7.9 2,304 5.3 11 Child driver fatality, other buses 0 0 1 0.0 1 0.0 12 Child passenger fatality in other buses 16 0.1 11 0.1 27 0.1 13 Child passenger fatality on motorcycle operated by an adult 89 0.3 85 0.5 174 0.4 14 Child driver (operator) fatality, motorcycle 550 2.0 441 2.8 991 2.3 15 Child passenger fatality on motorcycle operated by a child 55 0.2 46 0.3 101 0.2 Total   27,633 100.0 15,741 100.0 43,374 100.0

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 10 Child Fatalities During Non–School Travel Hours by Fatality Category and Age Group (N = 43,374) Category Description Age Group (years) 5–10 11–13 14–15 16–18 Total N % N % N % N % N % 1 Child school bus passenger fatality in a school bus–related crash 4 0.1 3 0.1 2 0.0 5 0.0 14 0.0 2 Child pedestrian fatality in a school bus–related crash 16 0.2 4 0.1 3 0.0 1 0.0 24 0.1 3 Child passenger fatality in all other vehicles driven by an adult 3,735 53.5 1,684 41.5 1,574 25.0 4,442 17.1 11,435 26.4 4 Child driver fatality, all other vehicles 66 0.9 211 5.2 984 15.7 13,088 50.3 14,349 33.1 5 Child passenger fatality in all other vehicles driven by a child 262 3.8 625 15.4 2,355 37.5 5,861 22.5 9,103 21.0 7 Child pedestrian fatality, not school bus–related 1,954 28.0 775 19.1 743 11.8 1,377 5.3 4,849 11.2 9 Child bicyclist fatality in a school bus–related crash 1 0.0 – – – – 1 0.0 2 0.0 10 Child bicyclist fatality not in a school bus–related crash 900 12.9 650 16.0 402 6.4 352 1.4 2,304 5.3 11 Child driver fatality, other buses – – – – – – 1 0.0 1 0.0 12 Child passenger fatality in other buses 5 0.1 8 0.2 7 0.1 7 0.0 27 0.1 13 Child passenger fatality on motorcycle operated by an adult 23 0.3 17 0.4 21 0.3 113 0.4 174 0.4 14 Child driver (operator) fatality, motorcycle 13 0.2 69 1.7 161 2.6 748 2.9 991 2.3 15 Child passenger fatality on motorcycle operated by a child 5 0.1 15 0.4 34 0.5 47 0.2 101 0.2 Total   6,984 100.0 4,061 100.0 6,286 100.0 26,043 100.0 43,374 100.0

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 11 Child Fatalities During Non–School Travel Hours by Fatality Category and Age (N = 43,374) Category Description Age (years) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Total 1 Child school bus passenger fatality in a school bus-related crash 1 0 1 0 2 0 0 0 3 1 1 3 2 0 14 2 Child pedestrian fatality in a school bus-related crash 6 5 2 3 0 0 1 1 2 1 2 1 0 0 24 3 Child passenger fatality in all other vehicles driven by an adult 690 611 660 631 595 548 529 568 587 731 843 1,082 1,404 1,956 11,435 4 Child driver fatality, all other vehicles 1 2 7 10 19 27 34 57 120 287 697 3,423 4,255 5,410 14,349 5 Child passenger fatality in all other vehicles driven by a child 33 35 31 46 53 64 93 174 358 801 1,554 2,158 2,039 1,664 9,103 7 Child pedestrian fatality, not school bus-related 430 361 354 316 259 234 225 244 306 348 395 421 412 544 4,849

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment Category Description Age (years) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Total 9 Child bicyclist fatality in a school bus-related crash 0 0 0 1 0 0 0 0 0 0 0 0 1 0 2 10 Child bicyclist fatality not in a school bus-related crash 97 114 165 173 170 181 222 192 236 229 173 145 113 94 2,304 11 Child driver fatality, other buses 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 12 Child passenger fatality in other buses 0 0 0 1 4 0 3 2 3 2 5 4 2 1 27 13 Child passenger fatality on motorcycle operated by an adult 2 3 3 2 7 6 3 6 8 10 11 32 36 45 174 14 Child driver (operator) fatality, motorcycle 0 0 3 0 1 9 10 25 34 68 93 103 215 430 991 15 Child passenger fatality on motorcycle operated by a child 0 0 0 2 1 2 1 3 11 16 18 18 10 19 101 Total   1,260 1,131 1,226 1,185 1,111 1,071 1,121 1,272 1,668 2,494 3,792 7,390 8,489 10,164 43,374

