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.

96 A P P E N D I X H Introduction This appendix reviews and summarizes research studies that have evaluated the safety effects of selected types of pedestrian crossing treatments at unsignalized crossing locations. Specifi- cally, this appendix is the result of Task 2 of NCHRP Project 17-56, “Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments.” The following types of pedestrian treatments are included: • Unsignalized pedestrian crosswalk signs and pavement markings, including advanced yield or stop markings and signs; • High-visibility crosswalk marking patterns; • High-intensity activated crosswalk (HAWK) signals; • Rectangular rapid flashing beacons (RRFBs); • In-pavement warning lights; • Pedestrian refuge areas; • Curb extensions, and • Raised pedestrian crosswalks. Although NCHRP Project 17-56 focused on the crash modification factors (CMFs) of those pedestrian treatments, this literature review includes not only crash-based studies, but also studies which investigated behavioral and operational measures of effectiveness (e.g., pedestrian/vehicle conflicts, vehicle speeds, driver yielding behavior). This is because there are very few research studies that have been able to use pedestrian crashes in their evaluation. Specifically, sample size issues may arise since pedestrian crashes usually do not cluster in large numbers at a single site; thus, a larger number of treatment and comparison sites may be needed to get statistical signifi- cance compared to evaluation of countermeasures for vehicle/vehicle crashes. Also, studies that use behavioral measures may at least be useful in gaining insights into potential safety behaviors by motorists and pedestrians that result from the treatments. However, the analysis performed in NCHRP Project 17-56 is strictly based on crash effects (CMFs) of the selected treatments. The following is a summary of available research for each of the eight pedestrian treatments. At the end of this appendix, a set of summary tables provides an overview of the studies and results of the research for each of the eight treatments of interest, which are listed above. The primary source used to compile this literature review was a UNC HSRC produced draft report of the Evaluation of Pedestrian-Related Roadway Measures: A Summary of Available Research by Mead, Zegeer, and Bushell1. Selected sections from this resource were extracted or adopted for use in this literature review. Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 1Mead, J., C. Zegeer, and M. Bushell (2013). Evaluation of Pedestrian-Related Roadway Measures: A Summary of Available Research. Chapel Hill, North Carolina: Federal Highway Administration.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 97 Treatment 1. Unsignalized Pedestrian Crosswalk Signs and Pavement Markings, including Advanced Yield or Stop Markings and Signs Signing In-Street Pedestrian Signs An early version of in-street pedestrian signs was studied as part of a 2000 report by Huang, Zegeer, Nassi, and Fairfax published by the Federal Highway Administration. The treatment con- sisted of pedestrian safety cones with the message “State Law—Yield to Pedestrians in Crosswalks in Your Half of the Road,” which were installed in New York State and Portland, Oregon. The cone was developed in 1996 and designed to be placed in the middle of a crosswalk. The use of the cones was evaluated at six sites in New York State and one site in Portland, Oregon. Pre- and post-treatment data were collected for seven sites, and the following measures of effectiveness (MOEs) were used: (1) percentages of pedestrians for whom motorists yielded, (2) percentage of motorists who yielded to pedestrians, (3) percentage of pedestrians who hesitated, rushed, or aborted in crossing, and (4) percentage of pedestrians crossing in the crosswalk. Of the three treatments that were evaluated in the report, pedestrian safety cones were the most successful in increasing the percentage of yielding drivers. When all study sites were combined, motorist yielding increased from 69.8 percent pre-treatment to 81.2 percent post-treatment, which is sig- nificant at the 0.001 level. Pedestrians who ran, aborted, or hesitated decreased, but the decrease was not statistically significant. The authors concluded that pedestrian safety cones were gener- ally effective in increasing the percentages of pedestrians for whom motorists yielded (1). A 1999 article sponsored by the Federal Highway Administration described the 2-year eval- uation of then-experimental in-street yield-to-pedestrian signs installed at various locations throughout the city of Madison, Wisconsin, in 1997. Three test sites were used in the first year of Figure H-1. Prototype in-street yield to pedestrians sign used in the 1990s. Source: pedbikeimages.org

98 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments the experiment, while five sites were used in the second year. The researchers were able to conduct before-after analysis at two of the sites and after-only analysis at two of the sites. The researchers used motorist yielding to pedestrians in the crosswalk as the measure of effectiveness (MOE). The proportion of drivers yielding was analyzed using a Z-test for proportions in the before-after study, and results indicated a statistically significant increase in yielding behavior at sign locations. Because of differences between test site geometry and results, the authors called for further research on the effectiveness of the signs. The City of Madison Traffic Engineering staff reported positive feedback from pedestrians regarding the signs, as well as citizen involvement in reporting damaged or missing signs at study locations (2). The following year, a 2000 report sponsored by the Iowa Department of Transportation summarized the results of placing in-street yield-to-pedestrian signs at three sites throughout Cedar Rapids, Iowa. The researchers measured vehicle speed, percentage of vehicles yielding to pedestrians, and percentage of failed or rushed crossings before and after the signs were installed. Results indicated that the presence of the signs had a positive effect on driver behavior, leading to speed reductions at one site and increased driver yielding at another. Similar to the Madison study, results were not uniform, with overall roadway configuration affecting driver behavior (3). A 2003 paper by Kamyab, Andrle, Kroeger, and Heyer discussed the effects of installing a removable pedestrian island and pedestrian crossing signs on a two-lane highway in rural Mahnomen County, Minnesota. Researchers collected pre- and post-treatment speed data to assess short and long-term effects of the treatments. Results showed a statistically significant reduction in mean speeds and increase in speed-limit compliance at the treatment site for both the long- and short term (4). With the goal of improving pedestrian safety in Pennsylvania, the Pennsylvania Department of Transportation organized a program to distribute Yield-to-Pedestrian Channelizing Devices (YTPCD) to interested municipalities. A 2006 report summarized a safety evaluation of the in-roadway yield-to-pedestrian signs installed at 21 midblock and intersection sites in four Pennsylvania cities. The researchers collected pre- and post-treatment driver and pedestrian behavior data at treatment and potential spillover sites. Driver yielding to pedestrians increased by 30 to 40 percent at intersections and by 17 to 24 percent at treated crosswalk sites. Pedestrian yielding to motorists decreased by 11 to 16 percent at intersections and by 8 to 13 percent at treated crosswalk sites. A statistically significant increase in pedestrians using the crosswalks Source: Kannel et al., 2000. (3) Figure H-2. Two types of impactable yield signs. Figure H-3. An in-street pedestrian sign used with a removable crossing island in Minnesota. Source: Kamyab et al., 2003. (4)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 99 was also observed. The researchers concluded that the signs were more effective at intersections than at midblock crossings, but that the in-roadway yield-to-pedestrian signs had an overall positive effect on increasing pedestrian safety. They also found that follow-up data collection was complicated by damaged, moved, or missing signs (5). A 2007 study by Banerjee and Ragland used video recordings to examine the changes in driver yielding rates as a result of impactable yield signs installed at three intersections in San Francisco. The researchers concluded that a large increase in yielding did occur following the installation of the signs. Figure H-5 shows increases in percentages of vehicles that yielded at each of the four sites (6). A 2007 report by Ellis, Van Houten, and Kim studied the effect of placing an in-roadway “Yield to Pedestrians” at different distances in advance of a crosswalk. Three marked crosswalks were chosen in Miami Beach, Florida, and baseline data about pedestrian and driver behavior were collected. To test optimal sign placement, signs were placed at one or all three of the crosswalks Table H-1. Mean driver speeds before and after the installation of an in-street pedestrian sign and removable pedestrian island. Observed Traffic Mean Speed (mi/h) t- statistic Significant (95%) Speed Compliance % t- statistic Significant (95%) Passenger Cars Before 1152 34.8 -- -- 31 -- -- After-1 1067 29.5 13.49 Yes 58 -12.80 Yes After-2 1331 30.7 11.05 Yes 51 -10.01 Yes Nonpassenger Cars Before 71 37.4 -- -- 24 -- -- After-1 46 28.8 4.11 Yes 65 -4.42 Yes After-2 60 29.5 4.01 Yes 57 -3.84 Yes All Vehicles Before 1237 35 -- -- 30 -- -- After-1 1113 29.5 14.20 Yes 58 -13.68 Yes After-2 1392 30.6 11.02 Yes 51 -10.85 Yes Source: Kamyab et al., 2003. (4) Source: pedbikeimages.org Figure H-4. In-roadway pedestrian sign installed mid-crosswalk.

100 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments on a rotating, random order either at the crosswalk itself, 20 feet in advance of the crosswalk, or 40 feet in advance of the crosswalk. A z-test for comparing proportions was utilized to analyze pedestrian and driver behavior as a result of the sign placement. While researchers determined that the presence of the signs alone was highly effective at increasing the percentage of drivers who yielded to pedestrians, the location and number of in-roadway signs placed in a crosswalk approach was not a critical factor in determining the magnitude of this outcome (7). Source: Banerjee and Ragland, 2007. (6) Figure H-5. Graph showing the percentage of yielding motorists before and after the installation of impactable yield signs in San Francisco, California. Figure H-6. Two placements of in-roadway pedestrian signs as tested in a Miami, Florida, evaluation study by Ellis et al., 2007. (7)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 101 A 2009 evaluation of the Pedsafe II project in San Francisco used video observation and inter- cept surveys to collect pre- and post-treatment data to evaluate the effectiveness of 13 counter- measures deployed at 29 sites throughout San Francisco, California. As part of the project, in-street “Yield-to-Pedestrian” signs were installed in the medians of uncontrolled crosswalks. At the four crosswalks where in-street pedestrian signs were installed, there was a significant increase in the percentage of yielding drivers, from 53 percent pre-treatment to 68 percent post-treatment. Of the 13 countermeasures that were tested, in-street “Yield-to-Pedestrian” signs were one of the six countermeasures considered the most effective in increasing pedestrian safety (8). A 2014 presentation by Bennett, Manal, and Van Houten at the Transportation Research Board conference evaluated the use of in-street pedestrian crossing signs in gateway configurations at locations in East Lansing, Michigan. The gateway configuration consists of one sign in the mid- dle of the roadway, and two signs installed in the gutter pans on each side of the roadway. Three conditions were alternated and evaluated at the two study sites: no in-street sign (baseline), one in-street sign in the median (typical configuration), and three in-street signs in the gateway con- figuration. For each data collection session, staged pedestrian crossings were conducted while research assistants measured motorist yielding behavior. At both sites, motorist yielding aver- aged 25 percent when no signs were present. The presence of one in-street sign was associated with motorist yielding of 57 percent at both locations. The gateway (three signs) configuration was associated with 79 percent and 82 percent at the two locations. Thus, the gateway configura- tion using three signs was associated with the highest motorist yielding rates (9). References 1. Huang, H., C. Zegeer, R. Nassi, and B. Fairfax. The Effects of Innovative Pedestrian Signs at Unsignalized Locations: A Tale of Three Treatments. Publication FHWA-RD-00-098, FHWA, U.S. Department of Transportation, Washington, D.C., 2000. 2. City of Madison Traffic Engineering Division. Year 2 Field Evaluation of Experimental ‘In-Street’ Yield to Pedestrian Signs. City of Madison Department of Transportation, Madison, Wisconsin, 1999. 3. Kannel, E. J., R. R. Souleyrette, and R. Tenges. In-Street Yield to Pedestrian Sign Application in Cedar Rapids, Iowa. Center for Transportation Research and Education, Iowa State University, Ames, Iowa, 2003. Source: Federal Highway Administration. Figure H-7. Two pedestrians use a high-visibility yellow crosswalk that has been enhanced with an in-street pedestrian sign in San Francisco, California.

102 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments 4. Kamyab, A., S. Andrle, D. Kroeger, and D. S. Heyer. Methods to Reduce Traffic Speed in High-Pedestrian Rural Areas. In Transportation Research Record: Journal of the Transportation Research Board, No. 1828, Transportation Research Board of the National Academies, Washington, D.C., 2003, pp. 31–37. 5. Strong, C. and M. Kumar. Safety Evaluation of Yield-to-Pedestrian Channelizing Devices. Pennsylvania Depart- ment of Transportation, Harrisburg, Pennsylvania, 2006. 6. Banerjee, I., and D. R. Ragland. Evaluation of Countermeasures: A Study on the Effect of Impactable Yield Signs Installed at Four Intersections in San Francisco. Presented at the 86th Annual Meeting of the Transpor- tation Research Board, Washington, D.C., 2007. 7. Ellis, R., R. Van Houten, and J.-L. Kim. In-Roadway “Yield to Pedestrian” Signs: Placement Distance and Motorist Yielding. In Transportation Research Record: Journal of the Transportation Research Board, No. 2002, Transportation Research Board of the National Academies, Washington, D.C., 2007, pp. 84–89. 8. Hua, J., N. Gutierrez, I. Banerjee, F. Markowitz, and D. R. Ragland. San Francisco Pedsafe II Project Outcomes and Lessons Learned. Presented at the 88th Annual Meeting of the Transportation Research Board, Washington, D.C., 2009. 9. Bennett, M. K., H. Manal, and R. Van Houten. A Comparison of Gateway In-Street Sign Treatment to Other Driver Prompts to Increase Yielding to Pedestrians at Crosswalks. Presented at the 93rd Annual Meeting of the Transportation Research Board, Washington, D.C., 2014. Other Signs Evaluation of pedestrian signs has shown them to be of moderate efficacy in increasing pedes- trian safety, with some variation across treatments and site characteristics. A 2006 report by Fitzpatrick et al. suggested that some of the factors that influenced driver yielding at sign loca- tions included the speed and volume of the roadway and whether the motorists perceived yield- ing as a courtesy or the law (1). Signs that are enhanced with flashing beacons or lights have been shown to be more effective when activated manually or automatically by pedestrians than those that blink continuously (1). The following paragraphs give an overview of some of the sign-related research and evaluation from the past 20 years. More information about in-street pedestrian signs can be found in the section of the same name. In the 1990s, a new manufacturing process allowed for the development of high-visibility, “fluorescent strong yellow-green” (SYG) sign material. In 1996, Clark, Hummer, and Dutt eval- uated the performance of pedestrian warning signs that used the new design at sites in central North Carolina. The use of this new sign was associated with increased numbers of cars that slowed down or stopped for pedestrians, although there was no decrease in conflict events fol- lowing sign installation (2). In 1998, Van Houten, Healey, Malenfant, and Retting evaluated the effects of two types of exper- imental signs on motorist yielding behavior. The first was a pictograph of a walking pedestrian that was added to a pedestrian-activated amber flashing beacon suspended over the roadway at the crossing site. It was coupled with a “Yield When Flashing” sign placed 50 m ahead of the crosswalk. Results indicated that both measures were effective in increasing motorist yield percentage, with the most effective treatment being the combination of the two. Only the “Yield When Flashing” sign was effective in reducing vehicle-pedestrian conflicts; the researchers theorized it was a result of the sign’s placement within adequate stopping distance of the crosswalk (3). In 1999, Nitzburg and Knoblauch studied the effectiveness of internally illuminated over- head crosswalk signs that were installed in conjunction with high-visibility crosswalks at two midblock crossing locations in Clearwater, Florida. Using case-control research design, they compared motorist and pedestrian behavior at the treatment sites with two similar sites, one that featured standard pedestrian crossing signage and crosswalk design, and one that had no crosswalk. The researchers found that during the day, drivers at the experimental crossing locations were 30–40 percent more likely to yield than drivers at the control locations. At night, there was a smaller and statistically insignificant increase in driver yielding of 8 percent. There was a significant increase in pedestrians using the crosswalk at the treatment sites compared to control sites. Although the individual effects of having the signs in place could not be analyzed

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 103 separately from the high-visibility crosswalk, the researchers concluded that the treatments had a positive effect on pedestrian safety at the two intersections that were studied (4). In 2000, the Federal Highway Administration published a report that evaluated two types of innovative pedestrian signs that were tested in Seattle and Tucson. Pre- and post-treatment data were collected for all sites, and the following measures of effectiveness (MOEs) were used: (1) percentages of pedestrians for whom motorists yielded; (2) percentage of motorists who yielded to pedestrians; (3) percentage of pedestrians who hesitated, rushed, or aborted in cross- ing; and (4) percentage of pedestrians crossing in the crosswalk (5). The first of the treatments was an overhead, yellow crosswalk sign installed in Seattle, Washington. Before-after data were collected at a single intersection, and analysis of results showed an increase in driver yielding from 45.5 percent before installation to 52.1 percent, which was significant at the 0.06 level. Following the installation of the sign, there was a statis- tically significant decrease in the percentage of pedestrians who ran, aborted, or hesitated in crossing. The researchers concluded that the overhead crosswalk sign was effective at encour- aging driver yielding behavior (5). The second of the treatments studied was a pedestrian-activated “Stop for Pedestrian in Crosswalk” overhead sign installed in Tucson, Arizona. The sign, activated by a pedestrian push button, can be seen to the right in Figure H-9. Two sites were studied. It was found that following the installation of the signs, motorist yielding to pedestrians decreased from 62.9 per- cent to 51.7 percent. The percentage of pedestrians who ran, aborted, or hesitated decreased from 16.7 percent to 10.4 percent. Both decreases were statistically significant. It was theorized that installing the devices on arterial roads with speed limits of 40 mi/h may have limited their effectiveness, and the authors concluded by giving several modifications to the design and test conditions that might improve treatment performance (5). Neither of the two treatments led to a statistically significant increase in crosswalk use; how- ever, the authors concluded that the overhead crosswalk sign and pedestrian regulatory sign were generally effective in increasing the percentages of pedestrians for whom motorists yielded. They cautioned that site characteristics would need to be taken into account when choosing or designing treatments to draw motorists’ attention to pedestrians in crosswalks (5). Figure H-8. An overhead crosswalk sign used in conjunction with double bar pair crosswalk markings and pedestrian crossing signs in Seattle, Washington. Source: Huang et al., 2000.

