U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

 
REPORT
This report is an archived publication and may contain dated technical, contact, and link information
Back to Publication List        
Publication Number:  FHWA-HRT-13-098    Date:  January 2014
Publication Number: FHWA-HRT-13-098
Date: January 2014

 

Human Factors Assessment of Pedestrian Roadway Crossing Behavior

Pedestrian factors

It is obvious that pedestrians, in most cases, attempt to select the fastest and most direct route from their origin to their preferred destination. However, route planning must take into account the features of the traveling environment, physical abilities, some level of risk (real or perceived), among other factors. For example, suppose that Route A is .5 miles in length and that Route B is 1 mi in length. While Route A is shorter in absolute distance, it includes two steep hill segments, and Route B is flat. Given these environmental characteristics, a young and physically fit pedestrian would likely select Route A. However, an older adult with a physical impairment might select Route B for its terrain advantages.

It is improbable that people use complex decisionmaking processes to determine which route to walk to reach a specific location. Rather, pedestrians are likely to plan a general route and rely on other factors to make smaller in-route choices, such as where to cross the roadway. For example, one person may “always” cross at a specific location, and for this reason, he or she will rely on habit and cross there again on subsequent trips. Another pedestrian may have mobility issues and will cross at the most accessible location. It is probable that pedestrians rely on a set of heuristics or tendencies when selecting when and where to cross the roadway. Here, the study team explores some of the intrinsic pedestrian characteristics that can influence route choice and crossing behaviors.

Gender

Overall, males are more likely to be involved in vehicle crashes than females.(6) This trend remains consistent with pedestrian fatalities. In 2009, the male pedestrian fatality rate was 1.86 per 100,000, compared with .82 per 100,000 females. In total, 69 percent of the pedestrian fatalities were male. This gender difference is consistent across all age groups.(7)

This evidence suggests that gender plays a role in pedestrian crossing decisions to some degree. This idea is further supported in a recent study by Holland and Hill.(8) The authors presented participants with a variety of potentially dangerous crossing scenarios. In each of the scenarios, women reported perceiving more risk than their male counterparts. Furthermore, the females were less likely to indicate an intention to cross the roadway. Taken together, these data suggest that, in general, males and females view and act on potential safety hazards in the roadway environment differently. These gender differences might prove useful in targeting interventions and education toward different groups.

Age

Age also plays a role in pedestrian–vehicle collisions. In 2009, pedestrians 65 and older had the highest rate of fatalities, 1.96 per 100,000. Moreover, older adults constituted 19 percent of all pedestrian fatalities.(7) On the opposite end of the spectrum, pedestrians 15 years and younger accounted for 25 percent of all pedestrian injuries and 7 percent of pedestrian fatalities in 2009.(7)

Many factors could result in these age-related fatality trends. One possibility is that younger and older pedestrians make riskier road-crossing choices. However, multiple studies have shown that when participants are presented with potentially dangerous crossing scenarios, older adults are less likely to indicate that they would be likely to attempt the crossing.(8,9,10) Thus, it appears, at least in regard to intention to cross the road, older adults are not necessarily making more risky choices. In fact, this suggests just the opposite. One can logically draw the conclusion that those who are less likely to engage in risky crossing behaviors are also less likely to be involved in pedestrian–vehicle collisions.

If older adults are less inclined than younger adults to intend to make risky crossings, then other factors must contribute to their large proportion of the total pedestrian fatalities. This appears to be the case. While older adults constitute 19 percent of all pedestrian fatalities, they only make up 8.5 percent of the total injuries.(7) These data imply that older adults are less likely to be struck by a vehicle. However, when an older adult is involved in this type of collision, it is more likely to result in a fatality.

It has been suggested, for example in Barton and Morrongiello, that children have not fully developed the cognitive reasoning skills to cross the street safely.(11) It is certainly a possibility that children make riskier crossing decisions. It is also likely that smaller children are more difficult to see from the drivers’ perspective. This condition makes it more difficult for a driver to proactively avoid collision with this special group of pedestrians.

Age differences in pedestrian injury and fatality rates may help to direct educational interventions to increase safe crossing behaviors. Further, age-related differences may help engineers to target different interventions for different age groups. For example, sidewalk markings outside elementary schools may lead children to walk to pedestrian-activated crosswalks. In addition, crosswalk activation buttons could be lowered to child level or made visually more attractive to push. Longer protected crossings could also be employed near senior citizen communities.

