Transportation Research Part F 41 (2016) 170–178

Contents lists available at ScienceDirect

Transportation Research Part F

journal homepage: www.elsevier.com/locate/trf

Pedestrian behaviours: Validation of the Serbian

version of the pedestrian behaviour scale
Boris Antić a, Dalibor Pešić a, Nenad Milutinović b, Marko Maslać b,⇑
a
University of Belgrade, The Faculty of Transport and Traffic Engineering, Vojvode Stepe 305, Belgrade, Serbia
b
The Higher Education Technical School of Professional Studies Kragujevac, Kosovska 8, Kragujevac, Serbia

a r t i c l e i n f o a b s t r a c t

Article history: Statistics show that each year a great number of pedestrians in Serbia are killed in traffic acci-
Received 10 February 2015 dents. The purpose of this paper is to determine the frequency of some risky behaviours of
Received in revised form 4 January 2016 pedestrians in traffic in Serbia. The current version of Pedestrian Behaviour Scale (PBS)
Accepted 13 February 2016
was conducted among the participants in Serbia, but it was also improved by the addition
Available online 8 June 2016
of questions concerning the use of mobile phones while crossing the street. The questions
were divided into 5 characteristic separate groups (violations, errors, lapses, aggressive
Keywords:
and positive behaviours), confirmed by factor analyses. The results shown that pedestrians
Pedestrians
Behaviour
in Serbia distinguish between violations, errors and lapses, which in the studies conducted
Serbia to date in other countries was not the case. The greatest number of violations in Serbia is
Crossing made by young persons and persons who walk by necessity, while persons who walk for
Mobile telephones pleasure and who walk the shortest distances show the biggest number of lapses. The oldest
persons and females generally showed positive behaviours to other road users, while the
increase of daily distances walked led to the increase of errors and positive behaviours.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Road safety depends on the behaviour of all road users, who are influenced by their attitudes. Therefore, the knowledge of
pedestrians’ attitudes and behaviour can help to understand the needs of this group of road users (Yannis, Golias, &
Papadimitriou, 2007). This is particularly significant for policy measures in the process of improving road safety in a partic-
ular local community.
Traffic accidents involving pedestrians most frequently happen when pedestrians cross the street. For example, in the
USA, during a three-year period, 63% of the traffic accidents including pedestrians happened while the pedestrian was cross-
ing the street (Da Silva, Smith, & Najm, 2003). The behaviour of both pedestrians and drivers can lead to an accident involv-
ing pedestrians. A study (Bungum, Day, & Henry, 2005) showed that around 15% of the pedestrians were killed in traffic
accidents due to the negligence of pedestrians.
Various studies have examined road users’ social attitudes and behaviour, out of which several (Assum, 1997; Louka,
Yannis, & Kanellaidis, 2004; Papadimitriou, Theofilatos, & Yannis, 2013; Vanlaar & Yannis, 2006) were based on the SARTRE
1-4 (Social Attitudes to Road Traffic Risk in Europe) research projects and aimed to analyse social attitudes and behaviour
towards risk in road traffic in Europe by means of questionnaires and personal interviews. The results of this research have

⇑ Corresponding author.
E-mail addresses: b.antic@sf.bg.ac.rs (B. Antić), d.pesic@sf.bg.ac.rs (D. Pešić), nenadmilu@yahoo.com (N. Milutinović), marko.maslac@yahoo.com
(M. Maslać).

http://dx.doi.org/10.1016/j.trf.2016.02.004
1369-8478/Ó 2016 Elsevier Ltd. All rights reserved.
 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178 171