104 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments In 2002, a Transportation Research Record article by Van Houten, McCusker, Huybers, Malenfant, and Rice-Smith gave the results of experiments that studied the effects of advance yield markings and fluorescent yellow-green RA 4 signs at 24 rural and urban crosswalks throughout Nova Scotia, Canada. The signs featured the message “yield here to pedestrians,” using the yield symbol and an arrow pointing in the direction of the crosswalk on a rectan- gular, fluorescent yellow-green sign. Once baseline data were collected for all 24 crosswalks, they were put into treatment groups of 4, with one of the groups serving as a control through- out the experiment. The other three treatments consisted of (1) advance yield line markings with white-background “yield here to pedestrian” signs, (2) fluorescent yellow-green “yield here to pedestrian” signs, and (3) advance yield line markings with fluorescent yellow-green “yield here to pedestrian” signs. Follow-up data were collected at 6 months following treat- ment installation. Results showed that there was no reduction of vehicle-pedestrian conflicts when the more conspicuous fluorescent yellow-green sign was used instead of the white sign. However, the average number of vehicle-pedestrian conflicts decreased from 11.1 percent and 12.8 percent to 2.7 percent and 2.3 percent, respectively, at sites with the advance yield bar and either a white or fluorescent sign. Advanced stop lines were associated with a statis- tically significant increase in motorist yielding from 69 percent to 85 percent. The authors conclude by recommending the installation of advanced yield markings 7 m to 18 m in advance of the crosswalk, in order to better increase pedestrian visibility of oncoming vehicles when crossing (6). A 2009 evaluation of the Pedsafe II project in San Francisco used video observation and inter- cept surveys to collect pre- and post-treatment data to evaluate the effectiveness of 13 counter- measures deployed at 29 sites throughout San Francisco, California. Two types of signs were installed: portable changeable message speed-limit signs used at midblock locations and “Turning Traffic Must Yield to Pedestrians” signs installed at the corners of intersections. At the midblock locations where portable changeable message speed-limit signs were placed, researchers found a significant reduction in vehicle speeds, by between 1-6 mi/h. At the four intersections where “Turning Traffic Must Yield to Pedestrians” signs were installed, there was a small, but significant increase in the percentage of drivers yielding at all four corners. Of the 13 countermeasures that were tested, these two types of signs were among the six countermeasures considered the most effective in increasing pedestrian safety (7). Figure H-9. Pedestrian regulatory signs used in Tucson, Arizona, in the 1990s. Source: Huang et al., 2000. (5 )

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 105 A 2011 Vermont Agency of Transportation (VTRANS) report described the agency’s experi- ence with installing, evaluating, and maintaining the SmartStud in-pavement crosswalk lighting system and BlinkerSign, a sign equipped with LED lights. In 2006, VTRANS installed and evaluated SmartStud at a crosswalk in Hartford, Vermont. While the results of a pre- and post-treatment analysis revealed that the SmartStud system was effective in increasing pedestrian safety, several of the markers failed as a result of damage from snowplows and vehicles. As a result, in 2008, VTRANS decided to install BlinkerSigns, a type of experimental flashing LED traffic sign that used the existing SmartStud wiring system. The system is activated by pushing a SmartButton or by applying weight to a SmartPed sensor located underfoot. The 2005 pre-SmartStud baseline Source: Lalani, 2001. (8) Figure H-10. A flashing beacon used in conjunction with a pedestrian crossing sign in Austin, Texas. Figure H-11. A pedestrian sign with blinking lights installed at a crosswalk. Source: Kipp and Fitch, 2011. (9)

106 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments data were used and new data were collected to evaluate the BlinkerSign. The researchers found that yielding compliance increased by 23 percent on average following the installation of the BlinkerSign, compared to 13 percent following the installation of SmartStud. Both systems had a comparable effect on approach speeds, leading to a decrease in average driver speed in five of the eight studied scenarios. At 2 years following its installation, BlinkerSign has not required any additional maintenance (9). References 1. Fitzpatrick, K., S. Turner, M. Brewer, P. Carlson, B. Ullman, N. Trout, E. S. Park, J. Whitacre, N. Lalani, and D. Lord. TCRP Report 112/NCHRP Report 562: Improving Pedestrian Safety at Unsignalized Crossings. Trans- portation Research Board of the National Academies, Washington, D.C., 2006. 2. Clark, K. L., J. E. Hummer, and N. Dutt. Field Evaluation of Fluorescent Strong Yellow-Green Pedestrian Warning Signs. In Transportation Research Record 1538, TRB, National Research Council, Washington, D.C., pp. 39–46. 3. Van Houten, R., K. Healey, J. E. L. Malenfant, and R. Retting. Use of Signs and Symbols to Increase the Efficacy of Pedestrian-Activated Flashing Beacons at Crosswalks. In Transportation Research Record: Journal of the Trans- portation Research Board 1636, TRB, National Research Council, Washington, D.C., 1998, pp. 92–95. 4. Nitzburg, M., and R. Knoblauch. An Evaluation of High-Visibility Crosswalk Treatments—Clearwater, Florida. Publication FHWA-RD-00-105, FHWA, U.S. Department of Transportation, 2001. 5. Huang, H., C. Zegeer, R. Nassi, and B. Fairfax. The Effects of Innovative Pedestrian Signs at Unsignalized Locations: A Tale of Three Treatments. Publication FHWA-RD-00-098, FHWA, U.S. Department of Transportation, 2000. 6. Van Houten, R., D. McCusker, S. Huybers, J. E. L. Malenfant, and D. Rice-Smith. Advance Yield Markings and Fluorescent Yellow-Green RA 4 Signs at Crosswalks with Uncontrolled Approaches. In Transportation Research Record: Journal of the Transportation Research Board, No. 1818, Transportation Research Board of the National Academies, Washington, D.C., 2002, pp. 119–124. 7. Hua, J., N. Gutierrez, I. Banerjee, F. Markowitz, and D. R. Ragland. San Francisco Pedsafe II Project Outcomes and Lessons Learned. Presented at the 88th Annual Meeting of the Transportation Research Board, Washington, D.C., 2009. 8. Lalani, N. Alternative Treatments for At-Grade Pedestrian Crossings. Institute of Transportation Engineers, Washington, D.C., 2001. 9. Kipp, W. M. E. and J. Fitch. Evaluation of SmartStud In-Pavement Crosswalk Lighting System and BlinkerSign Interim Report. State of Vermont Agency of Transportation, Montpelier, VT, 2011. Marked Crosswalks and Enhancements The marking of crosswalks at uncontrolled locations, locations where no traffic signals or stop signs exist on the approach at either intersection or midblock locations, has been the subject of debate in the United States. Recent safety research on crosswalks, as discussed below, has helped to resolve some of the controversy on this issue. Marked crosswalks are typically installed at signalized intersections, in school zones, and at unsignalized intersections. The MUTCD defines three types of crosswalk markings: standard Figure H-12. Examples of crosswalk marking patterns. Source: Zegeer et al., 2004. (3)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 107 parallel lines, ladder or continental stripes, and diagonal stripes (1). A 2002 study by Zegeer et al. found no statistically significant difference in pedestrian crash risk for various types of cross- walk markings (standard parallel lines, ladder, zebra, or continental style). Crosswalks may be raised (“speed tables”) or used in conjunction with supplemental signing, in-pavement flash- ing lights, overhead flashers, nighttime lighting, pedestrian refuge islands, signalization, and/or other devices. Several studies prior to comprehensive studies by Zegeer et al. in 2002 and Knoblauch et al. in 2000 produced a wide range of results concerning the safety effects of marked vs. unmarked crosswalks. However, none of these earlier studies attempted to analyze the effects of marked vs. unmarked crosswalks specifically for different numbers of lanes, traffic volume, or other roadway features. A number of studies conducted between 1972 and 2000 concluded that pedestrian crashes were higher in marked crosswalks compared to unmarked crosswalks. For example, an often-cited 1972 San Diego study by Herms concluded that crashes on marked crosswalks were twice as frequent per unit of pedestrian volume compared to unmarked crosswalks (Herms, 1972 as cited in [4]). Herms looked at 400 intersections in the city, each of which had one marked and one unmarked crosswalk leg on the same street. In an earlier version of the same study (Herms, 1970), the author mentioned San Diego’s 1962 warrants for determining where to paint crosswalks. The city’s warrants required marking crosswalks when traffic gaps were inadequate, pedestrian volume was high, speed was moderate, and/or there were other relevant factors such as previous crashes. These criteria suggest that crosswalks in San Diego were painted where the conditions were already most conducive to pedestrian crashes or which already had a history of pedestrian crashes. In 1974, Gurnett described a project in which painted crosswalk stripes were removed from three locations because of a recent bad crash history (Gurnett, 1974 as cited in [4]). There were fewer crashes after removal of the stripes, but these findings might simply be due to regression- to-mean, since the only sites that were “treated” (i.e., crosswalks were removed) were those that had a recent history of pedestrian crashes. In 1983, Tobey et al. examined crashes at both marked and unmarked crosswalks as a func- tion of pedestrian volume (P) multiplied by vehicle volume (V) and, unlike some of the previous studies cited, reported fewer accidents at marked crosswalks than at unmarked ones (Tobey et al., 1983 as cited in [4]). However, this may be due to the fact that Tobey’s study included signalized as well as uncontrolled crossings, and it is likely that more marked crosswalks were at controlled locations than unmarked crosswalks were. It should be mentioned that the study methodol- ogy was designed to determine the pedestrian crash rate for a variety of human and location conditions, but was not specifically intended to quantify the isolated safety effects of marked vs. unmarked crosswalks. In 1994, Gibby et al. analyzed crashes at 380 unsignalized highway intersections in California from among 10,000 candidate intersections throughout the state (Gibby et al., 1994 as cited in [4]). Crash rates per pedestrian–vehicle volume were two or three times higher in marked than in unmarked crosswalks at these sites. Like other older studies, this study combined all sites with marked crosswalks and unmarked crosswalks, and did not conduct a separate analysis for different cross sections, traffic volumes, and other roadway features. In 2000, Jones and Tomcheck evaluated pedestrian crashes at crosswalks at unsignalized arterial intersections in Los Angeles to test the validity of the city’s crosswalk policies. The study attempted to determine whether removing a crosswalk marking reduced pedestrian crashes at such locations, and/or increased pedestrian crashes at adjacent unprotected sites. Jones and Tomcheck analyzed pedestrian crashes at 104 unsignalized intersections on arterials where

108 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments parallel-line crosswalks had been removed due to resurfacing, rather than at sites with pedestrian accident histories. At many intersections, some legs had both marked and unmarked cross- walks before and after the study. An average of approximately 7 years of pedestrian crash data was collected for each of the before and after periods for the 104 sites. Traffic and pedestrian exposure data were not collected, but untreated comparison sites were identified and used in the analysis (9). When only the legs of the intersections that previously had marked crosswalks were considered, Jones and Tomcheck found that there was a 73 percent reduction (from 116 to 31) in pedestrian crashes after crosswalk markings were removed at the 104 sites combined. Considering both legs (previously marked and unmarked crosswalks) of the intersections, there was a statistically significant decline of 61 percent (from 129 to 50) in pedestrian crashes. There was no statistically significant increase in pedestrian crashes at intersections adjacent to intersections where cross- walk markings were removed. At the 15 intersections where crosswalk markings were retained, pedestrian crashes did not decrease. The authors recommended supporting “a policy of selectively installing or reinstalling marked, unprotected crosswalks only after careful consideration” (9). It should be noted that the study did not report the effects of removing crosswalk markings by road type (i.e., two-lane vs. multi-lane) or volumes at the study sites. In the most comprehensive study of marked crosswalks at uncontrolled intersection and mid- block locations to date, Zegeer, Stewart, Huang, and Lagerwey (2002) analyzed data from 1,000 marked and 1,000 matching unmarked crosswalk sites in 30 U.S. cities. Zegeer et al. determined that some site factors such as area type, speed limit, and crosswalk marking pattern were not associ- ated with pedestrian crashes. Site factors that were related to pedestrian crashes, which were used as control variables in the analysis, included pedestrian ADT, vehicle ADT, number of lanes, median type, and region of the United States. Poisson and negative binomial regression models were used to analyze the crash effects of marked vs. unmarked crosswalks (2). At uncontrolled locations on two-lane roads and multi-lane roads with low traffic volumes (ADT below 12,000 vehicles per day), it was found that a marked crosswalk alone, compared with an unmarked crosswalk, made no statistically significant difference in pedestrian crash rate. On multi-lane roads with an ADT of more than 12,000 vehicles per day, a marked crosswalk in the absence of other substantial improvements was associated with a statistically significant higher pedestrian crash rate compared to sites with an unmarked crosswalk. On multi-lane roads, the presence of raised medians in marked or unmarked crosswalks provided statistically significant lower crash rates than did no raised median. There were two potential explanations for some of the higher crash rates seen at higher vol- ume crosswalks. First, the crash rates for older pedestrians were higher than for other pedestrian age groups, considering pedestrian crashes and exposure by age. It was found that older pedes- trians were more likely than younger pedestrians to cross at a marked crosswalk, which may partially explain the higher pedestrian crash rate at marked crosswalks. Second, it was theorized that marked crosswalks led to higher crash rates due to multiple-threat crashes on multi-lane roads. Multiple-threat crashes occur when a vehicle in the curb lane stops for a pedestrian in the crosswalk, simultaneously screening the pedestrian’s view of an oncoming vehicle and the oncoming vehicle’s view of the pedestrian, leading to a failure of the vehicle to yield. Zegeer et al. suggested a number of potential improvements at unsignalized crossing locations on multi-lane, higher volume roads to enhance pedestrian safety. These recommendations include providing raised medians on multi-lane roads, installing traffic and pedestrian signals where warranted, adding curb extensions or raised islands to reduce street-crossing distance, installing adequate nighttime lighting at pedestrian crossings, constructing raised street crossings, and designing safer intersection and driveways (e.g., with tighter turn radii).

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 109 Source: Zegeer et al., 2002. (2) Source: Zegeer et al., 2002. (2) Figure H-13. Pedestrian crash rates by traffic volume and presence/absence of crosswalk markings.

110 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Revisiting Older Studies As documented by Campbell et al. (4), authors of the Zegeer et al. study (2002) attempted to compare their results with those of the 1972 Herms San Diego study. Taking all of the 2,000 sites together as one group and simply dividing the crashes by pedestrian crossing volume (as Herms did), the Zegeer group also found that marked crosswalks had a pedestrian crash rate that was slightly more than twice the rate of unmarked crosswalk sites. Only when a more sophisticated statistical analysis was applied did the researchers in the Zegeer et al. study find that marked crosswalks are associated with higher pedestrian crash risk only on high-volume, multi-lane roads (i.e., ADT above 12,000 veh/day). Similarly, in 1967, the Los Angeles County Road Department found that accident frequency increased from 4 to 15 after marked crosswalks were installed at 89 non-signalized intersections (as cited in [10]). All the locations that showed an increase in crashes after crosswalk installation had an ADT greater than 10,900 vehicles; sites with fewer vehicles experienced no change in pedestrian crashes, which was consistent with the findings of the Zegeer et al. study. At the same time as Zegeer et al.’s research, Knoblauch performed two studies published in 2000 and 2001 on pedestrian and motorist behavior. The first of these studies was an effort to assess the effect of crosswalk markings on driver and pedestrian behavior at 11 unsignalized loca- tions in four U.S. cities (5). All of the sites were two- or three-lane roads with relatively low speed limits (35 to 40 mi/h) and low volumes (fewer than 12,000 vehicles per day). Given these charac- teristics, the authors concluded that marking pedestrian crosswalks had no measurable negative effect on either pedestrian or motorist behavior. Crosswalk usage increased after markings were installed, but no evidence was found that pedestrians were less vigilant or more assertive in the marked crosswalk. Drivers were found to approach a pedestrian in the crosswalk rather slowly, but no changes in driver yielding were noted. Details on the duration of the study periods were not reported (5). Knoblauch’s second study was performed at six sites in Maryland, Virginia, and Arizona in 2000. All of the locations were uncontrolled intersection approaches without traffic signals or stop control and with a 35 mi/h speed limit that had been recently resurfaced. Using a staged Roadway Type (Number of Travel Lanes and Median Type) Vehicle ADT Ä9,000 Vehicle ADT >9,000 to 12,000 Vehicle ADT >12,000- 15,000 Vehicle ADT >15,000 Speed Limit Ä48.3 km/h (30 mi/h) 56.4 km/h (35 mi/h) 64.4 km/h (40 mi/h) Ä48.3 km/h (30 mi/h) 56.4 km/h (35 mi/h) 64.4 km/h (40 mi/h) Ä48.3 km/h (30 mi/h) 56.4 km/h (35 mi/h) 64.4 km/h (40 mi/h) Ä48.3 km/h (30 mi/h) 56.4 km/h (35 mi/h) 64.4 km/h (40 mi/h) Two lanes C C P C C P C C N C P N Three lanes C C P C P P P P N P N N Multilane (four or more lanes) with raised median C C P C P N P P N N N N Multilane (four or more lanes) without raised median C P N P P N N N N N N N Table 11 from Zegeer et al. (2002) giving recommended guidelines for installing marked crosswalks and other pedestrian improvements at uncontrolled locations. C = Candidate sites for marked crosswalks, P = Possible increase in pedestrian crash risk if other enhancements are not used, N = Marked crosswalks alone are insufficient due to increased crash risk. (2) Table H-2. Recommended guidelines for installing marked crosswalks and other pedestrian improvements at uncontrolled locations.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 111 pedestrian at sample crossing locations, speed data were taken under three conditions: no pedes- trian present, pedestrian looking, pedestrian not looking. Results indicated a slight reduction in vehicle approach speeds at most, but not all, of the locations after the crosswalk markings had been installed. There was a significant reduction in overall speed under conditions of no pedes- trians and where pedestrians were not looking (6). A 2002 JAMA article by Koepsell, McCloskey, Wolf, Moudon, Buchner, Kraus, and Patterson studied the effect of crosswalk markings at urban intersections on the risk of injury to older pedestrians. The researchers looked at 282 sites where a pedestrian 65 years old or older had been struck by a motor vehicle while crossing the street. They matched these case sites to 564 control sites chosen for proximity to case sites and street classification characteristics. On the same day of the week and at the same time of day when the accident occurred, trained observers collected data on environmental characteristics, vehicle flow and speed, and pedestrian use for each site. Once the data were adjusted based on pedestrian and vehicle flow, crossing length, and signalization, it was found that the risk of a pedestrian–motor vehicle accident was 2.1 times as great at sites with a marked crosswalk. This excess risk was due almost entirely to the higher risk associated with using marked crosswalks at uncontrolled, unsignalized locations. The researchers concluded that marked crosswalks, when used alone, put older pedestrians at elevated risk of being struck by vehicles (7). In 2007, Mitman, Ragland, and Zegeer conducted another study summarizing pedestrian and driver behavior at uncontrolled intersections using observations at marked and unmarked cross- walks. The data were collected on low speed, two-lane, and multi-lane arterials. Using statistical analysis, the study found that drivers are more likely to yield to pedestrians in marked crosswalks as opposed to unmarked crosswalks. The results led the research team to recommend the creation of a crosswalk inventory to prioritize improvements, using HAWK beacons, undertaking educa- tion initiatives, and using enforcement measures both for pedestrians and drivers (8). Despite contradictory findings of various studies, it is clear that marked crosswalks are generally not associated with any statistically significant difference in pedestrian crash risk (compared to unmarked crosswalk sites) on two-lane roads or on multi-lane roads with fewer than 12,000 vehicles per day. On multi-lane roads with ADT higher than 12,000 vehicles per day, marked crosswalks installed alone, without other substantial safety devices, carry signifi- cantly increased crash risk for pedestrians, unless more substantial pedestrian safety treat- ments are provided. On many roads (particularly for multi-lane roads with ADT higher than about 12,000 veh/day), the safety professional may consider such crossing treatments (e.g., raised medians on multi-lane roads, traffic and pedestrian signals, where warranted, adequate nighttime lighting at pedestrian crossings, etc.) to help pedestrians to cross streets more safely. The following is a summary of some of these studies that involved evaluating pedestrian behav- ior on marked vs. unmarked crosswalks. Studies of pedestrian and motorist behavior suggest that pedestrian behavior is generally improved by marking crosswalks, and no indication of reckless behavior has been found associated with marked crosswalks. However, most of these behavioral studies were on two- or three-lane roads, where no differences were found in pedestrian crash risk between marked and unmarked crosswalks. Pedestrian Behavior Knoblauch et al. (2001) launched a study intended to observe the type of reckless pedestrian behavior to which Herms and others attributed the negative crash results reported in some of the marked crosswalk studies (as cited in [4]). The researchers gathered data at 11 sites before and after marked crosswalks were installed, evaluating the information in terms of three hypotheses regarding pedestrian behavior. The first hypothesis was that pedestrians, feeling more protected

