Skip to main content

Currently Skimming:


Pages 100-200

The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.


From page 100...
... 100 This section of the report describes the development of pedestrian and bicycle SPFs and incorporates pedestrian and bicycle exposure data into the modeling approach for potential incorporation in HSM2. SPFs were developed to predict pedestrian and bicycle crashes for urban and suburban roadway segments and intersections.
From page 101...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 101   Based on the results of the literature review, survey of practice, and discussions with selected agencies, it was decided that the best approach for incorporating pedestrian and bicycle exposure in the SPF development was to obtain count data for pedestrians and bicyclists and then develop direct demand models to estimate pedestrian and bicycle volumes for the study sites. Priority was given to searching for potential count data from cities, MPOs, and local transportation agencies compared to potential data sources from state transportation agencies as counts collected by state transportation agencies were considered too spread out geographically for efficient data collection purposes.
From page 102...
... 102 Pedestrian and Bicycle Safety Performance Functions Additionally, information on the number of bus/transit stops, schools, and alcohol sales establishments located in proximity to the study locations was gathered electronically. Figure  8 is a screen capture of the data collection tool for roadway segments and illustrates the data elements that were manually collected for roadway segments.
From page 103...
... Shared Path Shared Path Buffer Sidewalk Sidewalk Buffer Outside Shoulder Travel Lanes Ln1..
From page 104...
... 104 Pedestrian and Bicycle Safety Performance Functions Variable/Parameter Definition/Description Range or Permitted Values General Roadway Attributes Route identifier Unique identification number for roadway segment Character value Two-way versus one-way operation Indicates if the roadway operates with two-way traffic or one-way traffic One-way, two-way Presence of lighting Indicates if overhead lighting is present along the roadway Yes, no Presence of traffic calming Indicates if any traffic calming features (e.g., speed humps, chicanes) are present along the roadway Yes, no Median width Measured from outside of adjacent motor vehicle traffic lane in the one direction to the outside of adjacent motor vehicle lane in the opposing direction Values in feet Median type Type of median separating opposing directions of travel None, painted, depressed, raised, two-way left-turn lane (TWLTL)
From page 105...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 105   Variable/Parameter Definition/Description Range or Permitted Values One-way versus two operation along the bike lane Indicates whether bicycles travel in one direction or two directions along the bike lane One-way, two-way Presence of colored pavement along the bicycle facility Indicates the presence of colored pavement (e.g., green) in the marked bike lane (or buffered bike lane)
From page 106...
... 106 Pedestrian and Bicycle Safety Performance Functions Figure 11. Illustration of data collection tool to manually collect intersection inventory data (approach leg)
From page 107...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 107   Variable/Parameter Definition/Description Range or Permitted Values General Intersection Attributes Intersection configuration (i.e., number of legs and type of traffic control) Indicates the number of legs and type of traffic control 3ST, 4ST, 3SG, 4SG Two-way versus one-way operation Indicates whether it is an intersection of two two-way streets or an intersection of a oneway street and a two-way street (only applicable for 4SG)
From page 108...
... 108 Pedestrian and Bicycle Safety Performance Functions Variable/Parameter Definition/Description Range or Permitted Values Width of buffer (inbound/outbound) Width of buffer associated with a buffered bike lane or protected bike lane Values in feet Presence of colored pavement in inbound/outbound bicycle facility Indicates the presence of colored pavement (e.g., green)
From page 109...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 109   for outliers on all variables collected; cross-referencing selected variables for consistency of terminology/definition of terms; checking the data for completeness; and other checks as deemed necessary for each specific dataset. 3.2 Descriptive Statistics of Final Databases Data for approximately 253 miles of urban and suburban roads and 271 urban and suburban intersections were available for the development of pedestrian and bicycle SPFs, incorporating pedestrian and bicyclist exposure data into the modeling approach.
From page 110...
... 110 Pedestrian and Bicycle Safety Performance Functions • Shared-use path (i.e., a bikeway outside the traveled way and physically separated from motorized vehicular traffic by an open space or barrier and either within the highway right-of-way or within an independent alignment. Shared-use paths may also be used by pedestrians.)
From page 111...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 111   Table 37 provides information on the posted speed limit for sites by road type. The average speed limit for all roadway types was in the range of 25 to 30 mph.
From page 112...
