scispace - formally typeset
Search or ask a question
Journal ArticleDOI

Pedestrian Crossing Behavior at Signalized Crosswalks

01 Aug 2017-Vol. 143, Iss: 8, pp 04017036
TL;DR: In this paper, pedestrian jaywalking at signalized crosswalks was investigated and observational surveys were conducted at seven cross-walks in different areas in Hong Kong, after which pedestrian info...
Abstract: This study investigated pedestrian jaywalking at signalized crosswalks. Observational surveys were conducted at seven crosswalks in different areas in Hong Kong, after which pedestrian info...

Summary (2 min read)

Jump to: [INTRODUCTION][DATA][METHODS][RESULTS][DISCUSSION] and [CONCLUSION]

INTRODUCTION

  • Signal control has been widely used around the world for more than 100 years.
  • They found that the type of gap (location and sequence of oncoming vehicles) and the stage of crossing (near end or far end) influenced pedestrians’ crossing decisions.
  • Pedestrians are only allowed to start crossing when the steady green signal is illuminated.

DATA

  • Seven signalized junctions were randomly selected from different areas of Hong Kong (Table 1).
  • Table 1 lists the numbers of observations obtained at each site with the corresponding signal cycle time and average flow.
  • It was found that 57% of the pedestrians who arrived at the crosswalks during the red signal jaywalked, and 100% of those who arrived during the flashing green signal jaywalked without waiting for another cycle.
  • The gender and age of the pedestrians were identified during the video tracking.

METHODS

  • The binary logit model was used to represent how the individual-specific and site-specific factors influence the pedestrians’ jaywalking behavior.
  • In using the basic binary logit model, each pedestrian was regarded as an individual observation, and parameters β were assumed to be constant for all individuals at all sites, i.e., a fixed-parameter model.
  • The same set of parameters β were applied to all observations at all sites.
  • Random variations in the effects of the explanatory variables among pedestrians and random effects across sites could have existed.
  • Therefore, observations in the same site were grouped as panel data, and a random effect binary logit model was used to account for both the within-site correlations and the intersite heterogeneity.

RESULTS

  • STATA 13 was used to estimate the three binary logit models.
  • Not surprisingly, the correlation analysis indicated that the two dummy variables of gender, M and F, were highly correlated.
  • Average vehicle flow was found to be highly correlated with cycle time and pedestrian red signal time, as the longer the time for vehicles, the larger the average vehicle flow.
  • Both the random effect and random parameter binary logit models have lower AIC values and larger values of McFadden’s adjusted pseudo R2 than the basic binary logit model.
  • Similarly, the statistic of the likelihood-ratio test between the random effect model and the random parameter model is 103.50, which again is much larger than χ2 (3, 99%) = 11.34.

DISCUSSION

  • In the random parameter model, 4 of the 8 variables (gender, walking speed, percentage of pedestrians jaywalking, and crossing time) produced statistically significant random parameters (all were normally distributed).
  • In addition to the abovementioned individual-specific factors, individual pedestrians were likely to be influenced by surrounding pedestrians who arrived during the same cycle.
  • The results in Table 5 indicate that both a larger number of pedestrians in the cycle and a higher percentage of those jaywalking in the cycle increased the probability that a particular pedestrian would jaywalk.
  • This result is the opposite of Rosenbloom’s (2009) finding that the tendency to cross on a red signal is lower when there are more people waiting at the curb, due to the power of social control.
  • The marginal effect (0.021 in the random parameter model) implies that a second increase in crossing time resulted in a 2% increase in the probability of jaywalking.

CONCLUSION

  • This study investigated the contributory factors of pedestrians’ jaywalking behavior at signalized crosswalks.
  • The pedestrian characteristics (gender, walking speed), the behavior of surrounding pedestrians (total number of pedestrians and the proportion of jaywalkers), the vehicle traffic, the timing of arrival and the length of signal, and the location of the crosswalk were found to significantly determine the probability of pedestrian jaywalking.
  • Because Hong Kong is a densely populated city, it would be well worth considering providing more informative signals rather than simply promoting the regulation.
  • A possible measure that policy makers could consider is to introduce a signal countdown with the conventional graphic signal, which has been shown to significantly increase the proportion of pedestrians who start to cross during the green signal (Keegan and O’Mahony, 2003) and to effectively enhance pedestrian safety (Schattler et al., 2002).