112 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments in a marked crosswalk, would act more aggressively toward motorists. An analysis of data by the research team found no statistically significant difference in blatantly aggressive behavior by pedestrians following the crosswalk installation. The second hypothesis involved whether the pedestrians crossed within the marked lines of the crosswalk, and the data showed that pedestrians walking alone tended to use the marked crosswalk, especially at intersections, while pedestrian groups did not. Additionally, there was a statistically significant increase in overall crosswalk usage following crosswalk installation. The third hypothesis dealt with pedestrian vigilance. It was thought that pedestrians might become less vigilant in monitoring oncoming traffic when using a marked crosswalk, but results showed that pedestrian vigilance increased following crosswalk installation (4). These findings were consistent with an earlier study of pedes- trian behavior done by Knoblauch et al. (1987) that considered the effect of marked crosswalks on pedestrian looking behavior and staying within the area defined by the markings (4). A 1979 study by Hauck evaluated 17 crosswalks at traffic signals that were re-painted in Peoria, IL (as cited in [4]). A before-after analysis found a decrease in both pedestrian and motor- ist violations at the sites after installation of marked crosswalks. Jaywalking was unchanged, but the number of people who stepped out in front of traffic decreased at 12 of the locations, and those crossing against the DON’T WALK signal phase decreased at 13 sites. Motorist Behavior In 2000, Knoblauch and Raymond took speed measurements at six locations before and after marked crosswalks were installed (as cited in [4]). Speeds were measured (1) with no pedestri- ans present, (2) with a member of the research team posing as a pedestrian who was looking at traffic, and (3) when the team member approached and stood at the curb looking straight across the road rather than at oncoming traffic. Motorist behavior was not consistent, so the results were not clear-cut. At one site, drivers slowed down considerably even when no pedestrians were present. When a pedestrian was present and looking at traffic, there was a small but not statistically significant decrease in speed at all six locations. Knoblauch reasoned that drivers might assume a pedestrian looking toward oncoming traffic would not try to cross the street, so vehicles did not need to slow down. However, when the pedestrian was present and not looking for oncoming cars, drivers approaching the marked crosswalk did slow down enough to register a statistically significant change. Knoblauch’s conclusion was that drivers usually respond to crosswalk markings, especially when a pedestrian is present but not watching traffic (4). In 2001, Knoblauch et al. studied motorist behavior on two- and three-lane roads with 35 to 40 mi/h speed limits, studying the effects of the crosswalk markings on motorist behavior. The researchers found that drivers slowed slightly more when approaching pedestrians in marked rather than unmarked crosswalks, as well as no effect on yielding behavior when comparing pedestrians in marked versus unmarked crosswalks. In 1975, Katz et al. studied driver-pedestrian interaction when members of the research team crossed the street under a variety of conditions in 960 trials. Drivers were more likely to stop for pedestrians when the vehicle approach speed was low, when the pedestrian was in a marked crosswalk, when the distance between the car and pedestrian was greater rather than lesser, when there was a group of pedestrians, and when the pedestrians did not make eye contact with the driver (as cited in [4]). References 1. Federal Highway Administration. Manual on Uniform Traffic Control Devices for Streets and Highways. Washington, D.C., 2009. http://mutcd.fhwa.dot.gov/htm/2009/part3/part3b.htm#section3B18 2. Zegeer, C. V., J. R. Stewart, H. H. Huang, and P. A. Lagerwey. Safety Effects of Marked vs. Unmarked Crosswalks at Uncontrolled Locations: Executive Summary and Recommended Guidelines. Publication FHWA-RD-01-075, FHWA, U.S. Department of Transportation, 2002.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 113 3. Zegeer, C. V., J. Stutts, H. Huang, M. J. Cynecki, R. Van Houten, B. Alberson, R. Pfefer, T. R. Neuman, K. L. Slack, and K. K. Hardy. NCHRP Report 500: Guidance for Implementation of the AASHTO Strategic Highway Safety Plan, Volume 10: A Guide for Reducing Collisions Involving Pedestrians. Transportation Research Board of the National Academies, Washington, D.C., 2004. 4. Campbell, B. J., C. V. Zegeer, H. H. Huang, and M. J. Cynecki. A Review of Pedestrian Safety Research in the United States and Abroad. Publication FHWA-RD-03-042, FHWA, U.S. Department of Transportation, 2004. 5. Knoblauch, R. L., M. Nitzburg, and R. F. Seifert. Pedestrian Crosswalk Case Studies: Richmond, Virginia; Buffalo, New York; Stillwater, Minnesota. Publication FHWA-RD-00-103, FHWA, U.S. Department of Transportation, 2001. 6. Knoblauch, R. L. and P. D. Raymond. The Effect of Crosswalk Markings on Vehicle Speeds in Maryland, Virginia and Arizona. Publication FHWA-RD-00-101, FHWA, U.S. Department of Transportation, 2000. 7. Koepsell, T., L. McCloskey, M. Wolf, A. V. Moudon, D. Buchner, J. Kraus, and M. Patterson. Crosswalk Markings and the Risk of Pedestrian-Motor Vehicle Collisions in Older Pedestrians. Journal of the American Medical Association, Vol. 288, No. 1, 2002, pp. 1–8. 8. Mitman, M. F., D. R. Ragland, and C. V. Zegeer. Marked Crosswalk Dilemma: Uncovering Some Missing Links in a 35-Year Debate. Transportation Research Record: Journal of the Transportation Research Board, No. 2073, Transportation Research Board of the National Academies, Washington, D.C., 2008, pp. 86–93. 9. Jones, T. L. and P. Tomcheck. Pedestrian Accidents in Marked and Unmarked Crosswalks: A Quantitative Study. ITE Journal, Vol. 70, No. 9, 2000, pp. 42–46. 10. Chicago Department of Transportation. Evaluation of School Traffic Safety Program Traffic Control Measure Effectiveness. Report to the FHWA, 2005. http://mutcd.fhwa.dot.gov/resources/policy/ygcrosswalkmarking/ chicagostudy/index.htm Advanced Yield or Stop Markings and Signs Advance yield or stop markings are a type of pavement marking placed before a crosswalk to increase the distance at which drivers stop or yield to allow pedestrians to cross. Increasing the distance between yielding vehicles and pedestrians increases the ability of motorists in other lanes to see the pedestrian as he or she crosses and to yield accordingly. Pedestrian visibility of oncoming traffic is likewise improved. In 1988, Van Houten used a combination of advanced stop markings and “Stop Here for Pedestrians” signs at three Dartmouth, Nova Scotia, crosswalks to analyze the effect of the Source: Van Houten, 1988. (1) Figure H-14. Graph showing the percentage of motorists stopping under two sets of study conditions.

114 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments treatments on vehicle-pedestrian conflicts and yielding behavior at the sites. An analysis of pre- and post-treatment data indicated that the markings and signs produced an 80 percent decrease in vehicle-pedestrian conflicts as well as an increase in percentage of yielding motorists at treat- ment sites. Based on the results of this study, the Nova Scotia Department of Transportation began using advance yield markings throughout the province in order to mark crosswalks (1). One year later, Malenfant and Van Houten (1989) studied advance stop lines used with signs as a means of increasing motorist yielding at 34 crosswalks in three Canadian cities in Newfoundland Figure H-15. Motorist yielding following the implementation of countermeasures in Canada. The vertical stepped line shows the introduction of the treatments, pedestrian signs, and advance stop lines, deployed with education and enforcement programs. The horizontal lines represent the mean yielding percentage during baseline. Source: Malenfant and Van Houten, 1989. (2)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 115 and New Brunswick. Baseline data were collected in each of the cities prior to treatment, which consisted of education and enforcement in addition to the engineering countermeasures. Motor- ist yielding at follow up increased from 54 percent to 81 percent in St. John’s (Newfoundland), from 9 percent to 68 percent in Fredericton (New Brunswick), and from 44 percent to 71 percent in Moncton-Dieppe (New Brunswick). Given the scope of the treatment program, it was unclear to what extent the advance stop lines and signs contributed to the increase in motorist yielding behavior (2). A 1992 article in Accident Analysis and Prevention by Van Houten and Malenfant continued to evaluate advance stop markings as used with a pedestrian warning sign. The researchers applied a sequential series of enhancements at two marked crosswalks in Dartmouth, Nova Scotia. Pre-treatment data were collected at baseline and then following the addition of “Stop Here for Pedestrian” signs, following the placement of a stop line 50 ft in advance of the crosswalk, and at 1 year following treatment installation. Following the installation of the signs, pedestrian–vehicle conflicts decreased from 53 percent to 25 percent on Portland Street and from 25 percent to 10 percent on Prince Albert Road. The introduction of the stop lines was associated with an addi- tional reduction of pedestrian–vehicle conflicts from 25 percent to 10 percent at Portland Street and from 10 percent from 6 percent on Prince Albert Road. The reduction in pedestrian–vehicle conflicts was maintained at follow-up one year following their installation. While the sign and advance stop line had little effect on the percentage of motorists who yielded to pedestrians, they did produce an increase in motorist yielding distance and a decrease in vehicle-pedestrian conflicts (3). Advance stop lines and signs, as well as education and enforcement were evaluated at 34 cross- walks in three Canadian cities in Newfoundland and New Brunswick provinces. Motorist yielding increased before to after in all three cities. Yielding increased from 54 to 81 percent in St. Johns (NL), from 9 to 68 percent in Fredericton (NB), and from 44 to 71 percent in Moncton-Dieppe (NB) in response to the treatments. Given the treatment combination, it is unclear to what extent the advance stop lines and signs and the education and enforcement measures contributed to the increase in motorist yielding behavior. Average pedestrian crashes and injuries also trended lower in St. John’s and Fredericton and in the after period, but there were no controls for other potential causes (4). In 1993, Cynecki, Sparks, and Grote studied the effects of a different type of advance stop indi- cator: transverse rumble strips installed in advance of marked crosswalks at 19 uncontrolled loca- tions. There was little change in vehicle speed; 85th percentile speeds showed no real change (5). Figure H-16. Photos of crosswalk sign and markings at three sites. Source: Van Houten et al., 2001. (6)

116 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments In 2001, Van Houten, Malenfant and McCusker studied the effectiveness of advance yield markings used with symbol signs at three crosswalks in Nova Scotia, Canada, where yellow flash- ing beacons were already in place. The researchers experimented with yield marking placement, finding that marking and sign placement was effective at distances between 10 and 25 meters in advance of the crosswalk. The addition of the sign and yield markings led to decreases in vehicle- pedestrian conflicts of 74 percent, 87 percent, and 67.1 percent at the three sites. Like previous studies, there was a small increase in motorist yielding behavior (6). A 2002 Transportation Research Record article by Van Houten, McCusker, Huybers, Malenfant, and Rice-Smith gave the results of experiments that studied the effects of advance yield markings and fluorescent yellow-green RA 4 signs at 24 rural and urban crosswalks throughout Nova Scotia, Canada. The signs featured the message “Yield Here to Pedestrians,” using the yield symbol and an arrow pointing in the direction of the crosswalk on a rectangular, fluorescent yellow-green sign. Once baseline data were collected for all 24 crosswalks, they were put into treatment groups of four, with one of the groups serving as a control throughout the experiment. The other three treatments consisted of (1) advance yield line markings with white-background “yield here to pedestrian” signs, (2) fluorescent yellow-green “yield here to pedestrian” signs, and (3) advance yield line markings with fluorescent yellow-green “yield here to pedestrian” signs. Follow-up data were collected at 6 months following treatment installation. Results showed that there was no reduction of vehicle-pedestrian conflicts when the more conspicuous fluorescent yellow-green sign was used instead of the white sign. However, the average number of vehicle-pedestrian con- flicts decreased from 11.1 percent and 12.8 percent to 2.7 percent and 2.3 percent, respectively, at sites with the advance yield bar and either a white or fluorescent sign. Advanced stop lines were associated with a statistically significant increase in motorist yielding from 69 percent to 85 percent. The authors conclude by recommending the installation of advanced yield markings 7 m to 18 m in advance of the crosswalk, in order to better increase pedestrian visibility of oncoming vehicles when crossing (7). Another study by Nambisan, Vasudevan, Dangeti, and Virupaksha in 2007 examined driver and pedestrian behavior at unsignalized intersections with respect to combinations of Danish offset, advance yield markings, and high-visibility crosswalk markings. This analysis was based on data from Las Vegas, Nevada, and used an observational study approach. Results indicated Source: Van Houten et al., 2001. (6) Figure H-17. A pedestrian crosses while a vehicle waits at the advance yield markings.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 117 that Danish offset and high-visibility crosswalk treatments lead to a yielding rate of just below 50 percent at two sites, while the use of advanced yield markings caused the yielding rate to increase. Following statistical tests, the study concluded that Danish offsets, median refuge islands, and high-visibility crosswalks do enhance pedestrian safety with advance yield mark- ings being more successful when coupled with Danish offsets as opposed to a combination with pedestrian refuge islands (8). A 2009 report by Pecheux, Bauer, and McLeod gave the results of an evaluation of advance stop lines installed at one signalized and one unsignalized intersection in San Francisco. Based on pre- and post-treatment measurements taken of driver yielding, vehicle stop position, and pedestrian–vehicle conflicts, the researchers concluded that the advance stop lines had no impact on driver behavior or pedestrian safety at the sites (9). A 2013 TRB paper by Samuel et al. summarized the results of two experiments conducted in Massachusetts to evaluate the effectiveness of advanced yield lines on drivers’ ability to scan for pedestrians and driver yielding behavior. The researchers conducted an observational study using a staged pedestrian attempting to use the crosswalk and a second experiment on an open- road course in Greenfield, MA. Results show that advanced yield markings (coupled with warn- ing signs) improved drivers’ compliance in scanning for pedestrians. Additionally, advanced yield markings (coupled with vacant on-street parking adjacent to the crosswalk) were found to improve driver yielding compliance (10). Figure H-18. Graph from Van Houten et al. (2002) showing the number of pedestrian-motor vehicle conflicts per 100 crossings at each site during each phase of the evaluation: (a) control sites, (b) advance yield marking sites, (c) yellow-green pedestrian sign sites, and (d) advanced yield marking and yellow-green pedestrian sign sites. (7)