... 112 Pedestrian and Bicycle Safety Performance Functions the number of bus stops present along the roadway was represented on a per-mile basis (i.e., number of bus stops/mi)
From page 113...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 113   Table 43 provides information on the number of alcohol sales establishments within 1,000 ft of the center of the roadway segment. This information was gathered from electronic data available through online portals.
From page 114...
... 114 Pedestrian and Bicycle Safety Performance Functions Road Type Total Number of Sites Lane Width (ft)
From page 115...
... Road Type Total Number of Segments Conventional Bike Lane Placement of Bike Lane Bike Lane Width (ft) None One Dir Both Dir Mixed Facilitya Other Bike Facilityb Mean Min.
From page 116...
... 116 Pedestrian and Bicycle Safety Performance Functions at the study site, in most cases a buffered bike lane was present in both directions of travel. For one-way roads, if a buffered bike lane was present at the study site, the buffered bike lane was located only on one side of the road in the direction of travel (i.e., no contraflow buffered bike lanes were included in the analysis)
From page 117...
... Road Type Total Number of Segments Travel Lane with Shared-Lane Markings Placement of Markings Lane Width (ft) None One Dir Both Dir Mixed Facilitya Other Bike Facilityb Mean Min.
From page 118...
... Road Type Total Number of Segments Sidewalk Placement of Sidewalk Buffer Width (ft) Sidewalk Width (ft)
From page 119...
... Road Type Total Number of Segments Parking Lane Location of Parking Lane Parking Lane Width (ft)
From page 120...
... 120 Pedestrian and Bicycle Safety Performance Functions In Minneapolis, annual average daily pedestrian and bicycle volume estimates were both available at 535 unique roadway segments. These estimates were provided by the City of Minneapolis, which applied expansion factors obtained from long-term counting stations to short-term counts at the individual sites.
From page 121...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 121   segment: employment and land use, infrastructure, transit, demographics, and street network connectivity. A summary of each of these variables, including its description, the scale at which it was collected, and the data source used to obtain it, is provided in Table 54.
From page 122...
... 122 Pedestrian and Bicycle Safety Performance Functions Variable Name Description Scale Minneapolis Data Source Philadelphia Data Source Employment/Land Use Emp Number of employees Buffer ESRI Business Analyst EmpSQFT Employment square footage of foot traffic land uses Buffer ESRI Business Analyst LUEntropy Land-use entropy Buffer General land use 2016 DVRPC 2015 land use Infrastructure Arterial Count is on a principal arterial. Point Minneapolis street centerline Delaware County road centerline, Montgomery County centerlines, Philadelphia street centerlines Minorart Count is on a minor arterial.
From page 123...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 123   For Philadelphia, land uses with similar broad categories were considered as follows: • Residential. • Commercial.
From page 124...
... 124 Pedestrian and Bicycle Safety Performance Functions Select summary statistics for the roadway segments included in the Minneapolis analysis database are provided in Table 57. This includes summary statistics for the pedestrian and bicycle volumes, the year during which counts were obtained, and the independent variables included in the final exposure models.
From page 125...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 125   Continuous Variable Mean Std.
From page 126...
... 126 Pedestrian and Bicycle Safety Performance Functions The corresponding mean-variance relationship is described in the following equation.
From page 127...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 127   A forward-selection process was used to estimate the pedestrian and bicycle exposure models. In this method, each potential independent variable shown in Table 54 was tested to examine its impact on the dependent variable and the most impactful variable selected.
From page 128...
... 128 Pedestrian and Bicycle Safety Performance Functions is White, and land-use entropy within 0.25 mi. Bicycle volumes decrease with the number of transit stops within 0.1 mi.
From page 129...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 129   that were in line with expectations. Like the Minneapolis models, the yearly indicators were also included to improve prediction accuracy.
From page 130...
... City Date Range Number of Sites Number of Site-Years Annual Average Daily Pedestrian Volume (AADP)
From page 131...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 131   Some of the crashes in the crash dataset were classified as property-damage-only crashes. By definition, it is difficult to understand how a pedestrian or bicycle crash could be classified as property damage only, but recognizing that some pedestrian and bicycle crashes are classified as such, these crashes were included in the count of total crashes (i.e., all severity levels combined)
From page 132...
