Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

88 A P P E N D I X G The following document was used as a guide for the completion of the database and to record/ code observed site characteristic data on relevant treatment and comparison site features. The most recently available Google Earth Satellite overhead imagery and street-view photographs were used to the extent possible for selecting sites, where needed, and observing the relevant treat- ment or comparison site standards, characteristics, and measures. Additionally, QC/Kittelson photo graphs taken during the project’s 2014 on-site pedestrian volume counts were used to verify, update, and augment site characteristic data captured from Google Earth imagery. As the data collection process continued, new columns were added for additional data and site charac- teristics required. Separate databases were created for treatment and comparison sites. Site Characteristics/Feature Coding Instructions Features List 1. Crosswalks a. no marked crosswalk b. white standard crosswalk c. yellow crosswalk d. staggered crosswalk e. ladder crosswalk f. zebra crosswalk g. piano crosswalk h. continental crosswalk i. double continental crosswalk j. diagonal crosswalk k. brick/block/stamped style crosswalk l. marked crosswalk unknown type m. raised crosswalk 2. Signs a. crosswalk warning sign b. children playing sign c. bicycle crossing sign d. other supplemental warning signs (including overhead signs) e. flashing beacon f. in-street sign g. school zone indication (sign or pavement marking) h. advisory speed signs i. driver speed feedback sign j. flashing LED border sign NCHRP Project 17-56 Database Creation and Data Entry Methodology Notes

NCHRP Project 17-56 Database Creation and Data Entry Methodology Notes 89 3. Other Features a. school crossing guard b. markings for crosswalk ahead warning c. curb extension d. in-pavement warning lights e. post mounted delineator f. painted/flush median g. nearby traffic calming (e.g., speed humps) Table G-1. Treatment sites methodology by column. (continued on next page) Column Item Description A Study ID Study ID designation assigned by the project team. Coding for treatment sites: the first three letters of the city/agency name followed by a three-digit ID number for the location (e.g., intersection) followed by a letter to denote the particular crossing. o Example: Phoenix site - PHO001A Crash data were collected and recorded in a different database and according to the applicable StudyID. B City Location city for the site. C Status (definite, verify) As the data collection process continued, some sites were excluded and some were added based on study requirements (e.g., installation dates 2010 or later, invalid intersection geometry, treatments no longer exist, unreliable or no crash or AADT data, could not collect ped-volume counts due to construction or other issues, unusual characteristics). The letter “D” denotes a confirmed site that has been fully verified as applicable for the study and has all required data available. Filter this column for “D” to observe all valid study sites. D County The county jurisdiction in which the site is located (necessary for contacting agencies for AADT, crash. and other data at the site). E Place, town, or other possible jurisdiction The place, township, or other jurisdictional level in which the site is located (necessary for contacting agencies for AADT, crash, and other data at the site). F Intersection Name The current street names at the intersection or the main road only at a midblock location. G Midblock or intersection M-Midblock I- Intersection H Site has ped- volume count Y- Yes N- No I Site has AADT Y- Yes N- No J Years of crash data available The range of years for which crash data are available by city. K Transit/bus stop Y- Yes N- No When selecting comparison sites or recording treatment and comparison site characteristics with Google Earth imagery, if the crossing was in close proximity to a transit or bus stop, the column was marked “Y.” L Bus stop/route name The number or name of the bus stop or route, as provided by Google Earth. M Street lighting Y- Yes N- No N Unusual characteristics (geometry etc…) Y- Yes N- No Treatment sites were screened for intersections that had unusual geometries or more than four legs, and if so they were marked with a “Y,” which eliminated them from the study. O Road 1 Name 1 Name(s) or route(s) of road being crossed. P Road 1 Name 2 Secondary name(s) or route(s) of road being crossed (if applicable). Q Road 2 Name 1 Name(s) or route(s) of intersecting road, if treatment is installed at an intersection. R Road 2 Name 2 Secondary name(s) or route(s) of intersecting road, if treatment is installed at an intersection (if applicable).