118 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments A 2013 TRB paper by Hengel presented the results of a study of a single site in Santa Barbara, CA, where a curb extension, pedestrian refuge island, and stop bars were installed. MOEs were crossing delay, number of motorists failing to yield, and distance drivers yield from the crosswalk. Results show that the combination of treatments is effective at reducing wait times to cross, decreasing percentage of vehicles that pass before yielding, and increasing the distance that vehicles yield in advance of the crosswalk (11). References 1. Van Houten, R. The Effects of Advance Stop Lines and Sign Prompts on Pedestrian Safety in Crosswalk on a Multilane Highway. Journal of Applied Behavior Analysis, Vol. 21, 1988, pp. 245–251. 2. Malenfant, L., and R. Van Houten. Increasing the Percentage of Drivers Yielding to Pedestrians in Three Canadian Cities with a Multifaceted Safety Program. Health Education Research, Vol. 5, No. 2, 1989, pp. 275–279. 3. Van Houten, R., and L. Malenfant. The Influence of Signs Prompting Motorists to Yield Before Marked Crosswalks on Motor Vehicle-Pedestrian Conflicts at Crosswalks with Pedestrian Activated Flashing Lights. Accident Analysis and Prevention, Vol. 24, 1992, pp. 217–225. 4. Malenfant, L. and Van Houten, R. (1990). Increasing the Percentage of Drivers Yielding to Pedestrians in Three Canadian Cities with a Multifaceted Safety Program. Health Education Research, 5(2), 275–279. 5. Cynecki, M. J., J. W. Sparks, and J. L. Grote. Rumble Strips and Pedestrian Safety. ITE Journal, 1993, pp. 18–24. 6. Van Houten, R., J. E. L. Malenfant, and D. McCusker. Advance Yield Markings: Reducing Motor Vehicle– Pedestrian Conflicts at Multilane Crosswalks with Uncontrolled Approach. In Transportation Research Record: Journal of the Transportation Research Board, No. 1773, TRB, National Research Council, Washington, D.C., 2001, pp. 69–74. 7. Van Houten, R., D. McCusker, S. Huybers, J. E. L. Malenfant, and D. Rice-Smith. Advance Yield Markings and Fluorescent Yellow-Green RA 4 Signs at Crosswalks with Uncontrolled Approaches. In Transportation Research Record: Journal of the Transportation Research Board, No. 1818, Transportation Research Board of the National Academies, Washington, D.C., 2002, pp. 119–124. 8. Nambisan, S., V. Vasudevan, M. Dangeti, and V. Virupaksha. Advanced Yield Markings and Pedestrian Safety: Analyses of Use with Danish Offsets and Median Refuge Islands. Presented at 87th Annual Meeting of the Transportation Research Board, Washington, D.C., 2008. Source: Pecheux et al., 2009. (9) Figure H-19. A pedestrian crosses in a crosswalk at an intersection where advance stop lines have been installed. These lines increase the distance between stopped cars and pedestrians, increasing visibility for both.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 119 9. Pecheux, K., J. Bauer, P. McLeod. Pedestrian Safety Engineering and ITS-Based Countermeasures Program for Reducing Pedestrian Fatalities, Injury Conflicts, and Other Surrogate Measures: Final System Impact Report. Federal Highway Administration, U.S. Department of Transportation, 2009. 10. Samuel, S., M. R. E. Romoser, L. R. Gerardino, M. Hamid, R. A. Gomez, M. A. Knodler, Jr., J. Collura, and D. L. Fisher. Effect of Advance Yield Markings and Symbolic Signs on Vehicle-Pedestrian Conflicts: Field Evaluation. In Transportation Research Record: Journal of the Transportation Research Board, No. 2393, Trans- portation Research Board of the National Academies, Washington, D.C., 2013, pp. 139–146. 11. Hengel, D. Build It and They Will Yield: Effects of Median and Curb Extension Installations on Motorist Yield Compliance. Presented at the 92nd Annual Meeting of the Transportation Research Board, Washing- ton, D.C., 2013. Treatment 2. High-Visibility Crosswalk Marking Patterns Evaluation Studies of High-Visibility Crosswalks A 2001 Federal Highway Administration report by Nitzburg and Knoblauch evaluated the effectiveness of high-visibility crosswalk markings used in conjunction with an illuminated over- head crosswalk sign at two sites in Clearwater, Florida. The researchers used case-control research design to compare motorist and pedestrian behavior at the treatment sites with two similar sites, one that featured standard pedestrian crossing signage and crosswalk design, and one that had no crosswalk. The researchers found that during the day, drivers at the experimental crossing locations were 30–40 percent more likely to yield than drivers at the control locations. At night, there was a smaller and statistically insignificant increase in driver yielding of 8 percent. There was a significant increase in pedestrians using the crosswalk at the treatment sites compared to control sites. Although the individual effects of having the signs in place could not be analyzed separately from the high-visibility crosswalk, the researchers concluded that the treatments had a positive effect on pedestrian safety at the two intersections that were studied (1). A 2005 report for the Chicago Department of Transportation gave the results of an evaluation of the experimental use of strong yellow/green (SYG) crosswalk markings at over 100 Chicago elementary school zone crosswalks. City officials measured traffic speeds before and after the installation of the SYG crosswalks to determine whether the color of the pavement markings led to an improved pedestrian safety environment at the crossings. An analysis of traffic speeds sug- gested that the use of SYG crosswalk markings failed to have a significant effect on the percentage of drivers exceeding the speed limit or median 85th percentile speeds at study locations. Based on the results of the Chicago Department of Transportation’s analysis, the FHWA concluded that the use of yellow-green pavement markings did not improve crosswalk safety compared to standard white markings (2). A 2010 article by Feldman, Manzi, and Mitman provided an Empirical Bayesian evaluation of the safety outcomes of installing high-visibility crosswalks at 54 school sites in San Francisco, California. The researchers used an equal number of control intersections and pre-treatment data to predict the number of collisions that would have been expected in absence of treatment. The results of their analysis demonstrated a statistically significantly reduction in collisions of 37 percent (3). A 2011 Federal Highway Administration report by Fitzpatrick, Chrysler, Iragavarapu, and Park evaluated the relative visibility of three types of crosswalk markings, transverse lines, conti- nental markings, and bar pair markings, under daytime and nighttime conditions. Seventy-eight participants were recruited, evenly divided by gender and age (over/under 55), and drove an instrumented vehicle on a route in College Station, Texas. The participants were given instruc- tions to identify crosswalks and other roadway features as they came into view, at which point researchers used the instrumentation to mark the location at which the crosswalk was visible. Results were adjusted to account for response delay. Detection distances were analyzed with regards to marking type, light conditions, site characteristics, traffic characteristics, vehicle type,

120 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments and driver characteristics. Analysis of results showed that detection distances for continental and bar pairs were statistically similar and are also statistically significantly longer than for transverse line markings at day and at night. Participants also preferred the continental and bar pair mark- ings to the transverse markings. The presence of traffic also had the effect of reducing detection distance. Age, gender, driver eye height, and vehicle type were found to have minimal signifi- cance by the research team. The researchers concluded by suggesting the addition of bar pairs to the MUTCD and to also make bar pairs or continental markings the default crosswalk marking across uncontrolled approaches (4). A 2012 article by Chen, Chen, Ewing, McKnight, Srinivasan, and Roe considered the effective- ness of high-visibility crosswalks in increasing pedestrian safety at intersections. The researchers used a two-group pre-test/post-test research design to compare collision statistics following the implementation of high-visibility crosswalks at 72 sites throughout New York City. Pedestrian collision statistics were collected for the 5 years preceding treatment installation, as well as the two years following it, and the authors used ANCOVA analysis in order to control for potential regression-to-the-mean effects. Analysis of their results indicated that the average pedestrian Figure H-20. Bar pair markings. Source: Fitzpatrick et al., 2011. (4) Figure H-21. High-visibility crosswalks and raised crossing islands help pedestrians cross safely. Source: safety.transportation.org

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 121 crash rate decreased by 44.9 percent at treatment sites and by 11.5 percent at comparison sites. This resulted in an ANCOVA-adjusted reduction in pedestrian collisions of 48 percent at treat- ment sites, results which were significant at the 0.05 level (5). References 1. Nitzburg, M., and R. Knoblauch. An Evaluation of High-Visibility Crosswalk Treatments—Clearwater, Florida. Publication FHWA-RD-00-105, FHWA, U.S. Department of Transportation, 2001. http://www.fhwa.dot.gov/ publications/research/safety/pedbike/0105.pdf 2. Chicago Department of Transportation. Evaluation of School Traffic Safety Program Traffic Control Measure Effectiveness. Report to the FHWA, 2005. http://mutcd.fhwa.dot.gov/resources/policy/ygcrosswalkmarking/ chicagostudy/index.htm 3. Feldman, M., J. G. Manzi, and M. Mitman. Empirical Bayesian Evaluation of Safety Effects of High-Visibility School (Yellow) Crosswalks in San Francisco, California. In Transportation Research Record: Journal of the Trans- portation Research Board, No. 2198, Transportation Research Board of the National Academies, Washington, D.C., 2010, pp. 8–14. 4. Fitzpatrick, K., S. Chrysler, V. Iragavarapu, and E. S. Park. Detection Distances to Crosswalk Markings: Trans- verse Lines, Continental Markings, and Bar Pairs. In Transportation Research Record: Journal of the Transportation Research Board, No. 2250, Transportation Research Board of the National Academies, Washington, D.C., 2011, pp. 1–10. Also summarized in Fitzpatrick, K., S. Chrysler, R. Van Houten, W. Hunter, and S. Turner. Evaluation of Pedestrian and Bicycle Engineering Countermeasures: Rectangular Rapid-Flashing Beacons, HAWKs, Sharrows, Crosswalk Markings, and the Development of an Evaluation Methods Report. Publication FHWA-HRT-11-039, FHWA, U.S. Department of Transportation, 2011. 5. Chen, L., C. Chen, R. Ewing, C. McKnight, R. Srinivasan, and M. Roe. Safety Countermeasures and Crash Reduction in New York City—Experience and Lessons Learned. Accident Analysis and Prevention. In print, 2012. Retrieved July 23, 2012. http://dx.doi.org/10.1016/j.aap.2012.05.009 Treatment 3. High-Intensity Activated Crosswalk (HAWK) Signals Pedestrian Hybrid Beacon (HAWK Signal) Pedestrian hybrid beacons, also known as HAWK beacons (short for High-Intensity Activated crossWalK beacon), were developed by Tucson traffic engineer Dr. Richard Nassi in the late 1990s as a means of providing safe pedestrian crossings where minor streets intersected with Figure H-22. Pedestrians cross at a crosswalk enhanced with a pedestrian hybrid beacon in Phoenix, Arizona. Source: pedbikeimages.org

122 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments major arterials (1, 2). The first pedestrian hybrid beacon was installed in Tucson in 2000. The pedestrian hybrid beacon (PHB) was considered an experimental treatment until 2009, when it was included for the first time in the Manual on Uniform Traffic Control Devices (MUTCD). Today, PHBs are widely used in Tucson, and, as of 2012, have been installed in Georgia, Minnesota, Virginia, Arizona, Alaska, and Delaware (3). Evaluation of Pedestrian Hybrid Beacons A 2006 report authored by Fitzpatrick et al. and published by TRB, TCRP Report 112/NCHRP Report 562: Improving Pedestrian Safety at Unsignalized Crossings, evaluated various midblock crossing treatments, including the PHB. The researchers used trained data collectors and video recordings to collect motorist and pedestrian behavior data at five PHB sites in Tucson, Arizona. Post-treatment data were collected for staged and non-staged pedestrians, and measures of effec- tiveness such as pedestrian crosswalk compliance, pedestrian–vehicle compliance, and motorist yielding were used to evaluate the safety performance of the treatments. Results from the five PHB sites showed an average of 97 percent motorist yielding across all sites, comparable to the other treatments in the red signal or beacon category (see Figure H-23). Nearly all of the red signals or beacons studied were used on high-volume, high-speed arterial streets. Based on the results of this study, the researchers recommended the addition of a red signal or beacon to the MUTCD, given that no such treatment had yet been included (4). A 2010 report by Fitzpatrick and Park published by the Federal Highway Administration evalu- ated the safety effectiveness of the PHB at 21 sites in Tucson, Arizona. The researchers used colli- sion data for the 3 years pre-treatment and for the 3 years following treatment, as well as nearby untreated reference sites, in order to calculate reduction in expected collisions using the Empirical Bayes method. Results of the analysis showed a statistically significant reduction in total crashes of 29 percent as well as a statistically significant 69 percent reduction in pedestrian crashes. There was a 15 percent reduction in severe crashes; however, this result was not statistically significant. The authors concluded that the PHB was effective at leading to a reduction in total collisions at treatment sites. However, there are two main shortcomings of the Fitzpatrick and Park study: (1) the crash sample size was quite small, particularly for the after period and (2) all of the sites were from one single city (Tucson, AZ), and results may vary in other cities (1). A 2013 TRB paper by Deng, Ni, and Li investigated the safety and effectiveness of four mid- block crossing signal controls (pedestrian actuated (PA), pedestrian light controlled (PELICAN), Source: Federal Highway Administration, 2010 (1). Figure H-23. Motorist yielding percentages by countermeasure type. The PHB, or HAWK, beacon is shown second from the left.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 123 high intensity activated crosswalk (HAWK), and pedestrian user-friendly intelligent (PUFFIN)) during the clearance interval phase of the signal operation. The researchers hypothesized that the safety benefit of these treatments may be minimized because pedestrians are not always crossing during the green phase of the signal but in fact are often crossing during the clearance interval phase. Results for the HAWK and PUFFIN signals show that they are better at balancing efficiency and safety for all road users. The HAWK signal was observed to have a “satisfactory performance” with “low” pedestrian flow; however, as pedestrian flow increased, so did the conflicts (6). Pulugurtha and Self (2013) evaluated the safety effects of PHBs at three sites in Charlotte, North Carolina. The researchers collected data from pedestrian crossings during weekday morning and evening peak times at five time points: before the installation, and at 1, 3, 6, and 12 months follow- ing installation. The chosen measures of effectiveness were average traffic speed, the percentage of yielding motorists, the proportion of pedestrians trapped mid-crossing, and pedestrian–vehicle conflicts. An analysis of the results showed an increase in the percentage of yielding motorists, a decrease in the percentage of trapped pedestrians, and a decrease in pedestrian–vehicle conflicts at all three sites; however, these changes were significant at only one of the three sites. At the same time, a statistically significant increase in average vehicle speed was also observed at one of three sites. An analysis of pre- and post-installation crash data showed no significant change in pedestrian–vehicle crashes, although the sample size was small in both cases. The results also indicated that changes in pedestrian and motor vehicle actions were more consistent after the PHBs had been in place for 3 months or more. Overall, the PHBs were effective at increasing motorist yielding and reducing trapped pedestrians and pedestrian–vehicle conflicts (7). References 1. Fitzpatrick, K. and E. S. Park. Safety Effectiveness of the HAWK Pedestrian Crossing Treatment. Publication FHWA-HRT-10-042. FHWA, U.S. Department of Transportation, 2010. This study has also been summarized in Fitzpatrick, K., S. Chrysler, R. Van Houten, W. Hunter, and S. Turner. Evaluation of Pedestrian and Bicycle Engineering Countermeasures: Rectangular Rapid-Flashing Beacons, HAWKs, Sharrows, Crosswalk Markings, and the Development of an Evaluation Methods Report. Publication FHWA-HRT-11-039, FHWA, U.S. Depart- ment of Transportation, 2011. 2. Nassi, R. B., and M. J. Barton. New Traffic Control for an Old Pedestrian Crossing Safety Problem. APWA Reporter, Vol. 75, No. 6, 2008, pp. 44–49. Source: Hunter-Zaworski and Miller, 2012. (8 ) Figure H-24. A pedestrian hybrid beacon installed in Portland, Oregon.

124 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments 3. Chalmers, M. “New Traffic Signals Make It Safer for Pedestrians.” USA Today, 8/9/2010. 4. Fitzpatrick, K., S. Turner, M. Brewer, P. Carlson, B. Ullman, N. Trout, E. S. Park, J. Whitacre, N. Lalani, and D. Lord. TCRP Report 112/NCHRP Report 562: Improving Pedestrian Safety at Unsignalized Crossings. Trans- portation Research Board of the National Academies, Washington, D.C., 2006. 5. Van Houten, R., R. Ellis, and E. Marmolejo. Stutter Flash Light-Emitting-Diode Beacons to Increase Yielding to Pedestrians at Crosswalks. In Transportation Research Record: Journal of the Transportation Research Board, No. 2073, Transportation Research Board of the National Academies, Washington D.C., 2008, pp. 69–78. 6. Deng, T., Y. Ni, and K. Li. Pedestrian Crossings at Mid-Block Locations: A Comparative Study of Existing Signal Operations. Presented at the 92nd Annual Meeting of the Transportation Research Board, Washington, D.C., 2013. 7. Pulugurtha, S. S., and Self, D. R. (2013). Pedestrian and Motorists’ Actions at Pedestrian Hybrid Beacon Sites: Findings from a Pilot Study. International Journal of Injury Control and Safety Promotion, (ahead-of-print), 1–10. 8. Hunter-Zaworski, K., and J. Miller. “Evaluation of Alternative Pedestrian Traffic Control Devices,” FHWA- OR-RD-12-09, March, 2012. Treatment 4. Rectangular Rapid Flashing Beacons (RRFBs) The rectangular rapid flash beacon (RRFB) is a type of amber LED installed to enhance pedes- trian crossing signs at midblock crossings or unsignalized intersections. RRFBs can be automated or pedestrian actuated and feature an irregular, eye-catching flash pattern to call attention to the presence of pedestrians. Research has demonstrated that installing RRFBs on roadside pedestrian crossing signs significantly increases motorist yielding behavior. The RRFB was given interim approval as a crossing sign enhancement by the FHWA in 2008 (1). A 2008 Transportation Research Record article by Van Houten, Ellis, and Marmolejo studied the effectiveness of RRFBs (referred to as stutter-flash LED beacons in the article) in increasing motorist yielding behavior. RRFBs were installed at two Miami-Dade County, Florida, multilane crosswalks. Baseline data were collected pre-treatment, and, during the post-treatment phase, the researchers alternated the activation of the beacons at the sites in order to take further con- trol measurements. Observers measured the numbers of yielding motorists, vehicle-pedestrian conflicts, trapped pedestrians, and motorist yielding distance. At the two crosswalks, motorist yielding to resident pedestrians increased from 0 percent and 1 percent to 65 percent and 92 per- cent, respectively. There was also a reduction in the number of vehicle-pedestrian conflicts and Measure of Effectiveness Site Before After percent Change p-value Percent Drivers Yielding (Staged Crossings, Daytime) NW 67th & Main St. 4.2 (n=2330) 55.2 (n=2131) +51 0.01 (daytime and nighttime combined at this site) S. Bayshore & Darwin 4.1 (n=2075) 60.1 (n=1361) +56 0.01 (daytime and nighttime combined at this site) Percent Drivers Yielding (Staged Crossings, Nighttime) NW 67th & Main St. 4.4 (n=703) 69.8 (n=223) +65.4 See above. S. Bayshore & Darwin 2.5 (n=139) 66 (n=225) +63.5 See above. Percent Drivers Yielding (Resident Crossings) NW 67th & Main St. 12.5 (n=137) 73.7 (n=259) +61.2 0.001 S. Bayshore & Darwin 5.4 (n=200) 83.4 (n=111) +78 0.001 Percent of Pedestrians Trapped in Roadway NW 67th & Main St. 44 0.5 -43.5 <0.01 Percent Of Vehicle-Pedestrian Conflicts NW 67th & Main St. 11 2.5 -8.5 <0.05 S. Bayshore & Darwin 5.5 0 -5.5 <0.01 Source: Pecheux et al., 2009. (3 ) Table H-3. Measures of effectiveness measured by researchers in an evaluation of RRFBs, Miami, Florida.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 125 trapped pedestrians, leading the authors to conclude that the stutter-flash beacon was effective in increasing pedestrian safety at multilane crosswalks (2). A 2009 report by Pecheux, Bauer, and McLeod gave the results of an evaluation of RRFBs at two sites in Miami, Florida. The study team used the following measures of effectiveness (MOEs) to assess the effect of the RRFB on pedestrian and driver behavior: the percentage of pedestrians trapped in the roadway, the percentage of drivers yielding to pedestrians, and the percentage of pedestrian–vehicle conflicts. The researchers found statistically significant improvements in all of the studied MOEs. Table H-3 gives a summary of the results (3). The researchers concluded that the RRFB offered clear safety benefits, and it was placed into the category of highly effective countermeasures (3). A 2009 evaluation of the Pedsafe II project in San Francisco used video observation and inter- cept surveys to collect pre- and post-treatment data to evaluate the effectiveness of 13 counter- measures deployed at 29 sites throughout San Francisco, California. As part of the project, two types of flashing beacons were evaluated: one that was activated by pedestrians and a second that automatically detected pedestrians using infrared technology. The flashing beacons were installed at one uncontrolled crosswalk each in order to assess their effectiveness. Based on pre- and post-treatment video recordings of pedestrian and driver behavior at the site, the push-button activated beacon led to a significant reduction in vehicle/pedestrian conflicts, from 6.7 per cent pre- treatment to 1.9 percent post-treatment, as well as a significant increase in vehicle yielding, from 70 percent pre-treatment to 80 percent post-treatment. It was also noted that only 17 percent of pedestrians activated the beacon, although an additional 27 percent of pedestrians crossed Source: Pecheux et al., 2009. (3 ) Figure H-25. A pedestrian crosses in a crosswalk where pedestrian crossing signs have been enhanced with RRFBs, Miami, Florida. Motorist response Before After Total Full stop 21 (1.9)1 217 (27.3) 238 (12.4) Major direction change 0 (0.0) 5 (0.6) 5 (0.3) Slows 5 (0.5) 65 (8.2) 70 (3.7) No change 1096 (97.7) 508 (63.9) 1604 (83.7) Total 1122 (58.5)2 795 (41.5) 1917 (100.0) 1Column percent 2Row percent Source: Hunter et al., 2009. (5) Table H-4. Motorist responses during interactions with bicyclists and pedestrians before and after RRFB installation.