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 12 Child Fatalities During School Session Hours by Fatality Category and Population (N = 2,719) Category Description Population Rural Urban Total N % N % N % 1 Child school bus passenger fatality in a school bus–related crash 1 0.1 5 0.5 6 0.2 2 Child pedestrian fatality in a school bus–related crash 2 0.1 14 1.4 16 0.6 3 Child passenger fatality in all other vehicles driven by an adult 484 28.0 176 17.8 660 24.3 4 Child driver fatality, all other vehicles 723 41.8 344 34.8 1,067 39.2 5 Child passenger fatality in all other vehicles driven by a child 372 21.5 219 22.2 591 21.7 7 Child pedestrian fatality, not school bus–related 66 3.8 122 12.3 188 6.9 10 Child bicyclist fatality not in a school bus–related crash 49 2.8 67 6.8 116 4.3 12 Child passenger fatality in other buses 1 0.1 1 0.1 2 0.1 13 Child passenger fatality on motorcycle operated by an adult 1 0.1 4 0.4 5 0.2 14 Child driver (operator) fatality, motorcycle 28 1.6 32 3.2 60 2.2 15 Child passenger fatality on motorcycle operated by a child 4 0.2 4 0.4 8 0.3 Total   1,731 100.0 988 100.0 2,719 100.0

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 13 Child Fatalities During School Session Hours by Fatality Category and Age Group (N = 2,719) Category Description Age Group (years) 5–10 11–13 14–15 16–18 Total N % N % N % N % N % 1 Child school bus passenger fatality in a school bus–related crash 3 0.7 1 0.5 1 0.3 1 0.1 6 0.2 2 Child pedestrian fatality in a school bus–related crash 12 2.6 2 0.9 1 0.3 1 0.1 16 0.6 3 Child passenger fatality in all other vehicles driven by an adult 265 57.9 110 49.5 94 24.5 191 11.5 660 24.3 4 Child driver fatality, all other vehicles 3 0.7 14 6.3 79 20.6 971 58.7 1,067 39.2 5 Child passenger fatality in all other vehicles driven by a child 23 5.0 33 14.9 159 41.4 376 22.7 591 21.7 7 Child pedestrian fatality, not school bus–related 98 21.4 30 13.5 23 6.0 37 2.2 188 6.9 10 Child bicyclist fatality not in a school bus–related crash 53 11.6 26 11.7 15 3.9 22 1.3 116 4.3 12 Child passenger fatality in other buses 0 0.0 0 0.0 0 0.0 2 0.1 2 0.1 13 Child passenger fatality on motorcycle operated by an adult 0 0 1 0.5 1 0.3 3 0.2 5 0.2 14 Child driver (operator) fatality, motorcycle 0 0 3 1.4 8 2.1 49 3.0 60 2.2 15 Child passenger fatality on motorcycle operated by a child 1 0.2 2 0.9 3 0.8 2 0.1 8 0.3 Total   458 100.0 222 100.0 384 100.0 1,655 100.0 2,719 100.0

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment ANNEX 2-2 TABLE 14 Child Fatalities During School Session Hours by Fatality Category and Age (N = 2,719) Category Description Age (years) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Total 1 Child school bus passenger fatality in a school bus–related crash 0 0 1 0 2 0 0 0 1 0 1 0 1 0 6 2 Child pedestrian fatality in a school bus–related crash 6 3 1 2 0 0 0 1 1 1 0 1 0 0 16 3 Child passenger fatality in all other vehicles driven by an adult 62 46 41 45 38 33 28 38 44 39 55 45 63 83 660 4 Child driver fatality, all other vehicles 0 0 0 0 1 2 0 4 10 19 60 221 295 455 1,067 5 Child passenger fatality in all other vehicles driven by a child 4 6 2 3 4 4 5 7 21 54 105 145 141 90 591 7 Child pedestrian fatality, not school bus–related 28 25 17 11 12 5 15 1 14 10 13 10 14 13 188

OCR for page 25
The Relative Risks of School Travel: A National Perspective and Guidance for Local Community Risk Assessment   Category Description Age (years) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Total 10 Child bicyclist fatality not in a school bus–related crash 5 8 13 5 9 13 8 9 9 6 9 8 7 7 116 12 Child passenger fatality in other buses 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2 13 Child passenger fatality on motorcycle operated by an adult 0 0 0 0 0 0 0 0 1 1 0 0 1 2 5 14 Child driver (operator) fatality, motorcycle 0 0 0 0 0 0 0 0 3 4 4 7 17 25 60 15 Child passenger fatality on motorcycle operated by a child 0 0 0 0 0 1 1 0 1 2 1 0 1 1 8 Total 105 88 75 66 66 58 57 60 105 136 248 437 541 677 2,719