90 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Table G-1. (Continued). Column Item Description Y Refuge island Y- Yes N- No Z Continuous median Y- Yes N- No AA Refuge island installation year If refuge island is present, year when refuge island was installed. If no refuge island was installed at this location, enter n/a. Code as the year of installation (e.g., “2006”) or a range of years (“2006- 2008”) if the exact year is unknown. When not sure when treatment originally installed, put in earliest known year according to Google Earth aerial imagery. AB Refuge island treatment leg of intersection The leg of the intersection where the RI treated crosswalk is located, denoted by cardinal direction. If RIs are located at two legs of the crossing, these are noted as cardinal directions. If site is a midblock treatment, or no RI is at site, entered “n/a.” AC RI width (ft) The RI width in feet using the Google Earth aerial ruler tool. AD RI length (ft) If the RI is located on a non-continuous median, the length of the island in feet is recorded using the Google Earth aerial ruler tool. AE Refuge island installation notes Any special notes about the installation of the island. This only applies to characteristics of the island, not other countermeasures that were installed with it. That is handled in another variable. Variables noted were Raised island Cut-out Diagonal cut-out Flushed median Staggered cut-out Etc…. AF Advanced stop/yield line presence Y- Yes N- No AG Advanced stop installation year If advance stop or yield line is present, year when advance line was installed according to city or Google aerial imagery. If no advance line was installed at this location, enter “n/a.” AH Advanced stop/yield line type Indication whether drivers are instructed to stop or yield, based on local statute or supplementary sign indication. Coding: S = Stop Y = Yield AI Advanced stop/yield sign type Determined using Google Earth street-view or QC/Kittelson ped-volume photographs No sign Stop sign Yield sign “n/a” if no AS line is present AJ AS treatment leg of intersection The leg of the intersection where the AS line treated crosswalk is located, denoted by cardinal direction. If AS lines are associated with crosswalks at two legs of the intersection, these are noted as cardinal directions. If site is a midblock treatment or no AS line is at site, entered “n/a.” S Coordinates Latitude/longitude coordinates of the crossing location, in decimal degrees. T Google map link Hyperlink to the location of the crossing in Google Maps. U Baseline condition (NOTES) A text description of the baseline condition of the crossing location before it received the safety treatment (or at time of treatment installation). Often includes a description of the type of crosswalk. No notes if the crosswalk was present prior to the year 2000. V Baseline condition The feature code designation of the baseline condition of the crossing location before it received the safety treatment (or at time of treatment installation). Often includes the feature code for the type of crosswalk. If no pavement marking or crosswalk was present before the treatment was installed, then the coding “1A” for “no marked crosswalk” was used. W Earliest known year of baseline condition The year that the crosswalk was visible on Google Earth imagery. If the crosswalk was visible prior to the year 2000, the year was noted as “pre- 2000.” X Condition before according to city If any features were present before the earliest treatment installation according to the city (not Google Earth imagery), a description and feature code was noted here.

NCHRP Project 17-56 Database Creation and Data Entry Methodology Notes 91 Table G-1. (Continued). Column Item Description AP Other measures installed with HAWK or other pedestrian improvement (according to city) Any other items installed with or as part of the PHB according to the city. AQ RRFB presence Y- Yes N- No AR RRFB installation Year If RRFB is present, year when RRFB was installed. If no RRFB was installed at this location, enter “n/a.” AS RRFB treatment leg of intersection The leg of the intersection where the RRFB treated crosswalk is located, denoted by cardinal direction. If RRFBs are associated with crosswalks at two legs of the intersection, these are noted as cardinal directions. If site is a midblock treatment, or no RRFB is at site, entered “n/a.” AT RRFB installation notes Any special notes about the installation of the RRFB. This only applies to characteristics of the RRFB, not other countermeasures that were installed with it. That is handled in another variable. AU Other countermeasures installed with primary project Any other countermeasures or modifications installed at this location with the primary treatment, as reported by city (not observed by Google Earth). AV Other countermeasures installation year Year of installation of the other countermeasures installed along with primary treatment. AW Other pedestrian safety devices found at location Features without designated feature codes. AX Crossing guard Y- Yes N- No If a crossing guard at site crosswalk was reported by city or seen on Google street-view photos. AY School crossing Y- Yes N- No If indications of a school zone were seen on Google aerial imagery, street- view photos, QC/Kittelson ped-vol count site photos, or as reported by the city. AZ Reason for safety installations If known, reason why the site was selected for safety installations as reported by city. If not, marked “unknown.” BA Other pedestrian safety devices found at location (current) Pedestrian safety devices with designated feature codes that were noted as currently being present