shown that positive attitudes of all road users largely reduce the risk of involvement in traffic accidents. Positive attitudes
are shown by the pedestrians who have expressed their satisfaction with the traffic environment, were not annoyed by the
other road users, agree with reinforcement, traffic rules, in-vehicle devices and have careful behaviours (Papadimitriou et al.,
2013). The influence of gender on attitudes has shown that women show significantly more positive attitudes than men,
while the ratio between correct attitudes and age is positive among all road users in Europe.
Characteristics and personality traits are associated with differences in the behaviour of pedestrians when crossing the
street. A research that was based on the theory of planned behaviour, and which measured the tendency towards social con-
formity (Zhou, Horrey, & Yu, 2009) showed that subjective norms, perceived risk, self-identity and perceived behavioural
control were positively associated with crossing the street in potentially dangerous situations. Several studies, conducted
worldwide, have examined gender and age differences in pedestrian behaviour. Male pedestrians tend to violate traffic rules
more frequently than females and are more likely to cross in risky situations (Moyano Díaz, 2002; Rosenbloom, Nemrodov, &
Barkan, 2004). Generally, young adults and adolescent pedestrians are more likely to commit violations than older pedestri-
ans (Moyano Díaz, 2002), and older road users express more appreciation for controlled pedestrian crossings and signalized
intersections than younger pedestrians do (Bernhoft & Carstensen, 2008). Therefore, understanding pedestrian behaviours
when crossing remains a road safety challenge (Evans & Norman, 2003).
A self-report study which measured pedestrian behaviour was developed by several authors and was confirmed in their
countries. Moyano Díaz (1997) developed Pedestrian Behaviour Questionnaire PBQ in Chile. This tool made a difference
between violations, errors and lapses. After that, Yildirim (2007) developed a version of this tool in Turkey, differentiating
between violations, aggressive behaviours and errors. In both cases, the results showed that more violations were made
by men. Furthermore, more errors were noticed among young pedestrians (17–25 years) than among older pedestrians
(25–49 years). In contrast to them, Granié, Pannetier, and Guého (2013) developed a Pedestrian Behaviour Scale PBS in
France. This study validated the PBS for all ages, and made a difference between violations, errors and lapses, and also
provided an understanding of aggressive and positive behaviours by pedestrians towards other road users.
In 2014 in Serbia 128 pedestrians were killed in traffic accidents, which makes 23.9% of all fatalities. In the same year the
total number of injured pedestrians was 2827. In 1033 traffic accidents that occurred in 2014, it was pedestrians who were
responsible for the occurrence of traffic accidents (Road Traffic Safety Agency, 2015). Bearing this in mind, the aim of this
study is to obtain knowledge on pedestrian behaviours in Serbia, and to identify the differences in pedestrian behaviour
comparing them by gender, age groups, reasons for walking and daily distances walked. Of course, it is necessary to validate
a Serbian version of the Pedestrian Behaviour Scale (PBS), because until now these tools have not been used to evaluate the
behaviour of pedestrians in our country.
The Serbian version of the Pedestrian Behaviour Scale (PBS) suggests questions about the use of mobile phones while
crossing the street. The reason for the introduction of these questions is reflected in the fact that one of the pedestrians’ prob-
lems in terms of traffic safety is mobile phone use while crossing the street, as Hatfield and Murphy (2007) and Nasar, Hecht,
and Wener (2008) confirmed in their studies. The results of a study conducted in Serbia showed that the mobile phone use at
unsignalized intersections was 11.5%, while the mobile phone use at signalized intersections was 13.4% (Milenković, Maslać,
& Trifunović, 2014). These data clearly indicate the necessity of introducing the questions related to the use of mobile phones
while crossing the street, which was made in the Serbian version of the Pedestrian Behaviour Scale (PBS).

2. Material and methods

The questionnaire conducted in Serbia was based on the already confirmed version of Pedestrian Behaviour Scales (PBS)
(Granié, 2008; Granié et al., 2013) conducted in France. This version represents the unified whole of all previous tools which
were used for measuring the behaviour of pedestrians (Moyano Díaz, 1997; Torquato & Bianchi, 2010; Yildirim, 2007)
conducted in Chile, Brazil, Turkey. The Serbian version of the Pedestrian Behaviour Scale (PBS), in addition to conducting
questionnaires among the Serbian population, introduces an innovation related to the use of mobile phones by pedestrians
when crossing the street.
In Serbia, on the basis of the Law on the Road Traffic Safety, Article 96 (paragraph 2) was defined and reads as follows:
‘‘When crossing the street pedestrians must not use a mobile phone or use the headset on their ears”. Bearing this in mind,
the questions related to mobile phone use while pedestrians are crossing the street are grouped with violations. The ques-
tions refer to talking on mobile phones, reading the contents (text messages, the Internet) and listening to music.
The method of collecting the data was a questionnaire filled in by 415 participants. The questionnaires were distributed
to random pedestrians at several locations in Serbia. Items were presented in random order to avoid bias in participants’
answers.
The questionnaire had two parts. The first part contained the questions about the social and demographic characteristics
(gender, age, education), and the questions about the most frequent reasons for walking, the daily distance that a pedestrian
walked, the possession of driving licence and participation in traffic accidents. All the questions were closed-type and par-
ticipants circled one of the given answers.
The second part of the questionnaire was about the attitudes and behaviour of pedestrians in traffic, and it contained
twenty questions presented in Table 1, divided into five groups (violations, errors, lapses, aggressive behaviour and positive
behaviour). The first group contained the questions about violations in traffic. A violation was defined as an intentional
172 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178

Table 1
The questionnaire containing 20 questions (divided in groups).