Did you find this useful? Give us your feedback

Content maybe subject to copyright    Report

TE3973
PEDESTRIAN CROSSING BEHAVIOR AT SIGNALIZED CROSSWALKS
S.Q. XIE (Corresponding Author)
Department of Civil Engineering, The University of Hong Kong, Pokfulam Road, Hong
Kong, Tel.: 852-2859-2662; fax: 852-2517-0124; e-mail: seakay@connect.hku.hk
S.C. WONG
Department of Civil Engineering, The University of Hong Kong, Pokfulam Road, Hong
Kong, Tel.: 852-2859-1964; fax: 852-2559-5337; e-mail: hhecwsc@hku.hk
Tsz Man NG
Department of Civil Engineering, The University of Hong Kong, Pokfulam Road, Hong
Kong, e-mail: utszman@connect.hku.hk
William H. K. LAM
Department of Civil and Environmental Engineering, Hong Kong Polytechnic University,
Hung Hom, Kowloon, Hong Kong, Tel.: 852- 2766-6045; fax: 852- 2365-9291; e-mail:
william.lam@polyu.edu.hk
ABSTRACT
This study investigated pedestrian jaywalking at signalized crosswalks. Observational surveys
were conducted at 7 crosswalks in different areas in Hong Kong, after which pedestrian
information and site condition data were incorporated into a database. A binary logit model
was used to identify possible factors that determine the probability of pedestrian jaywalking.

To address the variation in the effects of the explanatory variables among pedestrians and the
unobserved heterogeneity across sites, we used a random parameter model and a random effect
model, respectively. The results showed that the random parameter model performed the best
in terms of goodness-of-fit. It was found that the signal when a pedestrian arrives at the
crosswalk is critical for decision making, and the jaywalking of surrounding pedestrians also
influences the pedestrian’s decision to cross. The gender and walking speed of the pedestrian,
vehicle flow, and site location and condition of the crosswalk were also found to significantly
determine the probability of pedestrian jaywalking.
Keywords: Pedestrian crossing behavior; Jaywalking; Signalized crosswalks; Random
parameter model; Random effect model
INTRODUCTION
Signal control has been widely used around the world for more than 100 years. It provides a
safe, economic, and efficient means of coordinating conflicting traffic flows at junctions, and
is particularly popular in densely populated cities with heavy vehicle and pedestrian traffic
loads. Signalized junctions are the most common type of junction in Hong Kong. Although
pedestrian-vehicle collisions at signalized junctions have been reduced by 35% in the past 5
years, 387 pedestrian-vehicle crashes were still recorded, which comprised nearly 25% of the
accidents that occurred at signalized junctions. Drivers should be aware of the traffic
regulations as they must pass a written test on the Road Users’ Code before obtaining their
licenses, so pedestrian jaywalking is the most likely cause of pedestrian-vehicle accidents at
signalized junctions in Hong Kong.

Studies have attempted to identify the factors that influence pedestrian crossing behavior. In
terms of individual characteristics, it was observed that male pedestrians tend to jaywalk more
than female pedestrians (Tiwari et al., 2007, Rosenbloom, 2009, Brosseau et al., 2013). A
similar tendency was observed in questionnaire surveys based on the theory of planned
behavior, which investigated pedestrians’ attitudes toward jaywalking (Diaz, 2002, Zhou et al.,
2009). However, Ren et al. (2011) observed that middle-aged female pedestrians had the lowest
compliance rate in China. The elderly were also found to be more patient and less likely to
jaywalk (Guo et al., 2011, Zhuang & Wu, 2011, Ren et al., 2011, Brosseau et al., 2013). Oxley
(1997) conducted an experiment on pedestrian traffic judgment and observed that older adult
pedestrians generally adopted a less safe crossing strategy and performed worse than younger
pedestrians on two-way undivided roads, although their performance was similar to that of
younger pedestrians on one-way divided roads. The differences associated with age-related
physical, perceptual, and cognitive deficits were further discussed and validated in an
experimental study of the age differences in pedestrians’ gap selection (Oxley, 2005). Holland
and Hill (2009) pointed out that driving experience also affected pedestrians’ decisions to make
unsafe crossings. Surprisingly, pedestrians with driving experience left smaller safety margins,
although they were more likely to look both ways before crossing than non-drivers. Ren et al.
(2011) suggested that a possible reason for the low compliance rate of female pedestrians in
China was that fewer of them had driving licenses. They showed that individual characteristics
affect pedestrians’ judgement of the traffic conditions and gap selection. In view of this, Koh
and Wong (2014) used a binary logit model to predict the proportion of pedestrians who accept
a gap, and hence jaywalk. They found that the type of gap (location and sequence of oncoming
vehicles) and the stage of crossing (near end or far end) influenced pedestrians’ crossing
decisions.