at the site by Google street-view photos or QC/Kittelson ped-volume count site photos. Uses feature list codes. AK Advanced stop/yield line notes Any special notes about the installation of the advance line. This only applies to characteristics of the advance line, not other countermeasures that were installed with it. That is handled in another variable. Variables noted were Shark tooth Solid line “n/a” if no AS line is present AL PHB presence Y- Yes N- No AM PHB installation year If PHB is present, year when PHB was installed. If no PHB was installed at this location, enter “n/a.” AN PHB treatment leg of intersection The leg of the intersection where the PHB treated crosswalk is located, denoted by cardinal direction. If PHBs are associated with crosswalks at two legs of the intersection, these are noted as cardinal directions. If site is a midblock treatment, or no PHB is at site, entered “n/a.” AO PHB installation notes Any special notes about the installation of the PHB. This only applies to characteristics of the PHB, not other countermeasures that were installed with it. That is handled in another variable. Example 1: “Supplemental sign to clarify what driver should do on flashing red.” (continued on next page)

92 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Table G-1. (Continued). Column Item Description BQ One-way vs two- way Indicates if the road being crossed is one- or two-way traffic. BR Notes Includes any notes (center turn lane, bike lanes, or interesting and unique features at the site in regard to geometry, crossing, characteristics, history, and other features). BS Two-stage crossing installation year If a two-stage crossing is present, indicates the year of installation as determined by Google Earth aerial imagery or reported by city. Two-stage crossing has an RRFB, PHB, or flashing LED border sign and the RI is staggered. Each stage has a pedestrian push-button activation for the flashing LED border sign. BT AADT major rd site ID If a site ID associated with the study site location is provided with the source city’s AADT data, it is noted here. If not, then it is marked “not provided.” BU AADT data source This is the jurisdiction that provided the AADT data or from which they were found. BV Possible anomalous AADT data If the only data available for the site were estimates from the data source contacts or from the source data, or if there were unusual characteristics about the data, it is noted here. If not, it is marked “no.” BW AADT count timeframe This indicates the type of count as indicated by the data source city (e.g., AADT, 24-hour count, average daily count, 24-hour weekday, etc.) BX-CK Major rd AADT/major rd AADT year This is the AADT count on the road being crossed and the date of the count (if specific day and month are provided, they are noted). Data are recorded from most recent to oldest. CJ Minor rd site ID If the site was at an intersection and the minor road(s) (road not being crossed) has AADT available, the site ID is noted here. If no data at the minor road are available, it is marked “none available.” CK-CR Minor rd AADT/minor rd AADT year This is the AADT count on the road not being crossed and the date of the count (if specific day and month are provided, they are noted). Data are recorded from most recent to oldest. CS QC ped count total The pedestrian volume count from the QC/Kittelson on-site video counts. CT QC number of hours counted The number of hours that QC/Kittelson counted pedestrians on site (1-2). CU QC count start time The time of day the QC/Kittelson on-site video count started. CV QC count date The date that QC/Kittelson conducted the on-site video count. CW QC count notes Any notes from QC/Kittelson about the count (usually to explain why a count could not be conducted, i.e., construction, no access, site now signalized, treatment removed or changed, etc.) CX-CY Ped volume (not QC)/ped volume year (not QC) If there was a previous pedestrian volume count available, then QC/Kittelson did not conduct an on-site video count and the previous count data were recorded here, along with the year of the count (if day and month were available they were also noted). BB Other pedestrian safety device year & source The year that the safety device was observed on streetview (SV) and QC/Kittelson photos (e.g., SV-2012 & QC-2014). BC-BJ Subsequent change/year of change The oldest to newest observed change and year of change of site characteristics and features coded in the features list or noted as otherwise. BK Speed limit of road being crossed The marked or indicated speed limit of the road being crossed by the treatment. As reported by city or observed on Google street-view. If no speed limit is observed or reported, marked “unknown.” BL Number of lanes on road being crossed This is the total number of lanes (including turn lanes) that the crosswalk or crossing goes through. Determined from Google Earth aerial imagery. (If there was a center or right/left turn lane, it was noted in the notes column.) BM Crosswalk length in feet As measured with the Google Earth ruler tool, this is the total distance in feet of the crossing. BN Presence of median on road being crossed Y- Yes N- No Indicates if a median is or is not present at the crossing. BO Crosswalk type Using the feature list coding, indicates what type of crosswalk is currently present at the crossing. If no crosswalk or crosswalk markings are present, the code “1A” is used. BP Urban vs suburban Indicated if site is located in urban or suburban area type. (Urban was marked if the site is within the city CBD; or by use of Google Earth aerial imagery is located in a dense and urban zone.)

