Driver Behavior in Yielding to Sighted and Blind Pedestrians at Roundabouts

This study evaluated drivers' behavior in yielding the right-of-way to sighted and blind pedestrians who stood at different stopping distances from the crosswalk lines at entry and exit lanes at two different roundabouts. The findings demonstrate that drivers' willingness to yield to pedestrians is affected by whether they are attempting to cross at the entry or exit to the roundabout and, under some conditions, by the presence of a long cane. This research was supported by Grant Number EY12894 from the National Eye Institute, National Institutes of Health, to Richard Long of Western Michigan University. The authors thank Richard Long and David Guth for their editorial comments and William De l'Aune for his assistance with the statistical analyses. At roundabouts, where there are no traffic signals or stop signs, when the volume of traffic is low, sighted pedestrians simply recognize the absence of traffic before they cross. When the volume of traffic is high, they must identify a gap in vehicular traffic that is of sufficient duration to allow them to cross safely or must cross in front of vehicles whose drivers have yielded to them. Pedestrians who are blind or have low vision may find it challenging to acquire the necessary information to cross roundabouts safely. Blind pedestrians, who use their hearing to determine whether gaps in traffic are crossable, often have difficulty deciding when it is safe to cross at roundabouts. Guth, Ashmead, Long, Ponchillia, and Wall (2003) and Ashmead, Guth, Wall, Long, and Ponchillia (in press) found that blind pedestrians took significantly longer to detect crossable gaps than did sighted pedestrians. For roundabouts assessed in Baltimore (Guth, Ashmead, Long, Wall, & Ponchillia, in press) and in Tampa (Guth et al., 2003), blind pedestrians took, on average, 3 to 5 seconds more than did sighted pedestrians to indicate (by pushing a button) when it was safe to start crossing, and were likely to miss crossable gaps altogether. Furthermore, blind pedestrians were about 2.5 times less likely than were sighted pedestrians to make correct judgments. These findings suggest that blind pedestrians face greater challenges than do sighted pedestrians when crossing at roundabouts. However, if drivers yielded to blind pedestrians, and blind pedestrians could reliably determine that drivers had yielded, the need to detect crossable gaps would be eliminated, at least at single-lane roundabouts. In an attempt to increase the safety of pedestrians, states have introduced laws that grant pedestrians the right-of-way in marked crosswalks and at other locations, such as intersections without marked crosswalks. For example, in Maryland, where this study was conducted, Statute 21-602(a2) states: The driver of a vehicle shall come to a stop when a pedestrian crossing the roadway in a crosswalk is: (i) On the half of the roadway on which the vehicle is traveling; or (ii) Approaching so closely from the other half of the roadway as to be in danger. Furthermore, because of a specific incident in which a blind pedestrian was struck and killed, Maryland enhanced the right-of-way for blind or deaf pedestrians by eliminating the requirement to be in a crosswalk. Transportation Article 21-511 states: The driver of a vehicle shall yield the right-of-way to: (1) A blind or partially sighted pedestrian using a guide dog or carrying a cane predominantly white or metallic in color (with or without a red tip). This Maryland law applies only when pedestrians are physically in a crosswalk or, in the case of blind pedestrians, in a crosswalk or a street. When a pedestrian is standing at the curb, the law does not require a driver to yield. We believe that the use of "in a crosswalk," rather than "at a crosswalk," places a heavy burden on pedestrians who must step into the street to obligate drivers to yield. For the purposes of this study, yielding was defined as completely stopping a vehicle before the lines of a crosswalk to allow a pedestrian to cross in front of the vehicle. Pedestrians should cross in the crosswalk if they expect drivers to yield in compliance with the law. However, it appears to be too much of an assumption that drivers will yield even then. Koepsell et al.'s (2002) retrospective analysis of crosswalk markings and motor vehicle collisions with pedestrians aged 65 and older found that the risk of a collision between a pedestrian and a motor vehicle increased at crosswalks that did not have traffic signals or stop signs. Varhelyi (1998) passively recorded 790 driver-pedestrian encounters at a crosswalk in Sweden to determine, among other things, the frequency of drivers giving way to pedestrians and concluded that drivers did not observe the law, since they yielded to pedestrians only 5% of the time. The effect of a pedestrian with a disability on drivers' willingness to yield has also been studied. At a busy four-lane street in Edmonton, Alberta, Canada, with a striped crosswalk not regulated by a traffic signal, Harrell (1992) found that drivers were three times more likely to yield to an experimenter who was equipped with an orthopedic cane than to one who was not. Harrell (1994a) studied the factors that influenced drivers to stop for blind pedestrians at a non-signal-regulated pedestrian crossing, finding that drivers were significantly more likely to yield to blind than to sighted pedestrians. Ashmead et al.'s (in press) study at a multilane roundabout in Nashville, Tennessee, reported that overall, drivers yielded to pedestrians 27% of the time at the entry lane, but only 4% of the time at the exit lane. Furthermore, drivers seemed to be more willing to yield at entry lanes and to yield more to blind pedestrians than to sighted pedestrians. They