... 132 Pedestrian and Bicycle Safety Performance Functions sites in Minneapolis and Philadelphia. Only those intersection configurations for which a sufficient amount of data was collected for potential model development are provided in the table.
From page 133...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 133   Table 71 provides information on the number of intersection legs where a marked parking lane is present in the inbound direction, outbound direction, or both. No distinction is made on whether the parking lane is present on the major or minor roads.
From page 134...
... 134 Pedestrian and Bicycle Safety Performance Functions Intersection Traffic Control Type and Configuration Total Number of Intersections Number of Intersection Legs with a Parking Lane Present 0 1 2 3 4 3ST (2×2)
From page 135...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 135   Intersection Traffic Control Type and Configuration Total Number of Intersections Number of Approaches with a Bike Lane or Buffered Bike Lane Entering the Intersection 0 1 2 3 4 3ST (2×2)
From page 136...
... 136 Pedestrian and Bicycle Safety Performance Functions Table 78 provides information on the number of lanes that a pedestrian must cross at an intersection by major- and minor-road approaches, taking into consideration the presence of a median refuge. Both through lanes and turning lanes that are crossed by a pedestrian along a crossing path are considered.
From page 137...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 137   3.2.2.2 Intersections: Traffic Volumes Traffic volume data were available for up to 13 years (from 2006 to 2018) for study sites in Minneapolis and up to 6 years (from 2013 to 2018)
From page 138...
... AADTmaj (veh/day)
From page 139...
... AADPmaj (ped/day)
From page 140...
... AADBmaj (bike/day) AADBmin (bike/day)
From page 141...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 141   3.2.2.4 Intersections: Crash Data Crash data were available for up to 13 years (from 2006 to 2018) for study sites in Minneapolis and up to 6 years (from 2013 to 2018)
From page 142...
... 142 Pedestrian and Bicycle Safety Performance Functions per year as a function of one or more explanatory variables. This is a common approach to modeling roadway segment and intersection crash frequencies (e.g., Miaou 1994; Shankar, Mannering, and Barfield 1995; Poch and Mannering 1996; El-Basyouny and Sayed 2006)
From page 143...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 143   the underlying statistical distribution assumed for the regression model. The likelihood function for the negative binomial model that was used in this study is shown in Equation 3-8.
From page 144...
... 144 Pedestrian and Bicycle Safety Performance Functions NMVolcrossing = sum of daily nonmotorized volume crossing all intersection legs (ped/day or bike/day) , β1, β2 = regression coefficients for major and minor road AADT, respectively, β1,nm, β2,nm = regression coefficients for major- and minor-road nonmotorized volume (i.e., AADP or AADB)
From page 145...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 145   SPFs. In this method, each potential independent variable considered was tested to examine its impact on the dependent variable and the most impactful variable selected.
From page 146...
... 146 Pedestrian and Bicycle Safety Performance Functions as well as other geometric and site characteristic features found to be significant predictors of total pedestrian crashes. • A reduced model to estimate FS pedestrian crashes (i.e., fatal and suspected serious injury crashes)
From page 147...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 147   Figure 13. Graphical representation of the SPF for predicted average total pedestrian crashes per year on two-lane undivided roads (reduced model)
From page 148...
... 148 Pedestrian and Bicycle Safety Performance Functions Table 88 presents an expanded model for total pedestrian crashes (i.e., all severity levels combined) for four-lane undivided and divided roads.
From page 149...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 149   Figure 15. Graphical representation of the SPF for predicted average total pedestrian crashes per year on four-lane undivided (4U)
From page 150...
... 150 Pedestrian and Bicycle Safety Performance Functions and associated p-values; the inverse of the overdispersion parameter and its standard error; the 2 x log-likelihood at convergence; the total number of crashes; and the total number of sites associated with the model. In addition to exposure measures for motor vehicles and pedestrians and indicator variables for the number of lanes, significant predictors of total pedestrian crashes on one-way roads include: • Presence/absence of a sidewalk buffer greater than 0 ft.
From page 151...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 151   Figure 17. Graphical representation of the SPF for predicted average FS pedestrian crashes per year on four-lane undivided and divided roads (reduced model)
From page 152...
... 152 Pedestrian and Bicycle Safety Performance Functions Variable Coefficient Std. Error p-Value Constant −9.339 2.160 <0.001 Natural log of AADT volume (veh/day)
From page 153...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 153   Figure 19. Graphical representation of the SPF for predicted average total pedestrian crashes per year on one-way roads (expanded model)
From page 154...