Violation
1. I cross the street diagonally in order to save time
2. I do not cross the street at the marked pedestrian crossing even though there is one at the distance smaller than 50 m
3. I cross the street even though the pedestrian light is red
4. I cross the street even though the light is still green for vehicles
5. I use a mobile phone while crossing the street
1. I talk over the phone
2. I read the contents (text messages, the Internet)
3. I listen to music (I use earphones)

Error
6. I start street crossing at the marked pedestrian crossing but I finish outside of it
7. I cross the street between the parked cars
8. When I want to overtake a slow-moving person and I do not have enough space to do it on the sidewalk, I go onto the road
9. On a two-way street, I cross the first part and wait in the middle of the roadway to cross the second part

Lapse
10. Before crossing the street I did not look left and right because I was thinking about something else
11. Before crossing the street I did not look left and right because I was talking to a friend
12. Before crossing the street I did not look left and right because I was in a hurry
13. Before crossing the street I did not look left and right because there were pedestrians moving in front of me

Aggressive behaviour
14. When I get angry at the driver who did not give me priority at the marked pedestrian crossing, I insult him/her
15. When I get angry at the driver who did not give me priority at the marked pedestrian crossing, I wave my hand
16. I intentionally move slowly while crossing the street in order to annoy drivers

Positive behaviour
17. I am grateful to the drivers who give me priority to cross the street
18. While crossing the street I give priority to vehicles even though I am at the marked pedestrian crossing
19. I walk on the right-hand side of the sidewalk so as not to bother the pedestrians I meet
20. When there is a crowd at the pedestrian crossing I give priority to the pedestrians from the opposite direction

deviation from the legal rules guiding pedestrian behaviour (e. g. ‘‘I cross the street diagonally in order to save time”) and the
use of mobile phones while crossing the street (e.g. ‘‘I read the contents (text messages, the Internet)”. Errors were defined as
making decisions that put the pedestrian in danger, without disobeying the legal rules (e.g. ‘‘On a two-way street, I cross the
first part and wait in the middle of the roadway to cross the second part”). Lapses were defined as ill-suited behaviours
related to the lack of concentration on the task (e.g. ‘‘Before crossing the street I did not look left and right because I was
talking to a friend”). Positive behaviours were defined as behaviours that appease social interactions (e.g. ‘‘I am grateful
to the drivers who give me priority to cross the street”). Aggressive behaviours were defined as conflicting behaviours with
other road users (e.g. ‘‘I intentionally move slowly while crossing the street in order to annoy drivers”). In this part we used
the Likert scale with answers ranging from 1 to 6, 1 being ‘‘Never” and 6 ‘‘Very often”.
The question ‘‘I cross the street even though the pedestrian light is red‘‘ was used in the questionnaire to check the
consistency of answering, since it essentially has the same meaning as the question ‘‘I cross the street even though the light
is still green for vehicles”. These two items were highly correlated to each other (rho = 0.653, p < 0.001), meaning the con-
sistency was fair. This question (‘‘I cross the street even though the light is still green for vehicles”) is excluded from further
processing of results.
The data were analysed in the statistical software package IBM SPSS v. 22. The normality of distribution was tested using
the Kolmogorov–Smirnov test. Since distributions of the data largely deviated from the normal distribution, the non-
parametric methods were used. Absolute and relative frequencies, or median and the interquartile range were used as
descriptive measures of response values, while non-parametric analytical statistical methods (Spearman’s rank correlation,
Kruskal–Wallis test, Mann–Whitney’s U test and Pearson v2 test) were used to evaluate the statistical significance of
association or difference, where appropriate. The internal consistency of the questionnaire was assessed using Cronbach’s
alpha statistic.
A Principal Component Analysis (PCA) using the Kaiser’s criterion for factor extraction and the orthogonal Varimax
rotation method was performed to investigate the underlying structure of the questionnaire and to obtain dimensional
aggregated measures of the behaviours of interest. The threshold of the statistical significance has been set to the conven-
tional level of p 6 0.05.