yielded to blind pedestrians 36% of the time at the entry lane and 9% of the time at the exit lane, but yielded to sighted pedestrians only 17% of the time at the entry lane and not at all at the exit lane. file:///A|/G-Gerusc.txt (1 of 9)6/1/2005 8:05:43 AM file:///A|/G-Gerusc.txt To evaluate the yielding behavior of drivers for blind and sighted pedestrians at roundabouts, we presented drivers with a variety of pedestrian behaviors at two double-lane roundabouts. We hypothesized that the yielding behavior of drivers at roundabouts is affected primarily by the speed of the vehicle and whether the vehicle is entering or exiting the roundabout. We also wanted to assess whether a pedestrian's location relative to the curb altered the drivers' yielding behavior. In this regard, we hypothesized that the closer that pedestrians stood to the curb and crosswalk lines, the higher the percentage of drivers who would yield. Finally, we were interested in assessing the effect of the long cane that many blind persons use as a mobility aid on the willingness of drivers to yield. We hypothesized that the presence of a long cane would increase the yielding behavior of drivers. Method The roundabouts We studied yielding by drivers at two roundabouts that were approximately two miles apart in Annapolis, Maryland. Both roundabouts met the primary characteristics of modern roundabouts: They had no traffic-control signals other than yield signs for entering traffic; circulating vehicles had the right of way; access by pedestrians was allowed only on the legs of the roundabouts (that is, the roads that connect to the roundabout); no parking was allowed on the roundabouts; and all vehicles circulated counterclockwise (U.S. Department of Transportation, 2000). Roundabout 1 (see Figure 1), in the center of Annapolis at the intersections of Main Street and Compromise Street, is an urban doublelane roundabout. Roundabout 2 (see Figure 2), on the edge of the downtown Annapolis business district at the intersection of West Street and Taylor Avenue, is also an urban double-lane roundabout. Begin Figures 1 and 2: Figure 1. Bird's-eye view of Roundabout 1. The figure shows an aerial photograph of a double-lane roundabout with three legs. Its crosswalk identified with black arrows. Figure 2. Bird's-eye view of Roundabout 2. The figure shows an aerial photograph of a double-lane roundabout with four legs. Its crosswalk identified with black arrows. End of Figures 1 and 2. Roundabout 1 is in a high-density business district with signal-controlled intersections, one of which was located one block from the leg of the roundabout that was used in this study. Roundabout 2 is in a less-dense business environment than is Roundabout 1. A cemetery and a park are on each side (north and south) of the road from the crosswalk, and businesses are to the east and west of the roundabout. The nearest signal-controlled intersection is approximately one-third of a mile from the leg of the roundabout that was used in this study. We selected the entry lane and exit lane at one leg of each roundabout. The arrows in Figures 1 and 2 indicate the crosswalks where the study was conducted. For three of the four crossings, there were two lanes for vehicles. For the entry crossing at Roundabout 1, there was one lane for vehicles. Recognizing that measures of speed are static values of a dynamic situation and are dependent on the location of the vehicle at the time that the speed was recorded, we measured speeds from a variety of locations to characterize the driving environment of each roundabout more accurately. To describe each roundabout, we recorded the speeds of 20 vehicles at three locations, for a total of 60 vehicle speeds per roundabout. Although these 60 vehicles met the criteria for inclusion in the study (they were the lead vehicles and were passenger cars, vans, or half-ton pickup trucks only; no queuing vehicles [line of vehicles]; and if a second vehicle was following the lead vehicle, it was not tailgating), data on these drivers' yielding behavior were not collected; these speeds were recorded on a separate day from the one on which data on yielding were collected. The three locations at each roundabout where measurements of speed were taken were the leading edge of the entry and exit crosswalks, respectively, and 50 feet after the exiting crosswalk used in the study. Table 1 lists the means and standard deviations of the speeds that were measured at the three locations at each roundabout. Significant differences in speed between the two roundabouts were found across the different locations F (5, 19) = 33.2, p <.001. Post hoc analyses (Tukey's HSD) showed that the speeds that were recorded at the exit crosswalk and 50 feet from the exit crosswalk of Roundabout 1 were significantly lower than those that were recorded at the same locations at Roundabout 2. No significant difference was found between Roundabouts 1 and 2 for speeds that were recorded at the entry crosswalks. Begin Table 1: Table 1 Mean vehicular speeds (in mph) for the entry and exit crosswalks and 50 feet from the exit crosswalks at the two roundabouts during prestudy measurements. Description: There are 2 main column heads: Roundabout 1 and Roundabout 2. They each have 3 subheads: Entry crosswalk, Exit crosswalk, and 50 feet after exit. Mean speed (mph); Roundabout 1, Entry crosswalk: 15.7; Roundabout 1, Exit crosswalk: 12.2; Roundabout 1, 50 feet after exit: 15.0; Roundabout 2, Entry crosswalk: 13.9 ; Roundabout 2, Exit crosswalk: 16.8; Roundabout 2, 50 feet after exit: 22.7. Standard deviation; Roundabout 1, Entry crosswalk: 2.0; Roundabout 1, Exit crosswalk: 1.6; Roundabout 1, 50 feet after exit: 3.4; Roundabout 2, Entry crosswalk: 3.4; Roundabout 2, Exit crosswalk: 3.4; Roundabout 2, 50 feet after exit: 2.5.