... 154 Pedestrian and Bicycle Safety Performance Functions The coefficients of a, b, and c; the overdispersion parameters; and the base conditions for the SPFs to predict total pedestrian crashes and FS pedestrian crashes on roadway segments are provided in Table  92 and Table  93, respectively. In Table  92 and Table  93, the combined models for four-lane undivided and divided roads from Table 87, Table 88, and Table 89 are provided as separate models for 4U and 4D.
From page 155...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 155   3.4.1.1.2.2 AFs for Total Pedestrian Crashes on Two-Lane Undivided Roads (Expanded Model) AF1p,r 2 Presence of Sidewalk Buffer Adjustment factors applicable to the expanded model for total pedestrian crashes on two-lane undivided roads, accounting for the presence of a sidewalk buffer, are presented in Table 95.
From page 156...
... 156 Pedestrian and Bicycle Safety Performance Functions 3.4.1.1.2.4 AFs for Total Pedestrian Crashes on One-Way Roads (Expanded Model) AF1p,r 2 Number of Lanes Adjustment factors applicable to the expanded model for total pedestrian crashes on one-way roads, accounting for the number of lanes, are presented in Table 96.
From page 157...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 157   Suburban Arterials)
From page 158...
... 158 Pedestrian and Bicycle Safety Performance Functions Based on this assessment, the expanded model for total pedestrian crashes (i.e., all severity levels combined) on two-lane undivided roads may or may not be compatible with the existing HSM Part C model for urban and suburban two-lane undivided roads at lower AADTs.
From page 159...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 159   For this road type, predicted average pedestrian crashes are similar for both models. Based on this assessment, the expanded model for total pedestrian crashes (i.e., all severity levels combined)
From page 160...
... 160 Pedestrian and Bicycle Safety Performance Functions Comparisons of the expanded models with existing models in HSM Part C illustrate potential differences in future editions of the HSM if the models from this research are integrated into the HSM and potential compatibility issues with existing HSM models for motor vehicle crashes, excluding pedestrian and bicycle crashes. 3.4.1.2.1 Bicycle Models (Conventional Output)
From page 161...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 161   Figure 23. Graphical representation of the SPF for predicted average total bicycle crashes per year on two-lane undivided roads (reduced model)
From page 162...
... 162 Pedestrian and Bicycle Safety Performance Functions Figure 24. Graphical representation of the SPF for predicted average total bicycle crashes per year on two-lane undivided roads (expanded model)
From page 163...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 163   Figure 25. Graphical representation of the SPF for predicted average total bicycle crashes per year on four-lane undivided and divided roads (reduced model)
From page 164...
... 164 Pedestrian and Bicycle Safety Performance Functions Figure 26. Graphical representation of the SPF for predicted average total bicycle crashes per year on four-lane undivided and divided roads (expanded model)
From page 165...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 165   one or more of the exposure measures (i.e., AADT or AADB) was counterintuitive or not statistically significant.
From page 166...
... 166 Pedestrian and Bicycle Safety Performance Functions Each component of Equation 3-12 (i.e., Nbr, Npedr, and Nbiker) is estimated separately to obtain the estimated total crash frequency for an individual roadway segment, and then a calibration factor is applied.
From page 167...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 167   In Table 104, the combined models for four-lane undivided and divided roads (from Table 100 and Table 101) are provided as separate models for four-lane undivided and divided roads.
From page 168...
... 168 Pedestrian and Bicycle Safety Performance Functions AF2b,r 2 Average Lane Width The AF applicable to the expanded model for total bicycle crashes on two-way undivided roads, accounting for the average lane width is calculated as follows: (3-23) .AF LW 0 05812exp,b r2 #= - -b` j l where: LW = average lane width (ft)
From page 169...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 169   AF2b,r 2 Number of Alcohol Sales Establishments The AF applicable to the expanded model for total bicycle crashes on four-lane undivided and four-lane divided roads, accounting for the number of alcohol sales establishments located near the roadway segment is calculated as follows: (3-26) .ASEAF 0 018exp,b r2 #= ` j where: ASE = number of alcohol sales establishments within 1,000 ft of the center of the roadway segment.
From page 170...