Alcohol

Alcohol plays a significant role in pedestrian fatalities. In 2009, it was reported that 48 percent of crashes resulting in a pedestrian fatality involved alcohol. When focusing on drivers involved in these collisions, about 13 percent had a blood alcohol concentration (BAC) of .08 percent by volume or greater. This is compared with 35 percent of pedestrians. Further, both the driver and the pedestrian had BAC levels of .08 percent by volume or higher in 6 percent of the fatal crashes.(7)

Self-Identity

Self-identity as a safe or careful pedestrian has been shown to be correlated with roadway crossing behavior. For example, Holland, Hill, and Cooke asked pedestrians to make decisions on whether or not they would cross the roadway in various situations.(10) It was found that pedestrians who had high self-identities as careful pedestrians selected fewer gaps in which they would cross the roadway at non-signalized crossing locations. Similarly, Evans and Norman presented pedestrians with a set of potentially hazardous road crossing behaviors. Those people who identified themselves as “safe pedestrians” were less likely to state that they would cross the roadway in such a manner.(9) A similar study with adolescents also found that those who rated themselves as safe pedestrians reported that they were less inclined to cross at a potentially risky location.(12)

If people who identify themselves as cautious accept fewer gaps to cross the roadway, then it is ideal to further understand what factors influence this self-identity. Holland, Hill, and Cooke found that people who had previously been involved in a roadway collision viewed themselves as more willing to take a risk.(10) Contrarily, those with mobility impairments and older adults rated themselves as more cautious. Furthermore, it was found that the biggest predictor of a self-identity as a careful or safe pedestrian was age (even when accounting for factors such as experience and mobility). This suggests that older adults take fewer risks when crossing the roadway than do younger pedestrians. Furthermore, this helps to explain the age-related discrepancy in the number of collisions with younger persons than older persons. Beyond this, if it is known that older adults are pedestrians who take fewer risks, then efforts to improve roadway safety might be better directed at younger foot travelers.

The results of these studies suggest that people who consider themselves to be careful/safe pedestrians are more aware of the potential hazards involved in crossing the roadway. As a result, these people may be less likely to employ potentially risky road-crossing choices. Along similar lines, it may be reasonable to assume that increases in perceived roadway dangers (either via environmental changes or educational interventions) will lead to fewer crossings at unmarked roadway locations.

Perceived Control

In addition to the aforementioned factors, Evans and Norman found that perceived control influenced pedestrians’ intentions to cross the roadway.(12) (Perceived behavioral control is a component of the Theory of Planned Behavior, a widely used theory of safety- and health-related behaviors.(13)) Participants were presented with three scenarios: crossing during a gap in traffic of a multilane roadway to get to a vehicle after a shopping trip, crossing at signalized pedestrian crossing during the don’t walk phase to get to the dry cleaners before closing, and crossing a residential roadway to meet a friend. For each of the scenarios, it was probable that those participants who perceived having more behavioral control (as assessed by a planned behavior questionnaire) of the situation would indicate that they would cross the roadway.

Evans and Norman also attempted to apply the Theory of Planned Behavior to adolescent intentions to cross the roadway.(12) Middle school aged children were presented with a potentially dangerous road-crossing scenario. Much like the researchers’ previously mentioned findings, perceived behavioral control was the strongest predictor of expressed intention to cross the roadway.(9) In both studies, perceived control of the situation is the greatest predictor of pedestrians’ planned intent to cross the roadway. This suggests that if pedestrians’ perceptions of the ability to cross the roadway at unsafe areas are decreased, then overall pedestrian safety may increase.

Environmental factors

Many factors contribute to whether people attempt to cross the roadway at a specific location at a specific time. However, features of the roadway environment have been largely ignored when examining these causal factors. Here the study team discusses foot travel in three components: trip originators, destinations or attractions, and affordances.

Trip Originators

Trip originators are areas where pedestrians begin trips. Some originators tend to generate more trips than others. For example, a house generates a finite number of trips. However, other originators generate countless pedestrian trips. Some of these high trip generating sources are places such as shopping malls, Metro/subway stations, and bus stops. These high trip originators require special attention with regard to pedestrian flow. One can easily imagine how pedestrian travel patterns might differ between a bus stop placed in the middle of a block and a bus stop placed at the corner of an intersection.