NCHRP Project 17-56 Database Creation and Data Entry Methodology Notes 93 Table G-1. (Continued). CZ Number of hours counted The duration in number of hours of the previous count. DA Count times The time of day of the previous count. DB Projected expanded pedestrian count ADT 12hr* Expanded total pedestrian count ADT for all 1-2 hr counts attained by QC/Kittelson on-site counts. Used calculated seasonal expansion factors from analysis of Charlotte data and other studies. (Note: previous 12 hour counts from City of St. Petersburg were not expanded.) DC Projected expanded pedestrian count ADT 24hr** Expanded total pedestrian count for column DB total to projected expanded 24-hr pedestrian volume count ADT. DD-DH Install in 2010- 2014 FALSE TRUE Filter cells to indicate if any of the four treatments were installed in 2010- 2014. If so, these sites were eliminated from the study. Column Item Description Table G-2. Comparison sites methodology by column. Column Item Description A Study ID Study ID designation assigned by the project team. Coding for comparison sites: the first three letters of the city/agency name, followed by “C” (to indicate that it is a comparison site), followed by a three-digit ID number for the location (e.g., intersection). Example: Phoenix comparison site - PHOC001. Crash data were collected and recorded in a different database and according to the applicable StudyID. B City Location city for the site. C Status (definite, verify) As the data collection process continued, some sites were excluded and some were added based on study requirements (e.g., invalid intersection geometry, newly installed treatment at site, site has raised median island, unreliable or no crash or AADT data, could not collect ped-volume counts due to construction or other issues, unusual characteristics). The letter “D” denotes a confirmed site that has been fully verified as applicable for the study and has all required data available. Filter this column for “D” to observe all valid study sites. D County The county jurisdiction in which the site is located (necessary for contacting agencies for AADT, crash and other data at the site). E Place, town or other possible jurisdiction The place, township, or other jurisdictional level in which the site is located (necessary for contacting agencies for AADT, crash and other data at the site). F Midblock or intersection M-Midblock I- Intersection G Untreated raised median Y- Yes N- No If the site had a untreated raised median island and the crossing, it was invalidated as a comparison site and eliminated from the study. H Site has ped- volume count Y- Yes N- No I Site has AADT Y- Yes N- No J Years of crash data available The range of years for which crash data is available by city. K Transit/bus stop Y- Yes N- No When selecting comparison sites or recording treatment and comparison site characteristics with Google Earth imagery, if the crossing was in close proximity to a transit or bus stop, the column was marked “Y.” L Bus stop/route name The number or name of the bus stop or route, as provided by Google Earth. M Street lighting Y- Yes N- No (continued on next page)

94 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Table G-2. (Continued). AR-BE Major rd AADT/major rd AADT year This is the AADT count on the road being crossed and the date of the count (if specific day and month are provided, they are noted). Data are recorded from most recent to oldest. BF Minor rd site ID If the site was at an intersection and the minor road(s) (road not being crossed) has AADT available, the site ID is noted here. If no data at the minor road are available, it is marked “none available.” Column Item Description AE Speed limit of road being crossed The marked or indicated speed limit of the road being crossed by the treatment. As reported by city, or observed on google street-view. If no speed limit is observed or reported, marked “unknown.” AF Number of lanes on road being crossed This is the total number of lanes (including turn lanes) that the crosswalk or crossing goes through. Determined from Google Earth aerial imagery. (If there was a center or right/left turn lane, it was noted in the notes column.) AG Crosswalk length in feet As measured with the Google Earth ruler tool, this is the total distance in feet of the crossing. AH Crosswalk type Using the feature list coding, indicates what type of crosswalk is currently present at the crossing. If no crosswalk or crosswalk markings are present, the code “1A” is used. AI Urban vs suburban Indicated if site is located in urban or suburban area type. (Urban was marked if the site is within the city CBD; or by use of Google Earth aerial imagery is located in a dense and urban zone.) AJ One-way vs two- way Indicates if the road being crossed is one- or two-way traffic. AK School crossing Y- Yes N- No If indications of a school zone were seen on Google aerial imagery, street- view photos, QC/Kittelson ped-vol count site photos, or as reported by the city. AL Name of school near site If the name of the school could be determined from Google imagery or reported by city, it was noted here. If not, “n/a” is marked. AM Notes Includes any notes (bike lanes or interesting and unique features at the site in regards to geometry, crossing, characteristics, history, and other features). AN AADT