126 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments when the beacon was activated. The automated flashing beacon led to a significant reduction in vehicle/pedestrian conflicts (from 6.1 percent pre-treatment to 2.9 percent post-treatment), a significant reduction in the number of trapped pedestrians (from 4.1 percent pre-treatment to 0 percent post-treatment), and a significant increase in vehicle yielding (from 82 percent pre- treatment to 94 percent post-treatment). Of the 13 countermeasures tested, both the push-button and automated flashing beacons were among the 6 countermeasures from this study that were considered the most effective in increasing pedestrian safety (4). A 2009 report summarized the effects of installing a pedestrian-activated rectangular rapid flash beacon (RRFB) at the location of one uncontrolled trail crossing at a busy (15,000 ADT), four-lane urban street in St. Petersburg, Florida. The researchers used a mounted video camera to collect pre- and post-treatment data about pedestrian and driver interactions at the trail crossing. An analysis of the data showed a statistically significant reduction in trail user crossing delay and pedestrian yielding, as well as a statistically significant increase in motorist yielding (from 2 percent pre-treatment to 35 percent post-treatment and 54 percent when the beacon was activated) and ability of pedestrians to cross the entire intersection (from 82 percent pre- treatment to 94 percent post-treatment). The researchers concluded that there was an increase in safety at the intersection as a result of installing the RRFB (5). A 2011 report by the Federal Highway Administration by Shurbutt and Van Houten reported on the effects of installing a yellow rectangular rapid-flashing beacon (RRFB) at 22 multilane, uncontrolled crosswalks in St. Petersburg, Florida; Washington, D.C.; and Mundelein, Illinois. The study compared the performance of RRFBs to traditional overhead yellow flashing beacons and a side-mounted traditional yellow beacon at two of the sites. They also compared the perfor- mance of two beacons (one facing each direction of traffic) to four beacons (two per approach on both sides of the road). The researchers measured driver yielding behavior and pedestrian/ vehicle conflicts at baseline (pre-treatment) and compared it to post-treatment data collected eight times over the following 2 years to assess long-term effects. On average across all sites, 4 percent of drivers yielded to pedestrians pre-treatment, while at 2-year follow-up, an average of 84 percent of drivers yielded to pedestrians at all sites, demonstrating the measure’s main- tenance of effect over time (6). Figure H-26. RRFB used at a crosswalk in a St. Petersburg, Florida, evaluation. Source: Hunter et al., 2009. (5)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 127 The RRFB also produced increases in driver yielding behavior at the two sites where its per- formance was compared to overhead and side-mounted beacons. At the site of the overhead beacon, motorist yielding increased from 15.5 percent with the overhead beacon to 78.3 percent when two RRFBs were installed and to 88 percent when four RRFBs were installed. At the site of the side-mounted beacon, motorist yielding increased from 12 percent with the side-mounted beacon to 72 percent with the installation of two RRFBs. Data collected at night showed an increase in driver yielding behavior from 4.8 percent pre-treatment to 84.6 percent (two-beacon RRFB) and 99.5 percent (four-beacon) post-treatment. The authors also compared the perfor- mance of beacons aimed parallel to the roadway to beacons aimed toward the eyes of drivers upon approach, a measure that increased yielding behavior. The authors concluded that the RRFB appeared to be an effective tool for greatly increasing the number of drivers yielding to pedestrians at uncontrolled crosswalks (6). A 2011 Oregon Department of Transportation report by Ross, Serpico, and Lewis evaluated RRFB installation at two Bend, Oregon, crosswalks. Previous to the installation of the RRFBs, motorist yield rates were 23 percent and 25 percent at the intersections; these rates increased to 83 percent at both sites following treatment. Based on their experience, the authors gave 11 sug- gestions for the installation of RRFBs and their evaluation (7). Six locations in Calgary, Canada, with RRFBs and traffic volumes ranging from 4,800 to 14,600 on streets with one to five lanes were observed for driver yielding rates with staged crossings. The RRFBs were observed to increase driver yielding by an average of 15 percent, to nearly 100 percent motorist yielding at the majority of the study sites. Overall the average motorist yielding increased from 83 to 98 percent following RRFB installation (8). Fitzpatrick et al. in 2014 reported on driver yielding rates at multilane crossings in Texas with PHBs, RRFBs, and traffic control signals (TCSs) installed. In this study, the researchers also used Figure H-27. Pedestrian sign enhanced with RRFBs. Source: Shurbutt and Van Houten, 2011. (6 )

128 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments staged pedestrians and employed the same protocol for crossing and observing driver yielding across all sites to ensure more comparable study conditions. Yielding rates for 22 RRFB sites varied more across different cities (34 to 92% yielding) compared to yielding rates for TCS sites (7 sites, 95 to 100% yielding) or PHB sites (32 sites, 73 to 92% yielding). Yielding rates also varied by one- or two-way cross sections, with a decreasing yielding effect for two-way compared to one-way, even controlling for total crossing distance (and city effects). Total crossing distance also had an independent negative effect on driver yielding at RRFBs. Interestingly, there was a correlation of increased yielding with higher speed limits, even when two anomalous sites with low volumes and very low driver yielding rates were removed from the analysis. Closer analysis revealed that the results were related to very slight differences in yielding (1 percentage point between the two averages) on roads with posted speeds of 40 versus 35 mph, and were con- sidered not practically significant. All of the RRFB sites in this study also had School Crossing signs, which the researchers thought could have contributed to an overall average yielding of 86 percent, which they indicated was higher than national averages (9). Foster, Monsere, and Carlos (2014) reported on motorist yielding and pedestrians’ use of RRFBs when they were activated by the pedestrians versus not activated for an urban and a suburban arterial in Portland, Oregon: • Site 1 is a five-lane, 35 mi/h urban arterial, with 30,700 vpd, and a narrow median refuge. Three RRFBs are installed on each side of the crosswalk, one on the side, one in the median, and one overhead, facing each direction, for a total of six pedestrian-activated RRFBs at the crosswalk site. There is also one more RRFB in advance of the crosswalk facing each direction of traffic for a total of eight RRFBs surrounding the crosswalk. Motorist yielding was 92% when beacons were activated and 75% when beacons were not activated. There is high pedestrian crossing activity at this location, with transit stops on each side, and over 200 activations of the RRFB each weekday. • Site 2 is a TWLTL lane, 40 mph, suburban arterial, with 26,400 vpd, a median island, and a Z crossing (Danish offset, a type of path in the median that directs pedestrians to face oncoming traffic before completing their crossing). Four RRFB assemblies were implemented at this loca- tion. Motorist yielding was 91% when the RRFBs were activated and 45% when not activated. Pedestrian activity is also high at this location. Pedestrians activated the beacon 94 percent of the time at Site 1 and 83 percent of the time at Site 2. Researchers also documented that motor- ists yielded more often to pedestrians in the second stage of their crossing at both locations. At Site 2 (the Danish offset), 82 percent of pedestrians who crossed the roadway chose to use the crosswalk, which compared favorably with a 71 percent compliance rate for marked midblock crosswalks in general (10). Source: Ross, Serpico, and Lewis, 2011. (7 ) Figure H-28. RRFB installation in Bend, Oregon.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 129 References 1. Federal Highway Administration. Rectangular Rapid Flash Beacon (RRFB). Publication FHWA-SA-09-009, FHWA, U.S. Department of Transportation, 2009. 2. Van Houten, R., R. Ellis, and E. Marmolejo. Stutter Flash Light-Emitting-Diode Beacons to Increase Yielding to Pedestrians at Crosswalks. In Transportation Research Record: Journal of the Transportation Research Board, No. 2073, Transportation Research Board of the National Academies, Washington D.C., 2008, pp. 69–78. 3. Pecheux, K., J. Bauer, and P. McLeod. Pedestrian Safety Engineering and ITS-Based Countermeasures Program for Reducing Pedestrian Fatalities, Injury Conflicts, and Other Surrogate Measures: Final System Impact Report. Federal Highway Administration, U.S. Department of Transportation, 2009. 4. Hua, J., N. Gutierrez, I. Banerjee, F. Markowitz, and D. R. Ragland. San Francisco Pedsafe II Project Outcomes and Lessons Learned. Presented at the 88th Annual Meeting of the Transportation Research Board, Washington, D.C., 2009. 5. Hunter, W. W., R. Srinivasan, and C. A. Martell. Evaluation of the Rectangular Rapid Flash Beacon at a Pinellas Trail Crossing in St. Petersburg, Florida. Florida Department of Transportation, Tallahassee, Florida, 2009. 6. Shurbutt, J. and R. Van Houten. Effects of Yellow Rectangular Rapid-Flashing Beacons on Yielding at Multi- lane Uncontrolled Crosswalks. Publication FHWA-HRT-10-043, FHWA, U.S. Department of Transporta- tion, 2010. This research is also summarized in Fitzpatrick, K., S. Chrysler, R. Van Houten, W. Hunter, and S. Turner. Evaluation of Pedestrian and Bicycle Engineering Countermeasures: Rectangular Rapid-Flashing Beacons, HAWKs, Sharrows, Crosswalk Markings, and the Development of an Evaluation Methods Report. Publication FHWA-HRT-11-039, FHWA, U.S. Department of Transportation, 2011. 7. Ross, J., D. Serpico, and R. Lewis. Assessment of Driver Yielding Rates Pre- and Post-RRFB Installation, Bend, Oregon. Oregon Department of Transportation, Salem, Oregon, 2011. 8. Domarad, J., P. Grisak, and J. Bolger. Improving Crosswalk Safety: Rectangular Rapid-Flashing Beacon (RRFB) Trial in Calgary. In Calgary 2013-The Many Faces of Transportation-Technical Compendium, 2013. 9. Fitzpatrick, K., M. A. Brewer, and R. Avelar. Driver Yielding at Traffic Control Signals, Pedestrian Hybrid Beacons, and Rectangular Rapid-Flashing Beacons in Texas. Transportation Research Record, 2463, Trans- portation Research Board of the National Academies, Washington, D.C., 2014, pp. 46–54. 10. Foster, N., C. M. Monsere, and K. Carlos. Evaluating Driver and Pedestrian Behaviors at Enhanced, Multilane, Midblock Pedestrian Crossings: Case Study in Portland, Oregon. In Transportation Research Record: Journal of the Transportation Research Board, No. 2464, Transportation Research Board of the National Academies, Washington, D.C., 2014, pp. 59–66. Treatment 5. In-Pavement Warning Lights In-pavement lighting is sometimes used to alert motorists to the presence of a crosswalk at uncontrolled locations. Both sides of the crosswalk are lined with encased raised pavement markers, which sometimes contain LED strobe lighting. In-pavement lighting has shown positive results such as increasing driver compliance and motorist yielding to pedestrians in Washington State but not in Florida (1, 2). At the same time, there are several drawbacks to this method. For example, the whole system must be replaced whenever road surfacing or utility repairs occur. Also, in-pavement lights are generally visible to only the first car in a platoon. Headlights from oncoming traffic may obscure a driver’s view of the entire crossing. Furthermore, in-pavement lighting does not indicate the direction of a pedestrian’s travel or whether people are crossing simultaneously from both sides of the road. Finally, the in-pavement flashers may be difficult to see during bright daylight hours. A 2002 evaluation by Hakkert, Gitelman, and Ben-Shabat studied the effectiveness of in-pavement flashing light systems that automatically detect the presence of pedestrians using infrared sensors at four uncontrolled pedestrian crossings in two Israeli cities. One of the sys- tems, called ARMS (Active Road Marking System for Road Safety), was developed by an Israeli startup company. The second system, called Hercules, was a modified version of an American system. Pedestrian and driver behavior were studied pre- and post-treatment by trained field observers. Analysis of results suggested that the use of the in-pavement flashing light systems could bring about a reduction in vehicle speeds near the crosswalk by 2–5 kph, increase yield- ing to pedestrians by 35 percent at the beginning and 70 percent at the middle of the crosswalk, significantly reduce pedestrian/driver conflict rates to less than 1 percent, and increase the percentage of diverted pedestrians from 50 percent to 90 percent. The authors concluded that,

130 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments owing to the differences across sites and observed treatment effects, it would be advisable to further study the systems before considering them fully ready for field implementation (3). A 2003 paper by Van Derlofske, Boyce, and Gilson gave the results of a field evaluation of in-pavement flashing lights that were installed at a two crosswalk sites at one uncontrolled inter- section in Denville, New Jersey. The site was chosen for safety improvements by the Department of Transportation because it posed crossing difficulties for pedestrians. Successive improve- ments were made, and evaluations were carried out between treatment phases. The first evalu- ation was made pre-treatment, when there was only one, eroded, standard crosswalk marking. The second evaluation was made in 2000, when a second crosswalk was added and both cross- walks were striped. The final evaluation was made later in 2000, following the installation of an in-pavement flashing lights system with automatic, microwave pedestrian detectors. Follow-up evaluations were made at 9 and 12 months following the treatments. Analysis of the results of adding an in-pavement flashing light system indicated that it enhanced the noticeability of the crosswalk, reduced the mean speed at which vehicles approached the crosswalk, and reduced the mean number of vehicles failing to yield to a waiting pedestrian. The researchers also noted important safety benefits from using high-visibility crosswalk markings (4). A 2006 article by Nambisan et al. summarized the effectiveness of an in-pavement flashing light system installed at one uncontrolled crosswalk in the Las Vegas metropolitan area in Nevada. The researchers collected data on driver behavior (yielding, vehicle speeds, yielding distance, and vehicle/pedestrian conflicts) before and after treatment installation and compared the data to see if mean values differed statistically at 95 percent confidence levels. Analysis showed that the system appeared to be effective at increasing driver yielding behavior. There was a statistically significant reduction in mean driver speed when pedestrians were crossing or waiting to cross. While yielding distance was increased by 9 ft in one direction, it decreased by 20 ft in the opposite direction, perhaps due to driver confusion. There was no statistically significant reduction in pedestrian/vehicle conflicts. The authors concluded that the in-pavement lighting solution did appear to have pedestrian safety benefits at a low-traffic-volume location (5). A 2011 report from Vermont found that while the system installed increased yielding and decreased approach speeds, it was removed due to multiple malfunctions and damage during the winter (before an evaluation could be completed) and the DOT’s preference for a treatment Figure H-29. In-pavement flashing lights used at a crosswalk in Vermont. Source: Van Derlofske et al., 2003. (4)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 131 that would be visible during all seasons. Also, in-pavement lights are generally visible to only the first car in a platoon. Headlights from oncoming traffic may obscure a driver’s view of the entire crossing. Furthermore, in-pavement lighting does not indicate the direction of a pedestrian’s travel or whether people are crossing simultaneously from both sides of the road. Finally, the in-pavement flashers may be difficult to see during bright daylight hours (6). In 2010, Thomas conducted a critical review of previous studies that had been conducted on the effects of in-pavement flashing lights. A total of nine studies were reviewed primarily with respect to safety effectiveness of these devices on driver and pedestrian behavior and resulting conflicts (7). Thomas summarizes the findings of this critical review of In-Roadway Warning Lights (IRWL): In conclusion, while motorist yielding to pedestrians has improved to varying degrees at most loca- tions examined, yielding may not be raised to a sufficiently high degree at locations with poor pedestrian conditions. The effects of IRWL on traffic speeds and pedestrian use of the crosswalks is not at all clear, as results have varied among the studies. Positive effects observed may also degrade over time as found in several studies. Unfortunately, it is not clear from the evaluation studies reviewed, under what conditions flashing crosswalk treatments may be most beneficial over the long term, while not recommended for others. Clearly, the location should be carefully evaluated prior to implementation, and the specifics should be assessed thoroughly to determine if this treatment, alone or in combination with other treatments, is the best solution for a particular situation. Some communities have removed IRWLs due to both safety and efficiency reasons. If installed, the treatment should be carefully evaluated and monitored long term for effects on pedestrian safety and mobility. Ideally, comparison locations with similar environmental and user characteristics would be used to help control for time trends and other unknown effects (7). References 1. Huang, H. An Evaluation of Flashing Crosswalks in Gainesville and Lakeland. Florida Department of Trans- portation, Tallahassee, Florida, 2000. 2. Godfrey, D. and T. Mazella. Kirkland’s Experience with In-Pavement Flashing Lights at Crosswalks. Presented at ITE/IMSA Annual Meeting, Lynnwood, Washington, 1999. 3. Hakkert, A. S., V. Gitelman, and E. Ben-Shabat. An Evaluation of Crosswalk Warning Systems: Effects on Pedestrian and Vehicle Behaviour. Transportation Research, Vol. 5, No. 4, 2002, pp. 275–292. 4. Van Derlofske, J. F., P. R. Boyce, and C. H. Gilson. Evaluation of In-Pavement, Flashing Warning Lights on Pedestrian Crosswalk Safety. International Municipal Signal Association Journal, Vol. 41, No. 3, 2003. 5. Nambisan, S. S., G. Karkee, and S. S. Pulugurtha. An Evaluation of the Effectiveness of an In-Pavement Flash- ing Light System. Presented at the 85th Annual Meeting of the Transportation Research Board. Washington, D. C., 2006. 6. Kipp, W. M., and J. Fitch. Evaluation of SmartStud™ In-Pavement Crosswalk Lighting System and BlinkerSign® (Report No. 2011-3). State of Vermont, Agency of Transportation, 2011. Accessed 2015 at http://vtransplanning. vermont.gov/sites/aot_program_development/files/documents/materialsandresearch/completedprojects/ Evaluation_of_SmartStud_In_Pavement_Crosswalk_web.pdf 7. Thomas, L. J. Safety Effects of In-Pavement Warning Lights or “Flashing Crosswalk” Treatments: A Review and Synthesis of Research. Treatment 6. Pedestrian Refuge Areas Median refuge islands, sometimes referred to as a center islands, refuge islands, or pedestrian islands, are raised areas that help protect pedestrians who are crossing the road at intersections and midblock locations. The presence of a median refuge island in the middle of a street or intersection allows pedestrians to focus on one direction of traffic at a time as they cross and gives them a place to wait for an adequate gap between vehicles. Islands are appropriate for use at both uncontrolled and signalized crosswalk locations. Where the road is wide enough and on-street parking exists, center islands can be combined with curb extensions to enhance pedestrian safety (1). A 1994 study by Bowman and Vecellio was conducted to determine the effects of urban and suburban median types on the safety of vehicles and pedestrians. The study involved an analysis of