Trip Destinations

Trip destinations are end points of pedestrian trips, whether they are final destinations or attractions (e.g., coffee shop) en route. Often, destinations are also trip originators. Take the example of a shopping center. Many people may consider the shopping center the completion of their trip, that is, their destination. However, when people leave this locale, the shopping center becomes the trip originator. As a result, designers must attend to both how pedestrians enter and exit such locations.

Affordances

It is obvious from the pedestrian fatality data that simple risk perception is not an adequate source of determining whether or not one should cross the roadway at a given location at a given time. If risk perceptions were adequate, pedestrians would not take such potentially harmful actions. Instead, it is likely that people rely on action-oriented perceptions of affordances.

An affordance refers to the qualities (real or perceived) of the environment (or object).(14) These qualities/properties determine how the environment/object can be used. For example, a chair of sufficient size and stability affords sitting or climbing upon. Along similar lines, a gap in traffic of sufficient distance may afford crossing of the roadway. Affordances are egocentric. This means that a chair that affords sitting to a child may not afford sitting for an adult. Similarly, a gap in traffic may afford crossing for an able-bodied young adult but not for an older adult with mobility impairment. It is very important to note that although these qualities of the environment are directly perceived, they may or may not be real. As a result, the affordances of an environment can lead a person to take an incorrect or unsafe action. For example, imagine an elongated door handle with the word “push” above it. A user will likely directly perceive that the door handle appears to afford pulling and will attempt to pull the door open. This is an “incorrect” action in the sense that the door must be pushed to open.

When they act upon perceived affordances, people generally produce a behavior that is adequate, this is, unless the affordance leads the user to perform an unsafe action. According to this perspective, behavior is goal-driven (e.g., I want to cross the street) rather than avoidance-driven (e.g., I don’t want to be struck by that car). This results in generally adequate choices (e.g., I can cross the roadway now). Of course, irregular outside influences have the potential to increase the salience of avoidance-driven choices (e.g., seeing a police officer during a known target jaywalking enforcement zone).

Perceived affordances are, of course, not the same as actual safety. People often incorrectly perceive an action to be safe, when indeed, it is not. For example, as Tyrrell, Wood, and Carberry found, at night, pedestrians frequently overestimate the distance at which drivers are able to see them.(15) These perceptions can lead pedestrians to take unsafe actions. As such, it is important to consider affordances when designing an environment. This includes affordances that influence safety, aesthetics, and functionality. Take, for example, Norman’s classic teapot example. Norman describes a teapot that has the handle and the spout on the same side of the pot.(16) While the design is aesthetically pleasing, the pot is not functional, and could actually pose quite a burn hazard. This example can easily be applied to the roadway and pedestrian travel area. While it may be desirable to make pedestrian pathways pleasing to the eye, this should not be the top priority when designing roadway environments. Instead, pathways should be developed to increase the ability of pedestrians to safely (and quickly) move from an originator to a destination. In reaching this goal, aesthetically pleasing environmental furniture could be used. For example, attractive shrubbery might be placed along a curb to dissuade people from crossing midblock (by reducing the roadway crossing affordance). Ultimately, roadway environments should maximize the roadway-crossing safety while not inhibiting pedestrian or vehicle travel.

If one is able to design an area that maximizes the affordances related to safe pedestrian crossing behaviors (and minimizes those that increase more hazardous crossing), pedestrians should be more likely to make safe crossings. Furthermore, this should reduce the need for hazard, or warning signs. If a better understanding of pedestrian crossing affordances is gained, modifications can be made to increase pedestrian safety, including the implementation of marked crossings (in optimal locations that encourage use) or midblock crossing inhibitors (encouraging pedestrians to cross at marked intersection crossings), while not inhibiting pedestrian travel. Further, if pedestrian crossing schemas can be identified, possible educational interventions can be developed.

An iterative process is presented here that explored the factors surrounding pedestrian crossings at 20 different intersections in the Washington, DC, metropolitan area. Data were collected in three separate phases. Each phase is discussed separately. A combined summary of each of the phases is presented later. The influences of environmental factors on pedestrian crossing locations are discussed. The individual relationships of these factors and where (i.e., intersection or non-intersection) and how (e.g., walk, run, etc.) pedestrians cross the road are examined. In addition, models are provided to predict where pedestrians are likely to cross given a set of roadway environmental factors. The ability to readily predict problematic unmarked crossing locations before crashes occur could be useful in the proactive deployment of mitigation strategies. These may include educational interventions or engineering countermeasures, which have been shown to be effective (e.g., Pelican, Puffin, and Zebra crossings, or refuge islands).(17)

 

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101