... 170 Pedestrian and Bicycle Safety Performance Functions Figure 29 provides a comparison of the predicted average total bicycle crash frequency per year from the expanded two-lane undivided roads model and the predicted average total bicycle crash frequency and total crashes (i.e., multiple-vehicle + single-vehicle + pedestrian + bicycle crashes) from the existing HSM Part C model for urban two-lane undivided roads.
From page 171...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 171   compared to the existing HSM model for roads with higher bicycle volumes. For roads with low bicycle volumes, the results of the two models are similar in magnitude.
From page 172...
... 172 Pedestrian and Bicycle Safety Performance Functions There is no existing HSM model for urban and suburban one-way roads for comparison purposes with the expanded model for total bicycle crashes (i.e., all severity levels combined) for one-way roads, although it is anticipated that HSM2 will include new models for this road type.
From page 173...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 173   by the same models converted to a form more suitable for inclusion in the HSM. For each intersection type, three levels of models are presented as appropriate: • A reduced model to estimate total pedestrian crashes (i.e., all severity levels combined)
From page 174...
... 174 Pedestrian and Bicycle Safety Performance Functions suspected serious injury crashes) is provided because either the models did not converge or the coefficient(s)
From page 175...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 175   No expanded model to predict total pedestrian crashes (i.e., all severity levels combined) or FS pedestrian crashes (i.e., fatal and suspected serious injury crashes)
From page 176...
... 176 Pedestrian and Bicycle Safety Performance Functions associated p-values. e table also shows the inverse of the overdispersion parameter and its standard error, the 2 x log-likelihood at convergence, the total number of crashes, and the total number of sites associated with the model.
From page 177...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 177   overdispersion parameter and its standard error; the 2 x log-likelihood at convergence; the total number of crashes; and the total number of sites associated with the model. In addition to exposure measures for motor vehicles and pedestrians, significant predictors of total pedestrian crashes at four-leg signal control intersections with two-way/two-way operations include: • Prohibition of right-turn-on-red.
From page 178...
... 178 Pedestrian and Bicycle Safety Performance Functions factor. Using existing HSM notation and variable definitions as appropriate, Equation 3-28 and Equation 3-29 present the predictive models for intersections from the HSM Part C, Chapter 12 on urban and suburban arterials but slightly rearranged to properly integrate the new pedestrian and bicycle SPFs into the equations.
From page 179...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 179   and four-leg stop control intersections with two-way/two-way operations from Table 109 is provided as separate models. Similarly, the combined model for total pedestrian crashes (i.e., all severity levels combined)
From page 180...
... 180 Pedestrian and Bicycle Safety Performance Functions AF3p,i 2 Number of Alcohol Sales Establishments The AF applicable to the expanded model for total pedestrian crashes at four-leg signal control intersections with two-way/two-way operations, accounting for the number of alcohol sales establishments within 1,000 ft of the center of the intersection is as follows: (3-32) .AF ASE 0 0189exp,p i3 #= ` j where: ASE = Number of alcohol sales establishments within 1,000 ft of the center of the intersection.
From page 181...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 181   frequencies for the 60/40 scenario are basically equivalent to estimates for the 85/15 scenario up to a total entering-intersection volume of 22,000 veh/day. As pedestrian crossing volumes increase above 100 ped/day, the new model predicts significantly higher pedestrian crash frequencies compared to estimates from the existing HSM model; and at a total entering- intersection traffic volume (AADTtotal)
From page 182...
... 182 Pedestrian and Bicycle Safety Performance Functions (i.e., all severity levels combined) at four-leg stop control intersections with two-way/two-way operations does not appear compatible with the existing HSM Part C model for urban and suburban four-leg stop control intersections.
From page 183...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 183   There is no existing HSM model for four-leg signal control intersections with one-way/two-way operations so no comparison is made here between the reduced model for total pedestrian crashes (i.e., all severity levels combined) at four-leg signal control intersections with one-way/ two-way operations and an existing HSM model.
From page 184...
... 184 Pedestrian and Bicycle Safety Performance Functions signal control intersections with two-way/two-way operation may be suitable for integration into the HSM for use with four-leg signal control intersections, but calibration of the models will likely be critical. 3.4.2.2 Bicycle Models e nal bicycle SPFs for estimating bicycle crashes at intersections by intersection type are shown rst in tabular form presenting conventional statistical output results, followed by the same models converted to a form more suitable for inclusion in the HSM.