132 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments 32,894 vehicular crashes and 1,012 pedestrian crashes that occurred in three U.S. cities (Atlanta, Georgia; Phoenix, Arizona; and Los Angeles/Pasadena, California). The authors compared sites that had no median, those that had a raised median, and those with a flush or two-way, left-turn lane. A variety of statistical tests were used, including t-tests, analysis of variance, and the Scheffe multiple comparison test. The authors did not have pedestrian volume data, but used area type (CBD and suburban areas) and land use as surrogate measures for pedestrian activity and devel- oped pedestrian crash prediction models separately for the two area types (2). The results of this analysis provide evidence that having some area of refuge (either a raised median or TWLTL) on an arterial CBD or suburban street provides a safer condition for pedes- trians than having an undivided road with no refuge for pedestrians in the middle of the street. Furthermore, while this study found that suburban arterial streets with raised-curb medians had lower pedestrian crash rates compared to TWLTL medians, this difference was not statis- tically significant. This may be a clear indication that some refuge area in the middle of wide streets is more beneficial to pedestrian safety when crossing streets than having no refuge area. However, the safety benefits for a raised median vs. a TWLTL were not quantified in this study. Based on the study results, Bowman and Vecellio suggest that in CBD areas, whenever possible, divided cross sections should be used due to their lower crash rates for pedestrians and motor vehicles (2). A 1994 study by Claessen and Jones (4) found that replacing a 6 ft (1.8 m) painted median with a wide raised median reduced pedestrian crashes by 23 percent. According to Cairney in Pedestrian Safety in Australia (1999), this conclusion was consistent with Scriven’s finding that pedestrian crash rates for roads with 10 ft (3.05 m) medians were 33 percent lower than for roads with 4 ft (1.2 m) painted medians (3). In 2001, a study by Huang and Cynecki evaluated a variety of traffic calming measures in several U.S. cities, using before-after analysis of pedestrian and motorist behavior as measures of effectiveness. The study included an evaluation of four refuge islands at two unsignalized four-leg inter sections in Sacramento, California. The refuge islands constricted the width of the travel lanes and were expected to reduce vehicle speeds, increase the number of pedestrians for whom motorists yielded, and increase the percentage of pedestrians who crossed in the crosswalk. After installation of pedestrian refuge islands at the four crosswalk locations, the percentage of motorists who yielded to pedestrians increased from 32.6 percent to 42.1 percent, but this was not statistically significant at the 90 percent level, due to relatively small sample sizes of crossing pedestrians. However, there was a statistically significant increase in the percentage of pedestrians who crossed in the crosswalk, from 61.5 percent to 71.9 percent. There was no statistically sig- nificant difference in the pedestrian wait time after the refuge islands were installed. It would be expected that pedestrian wait time would more likely be improved in situations where refuge islands are installed on multi-lane roads (5). Also in 2001, Bacquie, Egan, and Ing conducted a before-after study to evaluate the safety effec- tiveness of raised pedestrian refuge islands. Pedestrian accidents that could have been prevented by a pedestrian refuge island were reduced at 28 sites for which data were available, from 22 in Location Treatment Before After Significant Corvallis, Oregon Refuge island and pavement markings 51.9% (n=79) 78.0% (n=113) No Sacramento, California Refuge islands with zebra crosswalks 4 locations 61.5% (n=314) 71.9% (n=224) Yes (p=0.012) Table from Huang and Cynecki (2001) showing the percentage of pedestrians who crossed in the crosswalk before and after the installation of median islands. (5)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 133 the 3 years before installation to 6 during the 3 years following installation of the refuge islands. However, there were 46 vehicle-island crashes during the after period, which were not possible during the 3 years prior to island installation. The study authors concluded that pedestrian safety had been enhanced by addition of the islands due to the 73 percent reduction in midblock pedes- trian crashes, but overall safety, as reflected in crash frequency, had decreased due to a 136 percent increase in total crashes. It was noted that the decrease in safety related to vehicle-island crashes might be helped by better island design and lane alignment (6). A 2002 study by Zegeer, Stewart, Huang, and Lagerwey that was primarily intended to deter- mine the safety effects of marked vs. unmarked crosswalks on pedestrian crashes provided some insight into the effectiveness of raised medians (7). The 2,000 crossing sites used in the study were uncontrolled crossings at intersection (i.e., no traffic signals or STOP control on inter- section approach of interest) or midblock locations. Zegeer et al. found that the presence of a raised median or crossing island was associated with a significantly lower rate of pedestrian crashes on multi-lane roads having either marked or unmarked crosswalks. This was true at marked as well as unmarked crosswalks. Comparing urban or suburban four- to eight-lane roads with a mini- mum ADT of 15,000 vehicles per day and marked crosswalks, the pedestrian crash rate (pedestrian crashes per million crossings) was 0.74 at crosswalks where there was a raised median and 1.37 for sites without a raised median. For similar sites (multi-lane with ADT above 15,000 veh/day) at unmarked crosswalk locations, the pedestrian crash rate was 0.17 with a raised median and 0.28 for sites without a raised median. Multi-lane road sites that had a center two-way-left-turn lane (TWLTL) or painted (but not raised) median did not correspond to safety benefits for pedestrians compared to multi-lane roads with no medians at all. Thus, this study found that raised medians clearly provide a significant safety benefit to pedestrians on multi-lane roads, particularly on such roads with ADT above 15,000 veh/day (7). A 2003 paper by Kamyab, Andrle, Kroeger, and Heyer discussed the effects of installing a remov- able pedestrian island and pedestrian crossing signs on a two-lane highway in rural Mahnomen County, Minnesota. Researchers collected pre- and post-treatment speed data to assess short- and long-term effects of the treatments. Results showed a statistically significant reduction in Figure H-30. A pedestrian crosses in a zebra crosswalk that has been enhanced with a refuge island in Sacramento, California. Source: Huang and Cynecki, 2001. (5 )

134 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments mean speeds and increase in speed-limit compliance at the treatment site for both the long and short term (8). A 2003 paper by King, Carnegie, and Ewing evaluated the effect of traffic calming measures involving signal, curb, sidewalk, and raised median installation and intersection redesign along a 3,200-foot section of a four-lane suburban roadway in New Jersey. The researchers used pre- and post-treatment on speed and volume counts, pedestrian tracking, video, and photography to evaluate the effect of the treatments on pedestrian safety. Results showed a 2 mi/h decrease in 85th percentile vehicle speed and a 28 percent decrease in pedestrian exposure risk without affecting vehicle volumes. The researchers predicted that $1.7 million would be saved due to avoided collisions over 3 years as a result of the roadway improvements (9). A 2009 report compiled by Pecheux, Bauer, and McLeod gave the results of an evaluation of median refuge islands installed at two signalized intersections in San Francisco, California. The Source: Kamyab et al., 2003. (8) Figure H-31. A removable pedestrian island installed in conjunction with an in-roadway yield to pedestrians sign at a crosswalk in Minnesota. Observed Traffic Mean Speed (mi/h) t-statistic Significant (95%) Speed Compliance % t-statistic Significant (95%) Passenger Cars Before 1152 34.8 -- -- 31 -- -- After-1 1067 29.5 13.49 Yes 58 -12.80 Yes After-2 1331 30.7 11.05 Yes 51 -10.01 Yes Nonpassenger Cars Before 71 37.4 -- -- 24 -- -- After-1 46 28.8 4.11 Yes 65 -4.42 Yes After-2 60 29.5 4.01 Yes 57 -3.84 Yes All Vehicles Before 1237 35 -- -- 30 -- -- After-1 1113 29.5 14.20 Yes 58 -13.68 Yes After-2 1392 30.6 11.02 Yes 51 -10.85 Yes Source: Kamyab et al., 2003. (8 ) Table H-5. Mean vehicle speeds before and after the installation of a removable pedestrian island and pedestrian crossing sign.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 135 researchers measured the percentage of pedestrians trapped in the roadway, the percentage of pedestrian–vehicle conflicts, the percentage of drivers yielding to pedestrians, and the average pedestrian delay before and after the medians were installed. The researchers found no signifi- cant impact on driver yielding, trapped pedestrians, or pedestrian–vehicle conflicts at either of the sites, and a statically significant increase in pedestrian delay at one of the sites. Based on these results, the researchers concluded that the median refuge islands were not very effective at alter- ing driver and pedestrian behaviors at the two San Francisco study sites (10). More recently, a 2012 paper by Pulugurtha, Vasudevan, Nambisan, and Dangeti evaluated four different infrastructure-based countermeasures, including median refuge and Danish offset, com- bined with high-visibility crosswalks at eight different sites in Las Vegas, Nevada. Pre- and post- treatment observations were collected on weekdays to record data regarding the following measures of effectiveness (MOEs): pedestrians trapped in the street, pedestrians looking for vehicles before beginning to cross, pedestrians looking for vehicles before crossing the second half of the street, percent of captured or diverted pedestrians, driver yield behavior and distance, and drivers block- ing the crosswalk. A two-proportion z-test was conducted to determine the statistical significance of post-treatment measurements. For median refuge, there was a statistically significant increase in the proportion of pedestrians who looked for vehicles before beginning to cross, the propor- tion of drivers yielding to pedestrians, and the distance at which drivers yielded to pedestrians. For Source: Pecheux et al., 2009. (10) Figure H-32. Pedestrians making use of a median refuge island in San Francisco. Source: Pulugurtha et al., 2012. (11) Figure H-33. A Danish offset median refuge island as used in Las Vegas, Nevada. This type of offset design is configured so that pedestrians view oncoming traffic as they walk to the second half of the crosswalk.

136 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Danish offset, there was a statistically significant increase in the proportion of diverted pedestrians, proportion of drivers who yielded to pedestrians, and driver yield distance (11). A 2013 TRB paper by Hengel presented the results of a study of a single site in Santa Barbara, California, where a curb extension, pedestrian refuge island, and stop bars were installed. MOEs were crossing delay, number of motorists failing to yield, and distance drivers yield from the crosswalk. Results show that the combination of treatments is effective at reducing wait times to cross, decreasing percentage of vehicles that pass before yielding, and increasing the distance that vehicles yield in advance of the crosswalk (12). References 1. Zegeer, C. V., J. Stutts, H. Huang, M. J. Cynecki, R. Van Houten, B. Alberson, R. Pfefer, T. R. Neuman, K. L. Slack, and K. K. Hardy. NCHRP Report 500: Guidance for Implementation of the AASHTO Strategic Highway Safety Plan, Volume 10: A Guide for Reducing Collisions Involving Pedestrians. Transportation Research Board of the National Academies, Washington, D.C., 2004. 2. Bowman, B. L. and R. L. Vecellio. Effects of Urban and Suburban Median Types on Both Vehicular and Pedestrian Safety. In Transportation Research Record 1445, TRB, National Research Council, Washington, D.C., 1994, pp. 169–179. 3. Cairney, P. Pedestrian Safety in Australia. Publication FHWA-RD-99-093, FHWA, U.S. Department of Transportation, 1999. 4. Claessen, J. and D. Jones. The Road Safety Effectiveness of Wide Raised Medians. Proc. 17th ARRB Conference, 17(5), 1994, pp. 269–284. 5. Huang, H. F. and M. J. Cynecki. The Effects of Traffic Calming Measures on Pedestrian and Motorist Behavior. Publication FHWA-RD-00-104, FHWA, U.S. Department of Transportation, 2001. 6. Bacquie, R., D. Egan, and L. Ing. Pedestrian Refuge Island Safety Audit. Presented at 2001 ITE Spring Con- ference and Exhibit, Monterey, CA, 2001. 7. Zegeer, C. V., R. Stewart, H. Huang, and P. Lagerwey. Safety Effects of Marked Versus Unmarked Crosswalks at Uncontrolled Locations: Executive Summary and Recommended Guidelines. Publication FHWA-RD-01-075, FHWA, U.S. Department of Transportation, 2002. 8. Kamyab, A., S. Andrle, D. Kroeger, and D. Heyer. Methods to Reduce Traffic Speed in High-Pedestrian Rural Areas. In Transportation Research Record: Journal of the Transportation Research Board, No. 1828, Transporta- tion Research Board of the National Academies, Washington, D.C., 2003, pp. 31–37. 9. King, M. R., J. A. Carnegie, and R. Ewing. Pedestrian Safety Through a Raised Median and Redesigned Intersections. In Transportation Research Record: Journal of the Transportation Research Board, No. 1828, Transportation Research Board of the National Academies, Washington, D.C., 2003, pp. 56–66. 10. Pecheux, K., J. Bauer, P. McLeod. Pedestrian Safety Engineering and ITS-Based Countermeasures Program for Reducing Pedestrian Fatalities, Injury Conflicts, and Other Surrogate Measures: Final System Impact Report. Federal Highway Administration, U.S. Department of Transportation, 2009. 11. Pulugurtha, S. S., V. Vasudevan, S. S. Nambisan, and M. R. Dangeti. Evaluating Effectiveness of Infrastructure- Based Countermeasures for Pedestrian Safety. In Transportation Research Record: Journal of the Transportation Research Board, No. 2299, Transportation Research Board of the National Academies, Washington, D.C., 2012, pp. 100–109. 12. Hengel, D. Build It and They Will Yield: Effects of Median and Curb Extension Installations on Motorist Yield Compliance. Presented at the 92nd Annual Meeting of the Transportation Research Board, Wash- ington, D.C., 2013. Treatment 7. Curb Extensions Curb extensions are a way of narrowing the roadway width by extending the curb line or side- walk into the parking line, which results in reduced vehicle speeds, improved visibility between pedestrians and oncoming motorists, and reduced crossing distance for pedestrians. A 1999 presentation by King on the subject of traffic calming evaluated the effect of curb exten- sions on crashes at six locations in New York City. Between 5 and 10 years of collision data were collected for the six curb extension sites. Each crash was given a value based on severity. Over- all, the curb extensions reduced overall severity rate at four of the six intersections, leading to increased pedestrian safety and the more widespread use of curb extensions in New York City (2).

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 137 A 2001 study by Huang and Cynecki involved evaluating curb extensions at eight residential and arterial crosswalk locations in Massachusetts, Washington, North Carolina, and Virginia, based on pedestrian wait time, vehicle speed, and motorist yielding behavior. The researchers employed pre- and post-treatment research design for the sites in Massachusetts and Washington and treatment and control design in North Carolina and Virginia. No significant improvements were found at most of the sample sites after curb extensions were installed. Huang and Cynecki stated that some of the results may have been due to traffic conditions at the study sites. The authors also stated that such devices cannot guarantee that motorists will slow down or yield to pedestrians, or that pedestrians will choose to cross at the crosswalk (3). A 2005 Federal Highway Administration case study analyzed the effect of curb extensions at one uncontrolled intersection in Albany, Oregon. Because there were no pre-treatment data avail- able, nor an appropriate control site, the researchers chose to observe driver behavior at a pedes- trian crosswalk that had a recently installed curb extension on only one side of the intersection, with the untreated curb acting as a control. Measures of effectiveness (MOEs) were the average number of vehicles that passed before a pedestrian could cross, the percent of pedestrians crossing with yield, and the percent of vehicles yielding at the advance stop bar. Difference in means was analyzed using a two-sample t-test. It was found that the curb extension contributed to a signifi- cant reduction in the mean number of vehicles passing a pedestrian before yielding, possibly due to the increased visibility offered by the curb extension. A 20 percent reduction was observed in vehicles stopping at the advance stop bar on the treatment side; however, this was not statistically significant. The researchers suggested that driver behavior, in addition to a lack of appropriate pedestrian facilities, also contributed to the observed failure to yield to pedestrians (4). A 2013 TRB paper by Hengel presented the results of a study of a single site in Santa Barbara, California, where a curb extension, pedestrian refuge island, and stop bars were installed. MOEs Source: Axelson et al., 1999. (1) Figure H-34. Sketch of a curb extension, demonstrating how it increases visibility for both drivers and pedestrians and reduces pedestrian crossing distances and vehicle turn speeds.