3. Results

The research included 415 participants, out of which 228 were men (55%) and 187 were women (45%). The greatest
number of participants belonged to the youngest age group, aged 15–25 (38%). The reason why participants in Serbia walked
was predominately by necessity (61%). There was a statistically significant tendency for men to walk by necessity, while
 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178 173

Table 2
Description of the sample.

N %
Gender Men 228 54.9
Women 187 45.1
Age 15–25 158 38.1
26–35 82 19.8
36–45 76 18.3
46–55 66 15.9
>56 33 8.0
Reasons for walking Necessity 285 68.7
Pleasure 130 31.3
Daily distances walked <300 m 55 13.3
300–800 m 173 41.7
800–1300 m 135 32.5
>1300 m 52 12.5

women stated that they more frequently walked for pleasure (v2 = 24.82, p < 0.001). The distances pedestrians would walk
during the day were usually in the range of 300–800 m (41.7%) or 800–1300 m (32.5%). Men walked on average longer
distances than women, daily medians being 800–1300 m and 300–800 m, respectively (ZU = 2.54, p = 0.011). The paper
examined the association of gender, age, reasons for walking and distances daily walked and the behaviour of pedestrians.
The description of the sample is presented in Table 2.
Bearing in mind that this is the first use of this tool in Serbia, internal consistency and principal component structure of
the questionnaire was determined. The questionnaire had a fair internal consistency (Cronbach’s alpha 0.749), and principal
component analysis with orthogonal Varimax rotation was carried out on all 20 items in the scale. The scree plot indicated
that the data best fit a five-factor solution, which accounted for 66.4% of the total variance. The PCA obtained a five-factor
solution as presented in Table 3. The factor loadings corresponded very well to the intended meaning of the items.
Factor 1 ‘‘violation”, explained 21.4% of the variance. The violation was defined by 6 items, three of which related to the
way of crossing the street (e.g. ‘‘I cross the street diagonally in order to save time”, factor loading: .558), and three to the use
of mobile phones while crossing the street (e.g. ‘‘I read the contents (text messages, the Internet)”, factor loading: .811). The
items loading on this axis all had in common the intentional nature of the dangerous behaviour or a deliberate offence
contrary to the legal rules.
Factor 2 ‘‘lapse” explained 14.3% of the variance. The lapse was defined by 4 items. In the classification by Reason,
Manstead, Stradling, Baxter, and Campbell (2011), lapses were defined as ill – suited behaviours related to the lack of
concentration on the task (e.g. ‘‘Before crossing the street I did not look left and right because I was talking to a friend”, factor
loading: .876).
Factor 3 ‘‘Positive behaviours” explained 12.5% of the variance. Positive behaviours were defined by 4 items. Positive
behaviours were defined as behaviours that appease social interactions, (e.g. ‘‘I am grateful to the drivers who give me
priority to cross the street”, factor loading: .513).
Factor 4 ‘‘error” explained 10.7% of the variance. The error was defined by 4 items. An error is a consequence of wrong
pedestrians’ decisions when crossing the street (e.g. ‘‘On a two-way street, I cross the first part and wait in the middle of
the roadway to cross the second part”, factor loading: .633).
Factor 5 ‘‘aggressive behaviours” explained 7.5% of the variance. Aggressive behaviours were defined by 3 items.
Aggressive behaviours were defined as conflicting behaviours with other road users, in this case towards the driver (e.g.
‘‘I intentionally move slowly while crossing the street in order to annoy drivers”, factor loading: .336).
Instead of single item scores, the component scores were used in further analyses as the measures of the underlying
dimensions of behaviour. Negative values of the median and percentiles were obtained in the following tables, because
the analysis used the factorial scores from the PCA.
Before presenting the most important results, it should be underlined that the most frequently self-reported behaviour
was the gratitude of the pedestrians to the driver who yielded priority to them at a marked pedestrian crossing with
60.6% of the participants circled very often. Thus, it can be concluded that pedestrians have an extremely high level of
positive behaviours towards drivers.
Table 4 shows all the questions related to the pedestrians’ behaviour in traffic. In addition to the median values, we also
presented the percentages of the answers on the Likert scale from 1 (Never) to 6 (Very often).