In addition to the pedestrian characteristics and types of gap, the environment and site
conditions may also affect the decision making of pedestrians. Lavalette et al. (2009) found
that the number of lanes of traffic, the presence of pedestrian crossing signals, and the presence
of a central traffic island influenced pedestrians’ decision making at crossings. Kruszyna and
Rychlewski (2013) investigated the influence of approaching trams on pedestrian behavior at
signalized crosswalks in Portland. Li and Fernie (2010) suggested that the weather also
influenced the compliance rate, particularly for pedestrians crossing a signalized two-stage
crossing with a center refuge island in the winter. The waiting time was also found to increase
the probability of pedestrians jaywalking (Tiwari, 2007, Li and Ferinie, 2010), and Li (2013)
proposed a model for pedestrians’ intended waiting time. To reduce the high incidence of
jaywalking and, hence, improve pedestrian safety at signalized crosswalks, pedestrian
countdown signals have been introduced in recent years to prevent pedestrians from
overestimating the waiting time (Keegan and Mahony, 2003) and taking the risk to jaywalk.
This measure has been proven to effectively reduce the number of pedestrians starting to cross
before the signal eventually turns green (Schattler et al., 2002).
Among the approaches used to identify the factors associated with pedestrian jaywalking
behavior, ANOVA has been used to analyze the differences among groups of pedestrians (Li
and Ferinie, 2010, Ren, et al., 2011) and logistic regression has been used to represent the
effects of explanatory variables in determining the probability of jaywalking (Rosenbloom,
2009, Brosseau et al., 2013). ANOVA is useful for evaluating the influence of demographic
factors, whereas logistic regression models are capable of linking the effects of the factors with
the probability of jaywalking. However, the effects of explanatory variables are considered to
be constant and fixed among all pedestrians in the simple logistic regression models, which
may lead to misleading outcomes if considerable variation exists in the effects among

individual pedestrians. In addition, although numerous previous studies have observed
pedestrian crossing behavior at different sites, few studies have discussed the possible
unobserved site differences.
In this study, observational surveys were conducted in 7 crosswalks in Hong Kong. The
relevant individual-specific factors and site-specific factors were extracted and incorporated
into a binary logit model to identify the contributory factors that determine the probability of
jaywalking. To address the heterogeneity across pedestrians and sites, a random parameter
model was used to accommodate the variation in the effects of the explanatory variables, and
a random effect models was used to account for the unobserved heterogeneity across sites.
In Hong Kong, the sequence of pedestrian signals is a steady green signal, a flashing green
signal, and a steady red signal. Pedestrians are only allowed to start crossing when the steady
green signal is illuminated. The flashing green signal indicates that the pedestrians already on
the crosswalk should continue and finish crossing at a reasonable speed. However, pedestrians
who have not started crossing should wait until the next steady green signal. No pedestrians
are allowed to cross during the red signal. In this study, pedestrians who entered a crosswalk
during the flashing green signal or the red signal were regarded as jaywalkers according to the
traffic regulations in Hong Kong. No countdown signals are provided at pedestrian crosswalks.
DATA
In this study, seven signalized junctions were randomly selected from different areas of Hong
Kong (Table 1). There were 4 sites in urban areas, including 2 in Hong Kong Island and 2 in
Kowloon, and the other 3 were in the New Territories. Video recording was conducted at each
site for about 90 minutes, during which pedestrian movements were captured for further
analysis. Preliminary analysis had previously been conducted based on the Travel

Citations
More filters
Journal ArticleDOI
TL;DR: Effects of both personal factors and environmental factors on the individual decision of red light running violation are examined, using the video observation surveys at the signalized crossings that are prone to pedestrian-vehicle crashes and have moderate pedestrian and vehicular traffic volumes in the urban area.
Abstract: Pedestrian is vulnerable to mortality and severe injury in road crashes. Red light running violation of pedestrians is one of the leading causes to the crashes at signalized intersections, at which the crash involvement rates of pedestrians are high. Therefore, it is important to identify the factors that affect the propensity of red light running of pedestrian. In this study, effects of both personal factors (pedestrians’ demographics and behavior) and environmental factors (presence and behavior of other pedestrians, signal time, and traffic condition) on the individual decision of red light running violation are examined, using the video observation surveys at the signalized crossings that are prone to pedestrian-vehicle crashes and have moderate pedestrian and vehicular traffic volumes in the urban area. Crossing behaviors of 6320 pedestrians are captured. Results of a random parameter logit model indicate that pedestrian gender, age, number of lanes, presence of a companion, number of pedestrians around, presence of other violators in the same cycle, time to green, red time, traffic volume, and percentage of heavy vehicles all affect the propensity of red light running violation of pedestrians. Also, there are significant interaction effects by pedestrian’s gender and age, presence of other violators, with a companion, and traffic volume on the propensity. Findings are indicative to the development of effective engineering, enforcement and educational initiatives combating the red light running violation behavior of pedestrians. Therefore, pedestrian safety level at the signalized intersections can be enhanced.

38 citations

Journal ArticleDOI
TL;DR: In this article, the influence of gender, age group, group-crossing, technological devices and carrying items on pedestrians crossing behaviors at signalized and unsignalized crosswalks simultaneously was examined.