138 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments were crossing delay, number of motorists failing to yield, and distance drivers yield from the crosswalk. Results show that the combination of treatments is effective at reducing wait times to cross, decreasing percentage of vehicles that pass before yielding, and increasing the distance that vehicles yield in advance of the crosswalk (5). References 1. Axelson, P. W., D. A. Chesney, D. V. Galvan, J. B. Kirschbaum, P. E. Longmuir, C. Lyons, and K. M. Wong. Designing Sidewalks and Trails for Access, Part I of II: Review of Existing Guidelines and Practices. Washington, D.C., Federal Highway Administration, 1999. 2. King, M. R. Calming New York City Intersections. In Transportation Research Board Circular E-CO19: Urban Street Symposium, TRB, National Research Council, Washington D.C., 2000. 3. Huang, H. F. and M. J. Cynecki. The Effects of Traffic Calming Measures on Pedestrian and Motorist Behavior. Publication FHWA-RD-00-104, FHWA, U.S. Department of Transportation, 2001. 4. Johnson, R. S. Pedestrian Safety Impacts of Curb Extensions: A Case Study. Publication FHWA-OR-DF-06-01, Oregon Department of Transportation, Salem, OR, 2005. 5. Hengel, D. Build It and They Will Yield: Effects of Median and Curb Extension Installations on Motorist Yield Compliance. Presented at the 92nd Annual Meeting of the Transportation Research Board, Wash- ington, D.C., 2013. Treatment 8. Raised Pedestrian Crossings Raised pedestrian crossings tend to be applied most often on two-lane business streets in urban environments and are applied both at intersections and midblock. They are generally installed on lower speed roads because they are designed for speeds in ranges below 35 mph and therefore would not be appropriate on higher speed roads. In 2001, Huang and Cynecki looked at how various pedestrian safety countermeasures, including raised pedestrian crossings, affected the behavior of pedestrians and drivers at three sites in North Carolina and Maryland. Each of the three treatment sites was matched with a control site. Overall, the use of raised crosswalks resulted in lower overall vehicle speeds. At the two North Carolina sites, 50th percentile vehicle speeds were 4.0 to 12.4 mph lower at treatment sites than at control sites. At the Maryland site, 50th percentile vehicle speeds were 2.5 miles per hour lower at treatment sites than at control sites; however, this difference was not statistically significant. At the North Carolina site where the raised crosswalk was installed, there was already an overhead flashing beacon and motorist yielding was significantly higher, while at the other Figure H-35. A raised pedestrian crosswalk. Source: Federal Highway Administration.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 139 North Carolina crosswalk, there were insufficient pedestrian crossings for comparison. At the Maryland site, the difference in motorist yielding was not statistically significant. The authors concluded that raised crosswalks are effective at reducing motor vehicle speeds, especially when combined with an overhead beacon. However, in the case of the intersection with the overhead beacon, it was impossible to gauge how much each countermeasure contributed to motorist yielding behavior (1). In the same study, Huang and Cynecki evaluated the installation of a raised intersection in Cambridge, Massachusetts. Before and after data were collected to assess the impact of the raised intersection on motorist yielding, percentage of pedestrians using the crosswalk, and average pedestrian wait time. There was a significant increase in the percentage of pedestrians who crossed in the crosswalk, from 11.5 percent to 38.3 percent. There was an increase in the percentage of motorists who yielded to pedestrians in the crosswalk, but this increase was not statistically sig- nificant due to small sample size (1). References 1. Huang, H. F. and M. J. Cynecki. The Effects of Traffic Calming Measures on Pedestrian and Motorist Behavior. Publication FHWA-RD-00-104, FHWA, U.S. Department of Transportation, 2001. Other Treatments A 2001 report entitled “Alternative Treatments for At-Grade Pedestrian Crossings” (1) contains a discussion of experimental measures used at uncontrolled crossings. A 2012 presentation by Dougald, Dittberner, and Sripathi detailed an experimental zig-zag pavement marking treatment in Loudoun County, Virginia. The Virginia Department of Trans- portation installed the markings at two locations where pedestrians and bicyclists use the Wash- ington and Old Dominion Trail to cross area highways in 2009. Researchers measured vehicle speeds and driver attitudes pre- and post-treatment and concluded that the use of the markings Figure H-36. Zig-zag pavement markings used in Virginia to call attention to the presence of pedestrians. Source: Dougald et al., 2016. (2)

140 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments increased motorist awareness of the crossings, as evidenced by lower mean vehicle speeds and self-reported yielding behavior. However, surveys revealed limited driver understanding of the markings’ purpose (2). References 1. Lalani, N. Alternative Treatments for At-Grade Pedestrian Crossings. Institute of Transportation Engineers, Washington, D.C., 2001. 2. Dougald, L. E., R. A. Dittberner, and H. K. Sripathi. Safer Mid-Block Environments for Pedestrian and Bicycle Crossings: Experiment with Zig-Zag Pavement Markings. In Transportation Research Record: Journal of the Transportation Research Board, No. 2299, Transportation Research Board of the National Academies, Washington, D.C., 2012, pp. 128–136. Comprehensive Countermeasure Evaluations While researchers have been successful in studying various pedestrian safety countermeasures on an individual basis, there are many cases where researchers have studied the effects of compre- hensive, often citywide, pedestrian safety programs. Such programs often involve simultaneous engineering, education, and enforcement efforts. While comprehensive programs are frequently successful at improving pedestrian safety, it can be impossible to measure the contribution of any individual countermeasures to the overall improvement in the pedestrian environment. Nonetheless, large-scale safety programs can publicly demonstrate how changes in the pedes- trian environment can lead to significantly improved streets for all travelers. In 1989, Malenfant and Van Houten studied the effects of a program that combined the instal- lation of advance stop lines with signs, education, and enforcement, as a means of increasing motorist yielding at 34 crosswalks in three Canadian cities in Newfoundland and New Brunswick. Baseline data were collected in each of the cities prior to treatment. Post-treatment data analysis revealed that motorist yielding at follow up increased from 54 percent to 81 percent in St. John’s (Newfoundland), from 9 percent to 68 percent in Fredericton (New Brunswick), and from 44 percent to 71 percent in Moncton-Dieppe (New Brunswick) (1). In 1999–2000, the small town of Storuman in Sweden reconstructed an arterial road that passed through its center, adding a collection of various traffic calming measures: pedestrian walkways, traffic islands, chicanes (“Danish buns”), a roundabout, and a bicycle path. At the same time, driver conduct codes became stricter, requiring drivers to yield to pedestrians at marked crosswalks at all times. Based on analysis of pre- and post-treatment observations, including video recordings at treatment sites, it was determined that the treatments had significant effects on mode and route choice, increasing pedestrian and bicyclist flow and perception of safety, and reducing the speed of motor vehicles. While fall injury incidence increased slightly in the post-treatment study period (it was theorized to be an effect of greater pedestrian use), collision data analysis indicated that safety increased not only along the arterial road, but also on adjacent roads (2). A 2011 study by Savolainen, Gates, and Datta evaluated the impact of a 2006 comprehensive education, enforcement, and engineering countermeasure program implemented in Detroit, Michigan, on improving pedestrian safety throughout the city. In addition to education and enforcement measures implemented in 2008 and 2009, the City of Detroit installed pedestrian countdown timers at 362 intersections, seven high-intensity activated crosswalk (HAWK) sig- nals, traversable medians, and new pavement markings at crosswalks. Additional pedestrian countdown timers, HAWK signals, and rectangular rapid flash beacons (RRFB) were planned for 2010. Collision data were analyzed from 2004 to 2009, with the period from 2004 to 2007 serving as the pre-treatment phase, and the period from 2008 to 2009 representing the treatment implementation phase. During the implementation period, there was a 17.9 percent reduction in pedestrian crashes, a 20.1 percent reduction in pedestrian injuries, and a marginal (1.7 percent)

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 141 increase in pedestrian fatalities, compared to statewide collision data. While the comprehen- sive program made determining specific program impacts difficult to analyze, the program has accompanied an improved pedestrian safety environment throughout the city (3). References 1. Malenfant, L., and R. Van Houten. Increasing the Percentage of Drivers Yielding to Pedestrians in Three Canadian Cities with a Multifaceted Safety Program. Health Education Research, Vol. 5, No. 2, 1989, pp. 275–279. 2. Leden, L., P. E. Wikstrom, P. Garder, and P. Rosander. Safety and Accessibility Effects of Code Modifications and Traffic Calming of an Arterial Road. Accident Analysis and Prevention, Vol. 38, No. 3, 2006, pp. 455–461. 3. Savolainen, P. T., T. J. Gates, and T. K. Datta. Evaluation of the Effectiveness of a Comprehensive Pedestrian Safety Program in a Major City. Presented at 90th Annual Meeting of the Transportation Research Board, Washington, D.C., 2011. Conclusions The primary source used to compile Appendix H was a UNC HSRC produced draft report of the Evaluation of Pedestrian-Related Roadway Measures: A Summary of Available Research by Mead, Zegeer, and Bushell2. Selected sections from this resource were extracted or adopted for use in this literature review. A summary is given below of the results of the literature review for the eight selected countermeasures: Unsignalized Pedestrian Crosswalk Signs and Pavement Markings, Including Advanced Yield or Stop Markings and Signs Many studies (roughly 30 identified) were found that evaluated pedestrian crosswalk signs, pavement markings, or advanced yield or stop markings and signs. The vast majority have been behavioral or observational studies. Only one study was able to develop a CMF for speed-limit reduction signs. Another study, in Los Angeles, found a crash effect of removing marked cross- walks at locations that did not meet their crosswalk policy. More research is definitely needed to better understand the safety effects of various types of signs and pavement markings for a variety of conditions. High-Visibility Crosswalk Marking Patterns Two studies were found that examined the crash effects of high-visibility crosswalk marking patterns, and two studies developed estimated CMFs. However, it is difficult to account for all confounding variables for such evaluations, and this is an issue of much interest to city traffic Table H-6. Crash trends before and during a citywide intervention in Detroit, Michigan. Period Pedestrian Crashes City of Detroit Pedestrian Crashes Michigan (Exclusive of Detroit) Crashes Injuries Fatalities Crashes Injuries Fatalities Before Intervention 576 464 29 1817 1772 107 During Intervention 473 371 30 1652 1616 88 Percent Reduction 17.9 20.1 -1.7 9.1 8.8 17.4 Source: Savolainen et al., 2011. (3) 2Mead, J., C. Zegeer, and M. Bushell (2013). Evaluation of Pedestrian-Related Roadway Measures: A Summary of Available Research. Chapel Hill, North Carolina: Federal Highway Administration.

142 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments engineers concerned with whether it is cost-effective to use high-visibility crosswalk marking pat- terns. More research is needed to determine safety effects of this treatment. High-Intensity Activated Crosswalk (HAWK) Signals Very few studies were found that examined HAWK signals; only one (Fitzpatrick and Park) of these studies developed CMFs and that study involved a limited number of sites, primarily from Tucson, Arizona. More work is needed to determine the safety effects of this treatment. Since this treatment is now in the 2009 MUTCD, and many city and state agencies have been installing or considering the installation of this measure, it would be valuable to better under- stand the crash effects in other cities and under a wide range of conditions. Rectangular Rapid Flashing Beacons (RRFBs) All six studies found on RRFBs were behavioral studies, but none developed CMFs. More research is needed to quantify the safety effects of this treatment, particularly since the RRFBs have become quite popular, and many are being installed in cities across the United States. In-Pavement Warning Lights Of the studies that were found related to in-pavement warning lights (IPWLs), none attempted to develop CMFs. Furthermore, there is evidence of malfunction and maintenance issues with some of these devices, which could cause difficulty in conducting a crash-based evaluation. However, there may also be reasons why CMF information would be of value for these devices if a suitable sample size of treatment locations could be found. Pedestrian Refuge Areas Several studies of pedestrian refuge areas have been identified, but only two developed CMFs, and these were from the mid-1990s, which had somewhat differing results. More work is needed to determine the safety effects of this treatment, particularly the promising nature of this treat- ment, combined with the large number of refuge islands that have been installed in recent years. Curb Extensions Of the four studies found on curb extensions, none developed CMFs. Curb extensions are a promising geometric treatment at unsignalized crossings, since they can result in improved sight distance for pedestrians, reduced speeds of oncoming and turning vehicles, and shorter crossing distance across the street for pedestrians. However, the CMF is not known for this treatment. Raised Pedestrian Crossings Only one study was found that examined raised pedestrian crossings, and no CMFs were developed. More work is needed to determine the safety effects of this treatment. Needs for Further Research While there were a few research studies that developed crash effects (CMFs) for these eight pedestrian crossing treatments, most of these studies used behavioral, speed, and/or conflict mea- sures only and few CMF values were determined. In summary, regarding the need for evaluating these pedestrian treatments for NCHRP Project 17-56, there is clearly a need to better quantify the CMFs of the eight proposed treatments listed above under various traffic and roadway conditions

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 143 (e.g., two-lane vs. multi-lane, high-speed vs. low speed, various area types, and road classes, etc.), to the extent possible. Research studies reviewed as a part of this literature review are listed in the summary tables that follow. Additional References Fitzpatrick, K., W. H. Schneider, IV, and E. S. Park. Operation and Safety of Right-Turn Lane Designs. In Trans- portation Research Record: Journal of the Transportation Research Board, No. 1961, Transportation Research Board of the National Academies, Washington, D.C., 2006, pp. 55–64. Garder, P. Pedestrian Safety at Traffic Signals. Accident Analysis and Prevention, Vol. 21, No. 5, 1989, pp. 435–444. Jonsson, T., J. N. Ivan, and C. Zhang. Crash Prediction Models for Intersections on Rural Multilane Highways: Differences by Collision Type. In Transportation Research Record: Journal of the Transportation Research Board, No. 2019, Transportation Research Board of the National Academies, Washington, D.C., 2007, pp. 91–98. Lalani, N. Road Safety at Pedestrian Refuges. Traffic Engineering & Control, Vol. 18, No. 9, 1977, pp. 429–431. Moore, V. M. and A. J. McLean. A Review of Pedestrian Facilities. Office of Road Safety, South Australian Depart- ment of Transport, 1995. Nambisan, S. S., S. S. Pulugurtha, V. Vasudevan, M. R. Dangeti, and V. Virupaksha. Effectiveness of Automatic Pedestrian Detection Device and Smart Lighting for Pedestrian Safety. In Transportation Research Record: Journal of the Transportation Research Board, No. 2140, Transportation Research Board of the National Academies, Washington, D.C., 2009, pp. 27–34. Vasudevan, V., S. S. Pulugurtha, S. S. Nambisan, and M. R. Dangeti. Effectiveness of Signal-Based Countermeasures for Pedestrian Safety: Findings from a Pilot Study. In Transportation Research Record: Journal of the Transporta- tion Research Board, No. 2264, Transportation Research Board of the National Academies, Washington, D.C., 2011, pp. 44–53. Van Houten, R., R. A. Retting, J. Van Houten, C. M. Farmer, and J. E. L. Malenfant. Use of Animation in LED Pedestrian Signals to Improve Pedestrian Safety. ITE Journal, Vol. 69, No. 2, 1999, pp. 30–38. Wang, X., M. Abdel-Aty, and P. A. Brady. Crash Estimation at Signalized Intersections: Significant Factors and Temporal Effect. In Transportation Research Record: Journal of the Transportation Research Board, No. 1953, Transportation Research Board of the National Academies, Washington, D.C., 2006, pp. 10–20. Wang, X. and M. Abdel-Aty. Temporal and Spatial Analysis of Rear-end Crashes at Signalized Intersections. Accident Analysis and Prevention, Vol. 38, 2006, pp. 1137–1150. Zegeer, C., D. Henderson, R. Blomberg, L. Marchetti, S. Masten, Y. Fan, L. Sandt, A. Brown, J. Stutts, and L. Thomas. Evaluation of the Miami-Dade Pedestrian Safety Demonstration Project. National Highway Traf- fic Safety Administration. Washington, D.C., 2008. Summary Tables Study Title Authors Year Summary CMFs? Application Notes Signs The Effects of Innovative Pedestrian Signs at Unsignalized Locations: A Tale of Three Treatments Huang et al. 2000 Study of in-street pedestrian signs "State Law - Yield to Pedestrians in Crosswalks in Your Half of the Road" installed in NY and Portland, OR. No Year 2 Field Evaluation of Experimental "In- Street" Yield to Pedestrian Signs City of Madison Traffic Engineering Division 1999 Two year evaluation of in- street yield-to- pedestrian signs installed at various locations in Madison, WI. No Treatment 1. Unsignalized pedestrian crosswalk signs and pavement markings, including advanced yield or stop markings and signs. (continued on next page)

144 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Treatment 1. (Continued). Study Title Authors Year Summary CMFs? Application Notes A Comparison of Gateway In- Street Sign Treatment to Other Driver Prompts to Increase Yielding to Pedestrians at Crosswalks Bennett et al. 2014 Studied sites with gateway configuration of in-street signs in Lansing and Detroit, MI. No San Francisco Pedsafe II Project Outcomes and Lessons Learned Hua et al. 2009 Evaluation of Pedsafe II project in San Francisco to study effectiveness of 13 countermeasures deployed at 29 sites throughout San Francisco, CA. No In-Street Yield to Pedestrian Sign Application in Cedar Rapids, Iowa Kannel et al. 2000 Summarized results of placing in-street yield-to- pedestrian signs at three sites in Cedar Rapids, IA. No Methods to Reduce Traffic Speed in High- Pedestrian Rural Areas Kamyab et al. 2003 Effects of installing removable pedestrian island and pedestrian crossing signs on a two-lane highway in rural Mahnomen County, MN. No Safety Evaluation of Yield-to- Pedestrian Channelizing Devices Strong and Kumar 2006 Report summarized safety evaluation of in-roadway yield-to- pedestrian signs installed at 21 midblock and intersection sites in four PA cities. No Evaluation of Countermeasures: A Study on the Effect of Impactable Yield Signs Installed at Four Intersections in San Francisco Banerjee and Ragland 2007 Video recordings to examine change in driver yielding rates as a result of impactable yield signs at three intersections in San Francisco. No In-Roadway "Yield to Pedestrian" Signs: Placement Distance and Motorist Yielding Ellis et al. 2007 Studied effect of placing an in- roadway "Yield to Pedestrians" sign at different distances in advance of a crosswalk. No

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 145 Treatment 1. (Continued). (continued on next page) Study Title Authors Year Summary CMFs? Application Notes Advanced Yield Markings and Fluorescent Yellow-Green RA 4 Signs at Crosswalks with Uncontrolled Approaches Van Houten et al. 2002 Studied effects of advance yield markings and fluorescent yellow-green RA 4 signs at 24 rural and urban crosswalks throughout Nova Scotia, Canada. No Decrease in conflicts San Francisco Pedsafe II Project Outcomes and Lessons Learned Hua et al. 2009 Evaluation of Pedsafe II project in San Francisco to study effectiveness of 13 countermeasures deployed at 29 No sites throughout San Francisco, CA. Other Signs TCRP Report 112/NCHRP Report 562: Improving Pedestrian Safety at Unsignalized Crossings Fitzpatrick et al. 2006 Some factors that influence driver yielding at sign locations may include speed and volume of roadway and whether motorists perceived yielding as a courtesy or the law. No Field Evaluation of Fluorescent Strong Yellow- Green Pedestrian Warning Signs Clark et al. 1996 Evaluated performance of pedestrian warning signs that used the new design at sites in central NC. No Decrease in conflicts Use of Signs and Symbols to Increase the Efficacy of Pedestrian- Activated Flashing Beacons at Crosswalks Van Houten et al. 1998 Evaluated the effects of two types of experimental signs on motorist yielding behavior. No Decrease in conflicts An Evaluation of High-Visibility Crosswalk Treatments - Clearwater, Florida Nitzburg and Knoblauch 1999 Studied effectiveness of internally illuminated overhead crosswalk signs that were installed in conjunction with high- visibility crosswalks at two midblock crossings in Clearwater, FL. No The Effects of Innovative Pedestrian Signs at Unsignalized Locations: A Tale of Three Treatments Huang et al. 2000 Evaluated two types of innovative pedestrian signs that were tested in Seattle and Tucson. No