3.1. Gender and pedestrians’ behaviour

The statistical comparison of the frequency of certain pedestrians’ behaviours across genders is presented in Table 5. Men
made statistically significantly more violations (ZU = 3.32, p = 0.001), errors (ZU = 2.48, p = 0.013) and were more prone to
174 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178

Table 3
Principal axis factor analysis of the 21 items in the scale (Varimax rotation).

Rotated component matrix

Component (factor)
1 2 3 4 5
Violation Lapse Positive Error Aggressive
I cross the street even though the pedestrian light is red .856
I read the contents (text messages, the Internet) .811
I listen to music (I use earphones) .757
I do not cross the street at the marked pedestrian crossing even though there is one at the .647
distance smaller than 50 m
I cross the street diagonally in order to save time .558
I talk over the phone .361
When I want to overtake a slow-moving person and I do not have enough space to do it on the .709
sidewalk, I go onto the road
On a two-way street, I cross the first part and wait in the middle of the roadway to cross the .633
second part
I cross the street between the parked cars .588
I start street crossing at the marked pedestrian crossing but I finish outside of it .557
Before crossing the street I did not look left and right because I was in a hurry .886
Before crossing the street I did not look left and right because I was talking to a friend .876
Before crossing the street I did not look left and right because I was thinking about something .866
else
Before crossing the street I did not look left and right because there were pedestrians moving in .673
front of me
When I get angry at the driver who did not give me priority at the marked pedestrian crossing, I .766
wave my hand
When I get angry at the driver who did not give me priority at the marked pedestrian crossing, I .702
insult him/her
I intentionally move slowly while crossing the street in order to annoy drivers .336
When there is a crowd at the pedestrian crossing I give priority to the pedestrians from the .817
opposite direction
While crossing the street I give priority to vehicles even though I am at the marked pedestrian .764
crossing
I walk on the right-hand side of the sidewalk so as not to bother the pedestrians I meet .723
I am grateful to the drivers who give me priority to cross the street .513
Percent of explained variance in rotated sums of squared loadings 21.4% 14.3% 12.5% 10.7% 7.5%

aggressive behaviours (ZU = 6.11, p < 0.001). While the lapses (ZU = 0.90, p = 0.370) and positive behaviours (ZU = 0.71,
p = 0.476) are not statistically significantly different between genders.

3.2. The association of age and pedestrians’ behaviours

Age was statistically significantly positively associated with lapses (rho = 0.25, p < 0.001) and positive behaviours
(rho = 0.16, p = 0.002), while negatively with violations (rho = 0.37, p < 0.001) and errors (rho = 0.23, p < 0.001). The
aggressive behaviours (rho = 0.01, p = 0.869) were not associated with age. The results also showed a statistically significant
difference in terms of violations (H = 5.681, p < 0.001) and errors (H = 6.585, p < 0.001) between the age groups 36–45 and
>56, and in terms of lapses (H = 3.901, p < 0.001) and positive behaviours (H = 4.081, p < 0.001) between the age groups
15–25 and 46–55 (see Table 6).

3.3. Reasons for walking and the behaviour of pedestrians

Table 7 represents the comparison of participants who stated they were walking predominantly by necessity with those
walking for pleasure. The pedestrians who walked by necessity reported making significantly more errors (ZU = 2.811,
p = 0.005), as well as displaying more aggressive behaviours (ZU = 3.138, p = 0.002) while crossing the street. Behaviours
relating to violations (ZU = 0.256, p = 0.798), lapses (ZU = 1.547, p = 0.122) and positive behaviours (ZU = 0.575,
p = 0.565) have not reached the threshold of statistical significance between the examined groups.

3.4. The association of daily distances walked and the behaviour of pedestrians

The daily distances walked were significantly positively associated with positive behaviours (rho = 0.13, p = 0.007) and
errors (rho = 0.13, p = 0.007). They were not associated with violations (rho = 0.05, p < 0.327), aggressive behaviours
(rho = 0.04, p < 0.371) and lapses (rho = 0.09, p = 0.06), i.e. p didn’t reach the threshold of statistical significance. The results
also showed a statistically significant difference in terms of violations (H = 4.382, p < 0.001) and lapses (H = 6.544,
p < 0.001) between the group of pedestrians who walked the distance <300 m and 800–1300 m (see Table 8).
 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178 175

Table 4
The distribution of the participants’ answers according to the Likert scale.