32 citations

01 Jan 1996
TL;DR: The authors investigated the road crossing practices of older adults relative to younger adults in order to determine whether they experience particular problems making traffic judgements and whether their road crossing behavior might render them more vulnerable to crashes because of declines in physical, sensory, perceptual, and cognitive abilities.
Abstract: There is considerable evidence that older pedestrians are much more likely to be severely injured or killed once involved in a crash than any other group of road users. This study set out to investigate the road crossing practices of older adults relative to younger adults in order to determine whether they experience particular problems making traffic judgements and whether their road crossing behavior might render them more vulnerable to crashes because of declines in physical, sensory, perceptual, and cognitive abilities. An initial "blackspot" accident analysis highlighted the types of crashes in which older (and younger) pedestrians were involved. From this a number of specific road crossing actions were identified that were investigated in more detail in two observational studies. Road crossing behaviors were measured from video recordings of individual road crossings of younger and older pedestrians. On two-way roads, older adults were observed to interact more with the traffic and left less time between completing the cross to the middle of the road and time-of-arrival of the closest near-side vehicle than younger pedestrians. The findings suggest age-related perceptual and cognitive deficits may contribute to many of the crashes in which older pedestrians are involved.

25 citations

Journal ArticleDOI
TL;DR: Wang et al. as discussed by the authors investigated the effect of personal characteristics, traffic attributes and environmental factors that affect the red light running propensities of pedestrians at two-stage crossings, based on the video observation surveys at six urban signal intersections in Hong Kong.

18 citations

Journal ArticleDOI
TL;DR: The proposed method is applied to a case study of Erzurum, Turkey that involves determining the best pedestrian crossing scenario and shows that the most suitable scenario is S.2.1, which provides an improvement of up to 50 % over the current situation.

15 citations

References
More filters
Journal ArticleDOI
TL;DR: The over-representation of older pedestrians in serious injury and fatal crashes compared to younger adults may be due, in part, to age-related diminished ability to select gaps in oncoming traffic for safe road-crossing.

350 citations

Journal ArticleDOI
TL;DR: Based on the theory of planned behavior, pedestrians' attitudes towards traffic violations and self-ratings of violations, errors and lapses were measured amongst a sample of 146 pedestrians as discussed by the authors.
Abstract: Based on the theory of planned behavior, pedestrians' attitudes towards traffic violations and self-ratings of violations, errors and lapses were measured amongst a sample of 146 pedestrians The evaluation of the planned behavior model using EQS showed a significant fit ( χ 2 (11)=1604 p =014; normal fit index =0937 ) The reported violations, errors and lapses appears causally related to the intention to violate regulations, and this in turn with positive attitudes, subjective norm and perceived behavioral control Young people have a more positive attitude towards committing violations as pedestrians than adults and perceive the subjective norm to be less inhibitory, have less control over violations, have a more positive intention to commit violations, and report more violations, errors, and lapses than adults Men report more frequent violations of the traffic rules than women do The conclusion is drawn that pedestrians, and in particular young males, are at least in part responsible for the high accident rates and remedial actions are suggested

262 citations

Journal ArticleDOI
TL;DR: The findings suggest that age-related perceptual and cognitive deficits may play a substantial role in many of the crashes involving older pedestrians.

252 citations

Journal ArticleDOI
TL;DR: In this article, the authors examined the road behavior of individual pedestrians at an intersection with a traffic signal compared to groups of pedestrians at the same intersection and found that males crossed on red more frequently than females.
Abstract: The present study examines the road behaviour of individual pedestrians at an intersection with a traffic signal compared to groups of pedestrians at the same intersection. In total, 1392 pedestrians were unobtrusively observed in an urban setting at a pedestrian street crossing of undivided streets; 842 were female (60.5%) and 550 were male (39.5%). The observations took place between 7:30 and 8:30 in the morning. Chi-square test revealed the males crossed on red more frequently than females. Logistic regression predicting red-light crossing for pedestrians arriving during a red-light phase indicated that, apart from gender, the tendency to cross on red was greater when there were fewer people waiting at the curb, either when a pedestrian arrived, or joining after arrival. The discussion refers to the theoretical explanations concerning the theory of ‘social control’ and to some practical implications of the results, such as using the positive value of social control in media campaigns and adjusting the red light duration in order to encourage people to obey the traffic light.

240 citations

Frequently Asked Questions (2)
Q1. What contributions have the authors mentioned in the paper "Te3973 pedestrian crossing behavior at signalized crosswalks" ?

This study investigated pedestrian jaywalking at signalized crosswalks. 

In the future, observational surveys conducted at more sites with different geometric features and signal phasing schemes would enable further insights to be obtained on the effects of site-specific factors.