major rd site ID If a site ID associated with the study site location is provided with the source city’s AADT data, it is noted here. If not, then it is marked “not provided.” AO AADT data source This is the jurisdiction that provided the AADT data or from which the data were found. AP Possible anomalous AADT data If the only data available for the site were estimates from the data source contacts or from the source data itself, or if there were unusual characteristics about the data, it is noted here. If not, it is marked “no.” AQ AADT count timeframe This indicates the type of count as indicated by the data source city. (e.g., AADT, 24-hour count, average daily count, 24-hour weekday, etc.). N Unusual characteristics (geometry, etc.) Y- Yes N- No Comparison sites were screened for intersections that had unusual geometries or more than four legs, and if so they were marked with a “Y,” which eliminated them from the study. O Road 1 Name 1 Name(s) or route(s) of road being crossed. P Road 1 Name 2 Secondary name(s) or route(s) of road being crossed (if applicable). Q Road 2 Name 1 Name(s) or route(s) of intersecting road, if treatment is installed at an intersection. R Road 2 Name 2 Secondary name(s) or route(s) of intersecting road, if treatment is installed at an intersection (if applicable). S Coordinates Latitude/longitude coordinates of the crossing location, in decimal degrees. T Google map link Hyperlink to the location of the crossing in Google maps. U Other pedestrian safety devices found at location (current) Pedestrian safety devices with designated feature codes that were noted as currently being present at the site by Google street-view photos or QC/Kittelson ped-volume count site photos. Uses feature list codes. V Other pedestrian safety device year & source The year that the safety device was observed on streetview (SV) and QC/Kittelson photos (e.g., SV-2012 & QC-2014). W-AD Subsequent change/year of change The oldest to newest observed change and year of change of site characteristics and features coded in the features list or noted as otherwise.

NCHRP Project 17-56 Database Creation and Data Entry Methodology Notes 95 Pedestrian 24-Hour ADT Expansion Method Step 1: Expansion to Projected 12-Hour Counts 1—All sites without previous count data provided by the city had in-person pedestrian volume counts conducted by Kittelson during 2014. These were all 1- or 2-hour counts, with start times at 1500–1700, and during June–November 2014. The seasonal expansion factors (Table E-3 in Appendix E) were used to expand these counts to projected 12-hour counts. Two-hour counts were first halved to get the 1-hour volume and then multiplied by the appropriate expansion factor to get the estimated 12-hour volume. 2—There were 28 treatment sites from St. Petersburg, FL, and 83 comparison sites from Charlotte, NC, that had previous 12-hour counts. The most recent year’s count was used, and these were not adjusted. Step 2: Expansion to Projected 24-Hour Pedestrian ADT 1—Adjustment factors by time of day and area type used to obtain estimated pedestrian ADT from Table 12 (p. 67) of Zegeer, C. V., Stewart, J. R., Huang, H. H., Lagerwey, P. A., Feaganes, J., & Campbell, B. J. (2005). Safety Effects of Marked versus Unmarked Crosswalks at Uncontrolled Locations: Final Report and Recommended Guidelines (No. FHWA-HRT-04-100) were used to expand all 12-hour adjusted and non-adjusted counts to estimated 24-hour pedestrian ADTs. The 12-hour count columns were divided by the 0.86 factor used in the Zegeer et al. (2005) report to arrive at the adjusted 24-hour pedestrian ADT numbers. Table G-2. (Continued). BG-BN Minor rd AADT/minor rd AADT year This is the AADT count on the road not being crossed and the date of the count (if specific day and month are provided, they are noted). Data are recorded from most recent to oldest. BO QC ped count Total The pedestrian volume count from the QC/Kittelson on-site video counts. BP QC number of hours counted The number of hours that QC/Kittelson counted pedestrians on site (1-2). BQ QC count start time The time of day the QC/Kittelson on-site video count started. BR QC count date The date that QC/Kittelson conducted the on-site video count. BS QC count notes Any notes from QC/Kittelson about the count (usually to explain why a count could not be conducted, i.e., construction, no access, site now signalized, treatment removed or changed, etc.). BT-CB Ped volume (not QC)/number of hours Counted/ped volume year (not QC) If there was a previous pedestrian volume count available, then QC/Kittelson did not conduct an on-site video count, and the previous count data were recorded here, along with the number of hours counted and the year of the count (if day and month were available they were also noted). CC Count times The time of day of all previous counts. CD Projected expanded pedestrian count ADT 12hr Expanded total pedestrian count ADT for all 1-2 hr counts attained by QC/Kittelson on-site counts. Used calculated seasonal expansion factors from analysis of Charlotte data and other studies. (Note: previous 12 hour counts from City of Charlotte were not expanded.) CE Projected expanded pedestrian count ADT 24hr Expanded total pedestrian count for column DB total to projected expanded 24-hr pedestrian volume count ADT. Column Item Description