From page 185...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 185   Table 116 presents a reduced model for total bicycle crashes (i.e., all severity levels combined) at three-leg and four-leg stop control intersections with two-way/two-way operations that primarily includes exposure variables.
From page 186...
... 186 Pedestrian and Bicycle Safety Performance Functions bicycle crashes (i.e., fatal and suspected serious injury crashes) is provided because either the models did not converge or the coefficient(s)
From page 187...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 187   Figure 43 graphically presents the SPF shown in Table 119 for various motor vehicle entering (AADTtotal) and bicycle crossing (AADBcrossing)
From page 188...
... 188 Pedestrian and Bicycle Safety Performance Functions Figure 42. Graphical representation of the SPF for predicted average total bicycle crashes per year at four-leg signal control intersections with two-way/two-way operations (reduced model)
From page 189...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 189   factor. Equation 3-28 and Equation 3-29 present the predictive models for intersections from the HSM Chapter 12 on urban and suburban arterials but slightly rearranged to properly integrate the new pedestrian and bicycle SPFs into the equations (see Section 3.4.2.1.2.)
From page 190...
... 190 Pedestrian and Bicycle Safety Performance Functions AADBcrossing = sum of daily bicycle volumes crossing all intersection legs (ped/day)
From page 191...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 191   AF2b,i 2 Type of Left-Turn Signal Phasing Adjustment factors applicable to the expanded model for total bicycle crashes at four-leg signal control intersections with two-way/two-way operations, accounting for type of left-turn signal phasing are presented in Table 122. AF3b,i 2 Number of Schools The adjustment factor applicable to the expanded model for total bicycle crashes at fourleg signal control intersections with two-way/two-way operations, accounting for number of schools within 1,000 ft of the center of the intersection is as follows: (3-35)
From page 192...
... 192 Pedestrian and Bicycle Safety Performance Functions Based on the comparison, at lower bicycle crossing volumes around 500 ped/day or less, the new model predicts bicycle crash frequencies similar to estimates from the existing HSM model. Although it is not clearly visible in Figure 44, the predicted average total bicycle crash frequencies for the 60/40 scenario are basically equivalent to estimates for the 85/15 scenario up to a total entering-intersection volume of 22,000 veh/day.
From page 193...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 193   (i.e., all severity levels combined) at four-leg stop control intersections with two-way/two-way operations appears compatible with the existing HSM Part C model for urban and suburban four-leg stop control intersections, although calibration of the models will likely be critical.
From page 194...
... 194 Pedestrian and Bicycle Safety Performance Functions Figure  47 provides a comparison of the predicted total bicycle crash frequency from the expanded model for total bicycle crashes (i.e., all severity levels combined) at four-leg signal control intersections with two-way/two-way operation and the predicted average total bicycle crash frequency and total crashes (i.e., multiple-vehicle + single-vehicle + pedestrian + bicycle crashes)
From page 195...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 195   adjust predictive models which were developed with data from one jurisdiction for application in another jurisdiction. Calibration provides a means to account for differences between jurisdictions in various factors such as climate, driver populations, and crash reporting thresholds.
From page 196...
... 196 Pedestrian and Bicycle Safety Performance Functions – Four-lane divided roads (4D) – One-way roads (OW)
From page 197...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 197   Reduced models (total crashes) 2U −5.214 0.327 0.224 1.267 Yesa No 4U −11.154 0.902 0.236 1.855 Yes No 4D −12.497 0.902 0.236 1.855 Yes No OW −10.651 0.829 0.337 1.513 Yes No Expanded models (total crashes)
From page 198...
... Reduced models (Total crashes) 2U −9.647 0.546 0.493 2.873 Yes No 4U −14.935 0.996 0.691 1.111 Yes No 4D −15.696 0.996 0.691 1.111 Yes No OW −10.388 0.446 0.749 0.002 Yes No Expanded models (Total crashes)
From page 199...
... Development of Pedestrian and Bicycle Models Incorporating Available Pedestrian and Bicyclist Exposure Data 199   Intersection Type Intercept ( ) AADTtotal ( )
From page 200...
... 200 Pedestrian and Bicycle Safety Performance Functions – Number of bus/transit stops within 1,000  ft of the center of the roadway segment (Equation 3-17)

Key Terms



This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More information on Chapter Skim is available.