146 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Treatment 1. (Continued). Study Title Authors Year Summary CMFs? Application Notes Advanced Stop Lines The Effects of Advance Stop Lines and Sign Prompts on Pedestrian Safety in Crosswalk on a Multilane Highway Van Houten 1988 Used a combination of stop markings and “Stop Here for Pedestrian” signs at three Dartmouth, Nova Scotia, crosswalks to analyze the effect of the treatments on vehicle-pedestrian conflicts and yielding behavior. No Increasing the Percentage of Drivers Yielding to Pedestrians in Three Canadian Cities with a Multifaceted Safety Program Malenfant and Van Houten 1989 Studied advance stop lines used with signs as a means of increasing motorist yielding at 34 crosswalks in three Canadian No Evaluation of SmartStud In- Pavement Crosswalk Lighting System and BlinkerSign Interim Report Kipp and Fitch 2011 Described experience installing, evaluating, and maintaining SmartStud in- pavement crosswalk lighting system and BlinkerSign. No Safety Countermeasure and Crash Reduction in New York City - Experience and Lessons Learned Chen et al. 2012 Evaluated the pedestrian safety impact of speed- limit reductions on roadway segments and at intersections. Yes – Not currently in CMF Clearinghouse (to be reviewed) At intersections, pedestrian crashes decreased by 41.49% at treatment sites and 15.58% at comparison sites. Marked Crosswalks Safety Effects of Marked versus Unmarked Crosswalks at Uncontrolled Locations: Executive Summary and Recommended Guidelines Zegeer et al. 2002 Found no statistically significant difference in pedestrian crash risk for various types of crosswalk markings. No A Review of Pedestrian Safety Research in the United States and Abroad Campbell et al. 2004 Summarized studies from many (Herms 1972, Gurnett 1974, Tobey et al. 1983, Gibby et al. 1994). No Pedestrian Accidents in Marked and Unmarked Crosswalks Jones and Tomcheck 2000 Evaluated pedestrian crashes at crosswalks at unsignalized arterial intersections in Los Angeles to test the validity of the city's crosswalk policies. Yes – Not in CMF Clearinghouse 73% reduction in pedestrian crashes after crosswalk markings were removed at the 104 sites combined with only one leg with previously marked crosswalks; considering both legs there was a 61% reduction in pedestrian crashes. cities in Newfoundland and New Brunswick.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 147 Treatment 1. (Continued). (continued on next page) Study Title Authors Year Summary CMFs? Application Notes Advanced Yield Markings and Pedestrian Safety: Analyses of Use with Danish Offsets and Median Refuge Islands Nambisan et al. 2007 Examined driver and pedestrian behavior at unsignalized intersections with respect to combinations of Danish offset, advance yield markings, and high-visibility crosswalk markings. No Pedestrian Safety Engineering and ITS-Based Countermeasures Program for Reducing Pedestrian Fatalities, Injury Conflicts, and Other Surrogate Measures Pecheux et al. 2009 Results of an evaluation of advance stop lines installed at one signalized and one unsignalized intersection in San Francisco. No The Influence of Signs Prompting Motorists to Yield Before Marked Crosswalks on Motor Vehicle- Pedestrian Conflicts at Crosswalks with Pedestrian- Activated Flashing Lights Van Houten and Malenfant 1992 Continued evaluation of advance stop markings as used with a pedestrian warning sign. No Reduction in conflicts Rumble Strips and Pedestrian Safety Cynecki et al. 1993 Studied effects of different types of advanced stop indicator. No Advance Yield Markings: Reducing Motor Vehicle- Pedestrian Conflicts at Multilane Crosswalks with Uncontrolled Approach Van Houten et al. 2001 Studied effectiveness of advance yield markings used with symbol signs at three crosswalks in Nova Scotia, Canada, where yellow flashing No Reduction in conflicts beacons were already in place. Advanced Yield Markings and Fluorescent Yellow-Green RA 4 Signs at Crosswalks with Uncontrolled Approaches Van Houten et al. 2002 Results of experiments that studied the effects of advance yield markings and fluorescent yellow-green RA 4 signs at 24 rural and urban crosswalks throughout Nova Scotia, Canada. No Reduction in conflicts

148 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Study Title Authors Year Summary CMFs? Application Notes Effect of Advanced Yield Markings and Symbolic Signs on Vehicle- Pedestrian Conflicts: A Field Evaluation Samuel et al. 2013 Two experiments conducted in MA to determine the effectiveness of driver yielding behavior and ability to scan for pedestrians at advanced yield markings. No Build It and They Will Yield: Effects of Median and Curb Extension Installations on Motorist Yield Compliance Hengel 2013 Single site study in CA of curb extension, pedestrian refuge island, and stop bar installation. No Treatment 1. (Continued). Treatment 2. High-visibility crosswalk marking patterns. Study Title Authors Year Summary CMFs? Application Notes An Evaluation of High-Visibility Crosswalk Treatments - Clearwater, Florida Nitzburg and Knoblauch 2001 Evaluated the effectiveness of high-visibility crosswalk markings used in conjunction with an illuminated overhead crosswalk sight at two sites in Clearwater, FL. No Evaluation of School Traffic Safety Program Traffic Control Measure Effectiveness Chicago DOT 2005 Results of an evaluation of experimental use of strong yellow/green crosswalk markings at over 100 Chicago elementary school zone crosswalks. No An Empirical Bayesian Evaluation of the Safety Effects of High-Visibility School (Yellow) Crosswalks in San Francisco Feldman et al. 2010 EB evaluation of safety outcomes of installing high-visibility crosswalks at 54 school sites in San Francisco, CA. Yes – Not in CMF Clearinghouse 37% reduction in collisions Detection Distances to Crosswalk Markings: Transverse Lines, Continental Markings, and Bar Pairs Fitzpatrick et al. 2011 Evaluated the relative visibility of three types of crosswalk markings, transverse lines, continental markings, and bar pair markings, under daytime and nighttime conditions. No

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 149 Treatment 2. (Continued). Study Title Authors Year Summary CMFs? Application Notes Safety Countermeasures and Crash Reduction in New York City - Experience and Lessons Learned Chen et al. 2012 Considered the effectiveness of high-visibility crosswalks in increasing pedestrian safety at intersections. Yes – Not currently in CMF Clearinghouse (to be reviewed) Authors estimated a 48% reduction in pedestrian crashes. Reduction was in crash rates. Evaluating the Effectiveness on Infrastructure- Based Countermeasures on Pedestrian Safety Pulugurtha et al. 2012 Evaluated four different infrastructure based countermeasures installed individually or in combination with other countermeasures at eight sites in Las Vegas, NV. No (continued on next page) Treatment 3. High-intensity activated crosswalk (HAWK) signals. Study Title Authors Year Summary CMFs? Application Notes Improving Pedestrian Safety at Unsignalized Crossings Fitzpatrick et al. 2006 Evaluated various midblock crossing treatments including PHB. No Safety Effectiveness of the HAWK Pedestrian Crossing Treatment Fitzpatrick and Park 2010 Evaluated safety effectiveness of PHB at 21 sites in Tucson, AZ. Yes – In CMF Clearinghouse (3&4 star ratings) 29% reduction (total crashes), 69% reduction (ped crashes), 15% reduction (severe crashes) Pedestrian Crossings at Mid-Block Locations: A Comparative Study of Existing Signal Operations Deng et al. 2013 Compared four midblock crossing signal types for safety and efficiency: pedestrian- activated (PA), pedestrian light controlled (PELICAN), high-intensity activated crosswalk (HAWK), and pedestrian user-friendly intelligent (PUFFIN). No

150 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Study Title Authors Year Summary CMFs? Application Notes Pedestrian and Motorists’ Actions at Pedestrian Hybrid Beacon Sites: Findings from a Pilot Study. International Journal of Injury Control and Safety Promotion Pulugurtha and Self et al. 2013 Researchers collected observational behavioral effects data from pedestrian crossings during weekday morning and evening peak times at five time points: before the installation, and at 1, three, 6, and 12 months following installation. No Increase in the percentage of yielding motorists, a decrease in the percentage of trapped pedestrians, and a decrease in pedestrian–vehicle conflicts at all three sites. Treatment 3. (Continued). Treatment 4. Rectangular rapid flashing beacons (RRFBs). Study Title Authors Year Summary CMFs? Application Notes Stutter-Flash Light Emitting-Diode Beacons to Increase Yielding to Pedestrians at Crosswalks Van Houten et al. 2008 Studied effectiveness of RRFBs in increasing motorist yielding behavior. No Reduction in conflicts, not crashes Pedestrian Safety Engineering and ITS- Based Countermeasures Program for Reducing Pedestrian Fatalities, Injury Conflicts, and Other Surrogate Measures: Final System Impact Report Pecheux et al. 2009 Used various measures of effectiveness to assess the effect of RRFBs on pedestrian and driver behavior. No Reduction in conflicts, not crashes San Francisco PedSafe II Project Outcomes and Lessons Learned Hua et al. 2009 Evaluated safety effectiveness of flashing beacons activated by pedestrians and ones that automatically detected pedestrians using infrared technology. No Reduction in conflicts, not crashes Evaluation of the Rectangular Rapid Flash Beacon at a Pinellas Trail Crossing in St. Petersburg, Florida Hunter et al. 2009 Summarized effects of installing a pedestrian- activated RRFB at the No location of one uncontrolled trail crossing.

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 151 Treatment 4. (Continued). Study Title Authors Year Summary CMFs? Application Notes Evaluating Driver and Pedestrian Behaviors at Enhanced Multi- lane Midblock Pedestrian Crossings: Foster et al. 2014 Reported on motorist yielding and pedestrians’ use of RRFBs No Site 1: Motorist yielding was 92% when beacons were activated and 75% when beacons were not activated. A Case Study in Portland, OR when they were activated by the pedestrians versus not activated. Site 2: Motorist yielding was 91% when the RRFBs were activated and 45% when not activated. Evaluation of Pedestrian and Bicycle Engineering Countermeasures: Rectangular Rapid- Flashing Beacons, HAWKs, Sharrows, Crosswalk Markings, and the Development of an Evaluation Methods Report. Fitzpatrick et al. 2011 Report on driver yielding with staged crossings rates at multilane sites. No Driver yielding increased across all sites. Effects of Yellow Rectangular Rapid- Flashing Beacons on Yielding at Multilane Uncontrolled Crosswalks Shurbutt and Van Houten 2010 Reported on effects of installing a yellow RRFB at 22 multilane uncontrolled crosswalks in St Petersburg, FL, Washington, DC, and Mundelein, IL. No Reduction in conflicts, not crashes Assessment of Driver Yielding Rates Pre- and Post-RRFB Installation, Bend, Oregon Ross et al. 2011 Evaluated RRFB installation at two Bend, OR crosswalks. No Improving Crosswalk Safety: Rectangular Rapid-Flashing Beacon (RRFB) Trial in Calgary. Domarod et al. 2013 Observed driver yielding with staged crossings at six locations in Calgary, Canada. No The RRFBs were observed to increase yielding by an average of 15 percent, to nearly 100 percent motorist yielding at the majority of the study sites. Overall the average motorist yielding increased from 83 percent to 98 percent following the installation of the RRFB. (continued on next page) Treatment 5. In-pavement warning lights. Study Title Authors Year Summary CMFs? Application Notes An Evaluation of Flashing Crosswalks in Gainesville and Lakeland Huang 2000 Effects of in- pavement lighting on driver compliance and motorist yielding to pedestrians. No Kirkland's Experience with In- Pavement Flashing Lights at Crosswalks Godfrey and Mazella 1999 Effects of in- pavement lighting on driver compliance and motorist yielding to pedestrians. No

152 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Treatment 5. (Continued). Study Title Authors Year Summary CMFs? Application Notes An Evaluation of Crosswalk Warning Systems: Effects on Pedestrian and Vehicle Behavior Hakkert et al. 2002 Studied the effectiveness of an in-pavement flashing light system that automatically detects the presence of pedestrians using infrared sensors (Israel). No Evaluation of In- Pavement, Flashing Warning Lights on Pedestrian Crosswalk Safety Van Derlofske et al. 2003 Results of a field evaluation of in- pavement flashing lights installed at two crosswalk sites at an uncontrolled intersection in Denville, NJ. No An Evaluation of the Effectiveness of an In-Pavement Flashing Light System Karkee et al. 2006 Summarized effectiveness of an in-pavement flashing light system installed at one uncontrolled crosswalk in the Las Vegas metropolitan area. No Evaluation of SmartStud™ In- Pavement Crosswalk Lighting System and BlinkerSign® Kipp and Fitch 2011 Observational study of in- pavement lights. No Found that while the system installed increased yielding and decreased approached speeds, it was removed due to multiple malfunctions and damage Treatment 6. Pedestrian refuge areas. Study Title Authors Year Summary CMFs? Application Notes Effects of Urban and Suburban Median Types on Both Vehicular and Pedestrian Safety Bowman and Vecellio 1994 Determine effects of urban and suburban median types on the safety of vehicles and pedestrians. Maybe Pedestrian Safety in Australia Cairney 1999 Crash rates decrease with wider medians. No Reduction in crash rates, not crashes The Road Safety Effectiveness of Wide Raised Medians Claessen and Jones 1994 Investigated effects of replacing painted medians with wide raised medians. Yes – Not in CMF Clearinghouse Reduced pedestrian crashes by 23% The Effects of Traffic Calming Measures on Pedestrian and Motorist Behavior Huang and Cynecki 2001 Evaluated a variety of traffic calming measures in several US cities. No

Effects of Pedestrian Treatments at Unsignalized Crossings: A Summary of Available Research 153 Treatment 6. (Continued). Study Title Authors Year Summary CMFs? Application Notes Evaluating the Effectiveness of Infrastructure- Based Countermeasures on Pedestrian Safety Pulugurtha et al. 2012 Evaluated four different infrastructure- based countermeasures including median refuge and Danish offset combined with high-visibility crosswalks at eight different sites in Las Vegas. No Build It and They Will Yield: Effects of Median and Curb Extension Installations on Motorist Yield Compliance Hengel 2013 Single site study in CA of curb extension, pedestrian refuge island, and stop bar installation. No Pedestrian Refuge Island Safety Audit Bacquie et al. 2001 Before-after study to evaluate the safety effectiveness of raised pedestrian refuge islands. Yes – Not in CMF Clearinghouse 73% reduction in midblock pedestrian crashes with addition of islands (however, there was a 136% increase in total crashes) Safety Effects of Marked Versus Unmarked Crosswalks at Uncontrolled Intersections Zegeer et al. 2002 Determine safety effects of marked vs. unmarked crosswalks on pedestrian crashes. No Methods to Reduce Traffic Speed in High- Pedestrian Rural Areas Kamyab et al. 2003 Effects of installing removable pedestrian island and pedestrian crossing signs on a two-lane highway in rural Mahnomen County, MN. No Pedestrian Safety Through a Raised Median and Redesigned Intersections King et al. 2003 Evaluated the effect of traffic calming measures involving signal, curb, sidewalk, and raised median installation and intersection redesign along a four-lane suburban road in NJ. No Pedestrian Safety Engineering and ITS-Based Countermeasures Program for Reducing Pedestrian Fatalities, Injury Conflicts, and Other Surrogate Measures: Final System Impact Report Pecheux et al. 2009 Results of an evaluation of median refuge islands installed at two signalized intersections in San Francisco, CA. No

154 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Treatment 7. Curb Extensions. Study Title Authors Year Summary CMFs? Application Notes Calming NYC Intersections King 1999 Evaluated the effect of curb extensions on crashes at six locations in NYC. No The Effects of Traffic Calming Measures on Pedestrian and Motorist Behavior Huang and Cynecki 2001 Evaluated curb extensions at eight residential and arterial crosswalk locations in MA, WA, NC, and VA. No Pedestrian Safety Impacts of Curb Extensions: A Case Study Johnson 2005 Case study analyzed effect of curb extensions at one uncontrolled intersection in Albany, OR. No Build It and They Will Yield: Effects of Median and Curb Extension Installations on Motorist Yield Compliance Hengel 2013 Single site study in CA of curb extension, pedestrian refuge island, and stop bar installation. No Treatment 8. Raised pedestrian crossings. Study Title Authors Year Summary CMFs? Application Notes The Effects of Traffic Calming Measures on Pedestrian and Motorist Behavior Huang and Cynecki 2001 Looked at how various pedestrian safety countermeasures, including raised pedestrian crossings, affected the behavior of pedestrians and drivers at three sites in NC and MD. Also evaluated (before/after study) a raised crosswalk in Cambridge, MA No

Abbreviations and acronyms used without definitions in TRB publications: A4A Airlines for America AAAE American Association of Airport Executives AASHO American Association of State Highway Officials AASHTO American Association of State Highway and Transportation Officials ACI–NA Airports Council International–North America ACRP Airport Cooperative Research Program ADA Americans with Disabilities Act APTA American Public Transportation Association ASCE American Society of Civil Engineers ASME American Society of Mechanical Engineers ASTM American Society for Testing and Materials ATA American Trucking Associations CTAA Community Transportation Association of America CTBSSP Commercial Truck and Bus Safety Synthesis Program DHS Department of Homeland Security DOE Department of Energy EPA Environmental Protection Agency FAA Federal Aviation Administration FAST Fixing America’s Surface Transportation Act (2015) FHWA Federal Highway Administration FMCSA Federal Motor Carrier Safety Administration FRA Federal Railroad Administration FTA Federal Transit Administration HMCRP Hazardous Materials Cooperative Research Program IEEE Institute of Electrical and Electronics Engineers ISTEA Intermodal Surface Transportation Efficiency Act of 1991 ITE Institute of Transportation Engineers MAP-21 Moving Ahead for Progress in the 21st Century Act (2012) NASA National Aeronautics and Space Administration NASAO National Association of State Aviation Officials NCFRP National Cooperative Freight Research Program NCHRP National Cooperative Highway Research Program NHTSA National Highway Traffic Safety Administration NTSB National Transportation Safety Board PHMSA Pipeline and Hazardous Materials Safety Administration RITA Research and Innovative Technology Administration SAE Society of Automotive Engineers SAFETEA-LU Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (2005) TCRP Transit Cooperative Research Program TDC Transit Development Corporation TEA-21 Transportation Equity Act for the 21st Century (1998) TRB Transportation Research Board TSA Transportation Security Administration U.S.DOT United States Department of Transportation

TRA N SPO RTATIO N RESEA RCH BO A RD 500 Fifth Street, N W W ashington, D C 20001 A D D RESS SERV ICE REQ U ESTED ISBN 978-0-309-44626-6 9 7 8 0 3 0 9 4 4 6 2 6 6 9 0 0 0 0 N O N -PR O FIT O R G . U .S. PO STA G E PA ID C O LU M B IA , M D PER M IT N O . 88 D evelopm ent of Crash M odification Factors for U ncontrolled Pedestrian Crossing Treatm ents N CH RP Research Report 841 TRB