Me Answers to %
1 2 3 4 5 6
I am grateful to the drivers who give me priority to cross the street 6 – 0.7 6.0 10.4 22.7 60.2
I walk on the right-hand side of the sidewalk so as not to bother the pedestrians I meet 5 2.2 6.7 19.5 14.5 42.5 14.7
While crossing the street I give priority to vehicles even thogh I am at the marked pedestrian 4 13.0 10.8 19.0 20.0 17.8 19.4
crossing
When there is a crowd at the pedestrian crossing I give priority to the pedestrians from the opposite 4 9.2 11.8 19.8 17.6 27.4 14.2
direction
I start street crossing at the marked pedestrian crossing but I finish outside of it 3 19.3 28.7 15.3 14.0 15.7 7.0
I cross the street between the parked cars 3 13.3 32.8 27.0 15.2 11.7 –
When I want to overtake a slow-moving person and I do not have enough space to do it on the 3 14.9 23.6 17.7 15.9 19.5 8.4
sidewalk, I go onto the road
On a two-way street, I cross the first part and wait in the middle of the roadway to cross the second 3 18.8 21.9 22.2 25.1 9.3 2.7
part
I cross the street diagonally in order to save time 3 12.8 32.0 22.7 14.9 7.0 10.6
I do not cross the street at the marked pedestrian crossing even though there is one at the distance 3 13.0 26.7 22.7 23.9 7.2 6.5
smaller than 50 m
I cross the street even tough the pedestrian light is red 2 44.6 34.0 12.8 5.1 1.3 2.2
Before crossing the street I did not look left and right because I was thinking about something else 2 41.9 23.4 17.3 13.3 2.7 1.4
Before crossing the street I did not look left and right because I was talking to a friend 2 40.0 23.1 15.7 15.2 4.6 1.4
Before crossing the street I did not look left and right because I was in a hurry 2 41.4 18.3 18.8 14.0 5.3 2.2
Before crossing the street I did not look left and right because there were pedestrians moving in front 2 27.0 28.2 13.7 16.1 8.9 6.1
of me
When I get angry at the driver who did not give me priority at the marked pedestrian crossing, I 2 34.7 21.9 17.3 11.1 12.8 2.2
insult him/her
When I get angry at the driver who did not give me priority at the marked pedestrian crossing, I wave 2 34.0 17.3 19.8 16.1 12.0 0.8
my hand
The use of mobile phone
 I talk over the phone 2 19.5 31.6 24.1 11.8 7.7 5.3
 I read the contents (text messages, the Internet) 2 40.5 35.2 13.0 6.3 2.3 2.7
 I listen to music (I use earphones) 1 57.7 9.6 14.9 6.5 5.8 5.5
I intentionally move slowly while crossing the street in order to annoy drivers 1 62.5 18.1 6.3 4.1 6.3 2.7
I cross the street even though the light is still green for vehicles 1 56.4 24.3 10.8 5.1 1.9 1.5

Table 5
Gender and pedestrians’ behaviour.

Gender Violation Error Lapse Aggressive behaviour Positive behaviour

Men Median 0.06 0.17 0.28 0.27 0.22
Perc. 25 0.63 0.84 0.93 0.37 0.45
Perc. 75 0.77 1.04 0.68 0.84 0.65
Women Median 0.36 0.39 0.01 0.36 0.27
Perc. 25 0.74 0.95 0.84 1.01 0.52
Perc. 75 0.43 0.94 0.96 0.29 0.80
Mann–Whitney U test ZU 3.32 2.48 0.90 6.11 0.71
p 0.001 0.013 0.370 <0.001 0.476

4. Discussion

Pedestrian Behaviour Scales (PBS) were validated for the pedestrians in Serbia. The scree plot indicated that the data best
fit a five-factor solution, which accounted for 66.4% of the total variance. The factor analysis identified five axes: violations,
errors, lapses, aggressive and positive behaviours. The results obtained from the Serbian version of the Pedestrian Behaviour
Scale (PBS) show the distinction between violations, errors and lapses. The presented Rotated Component Matrix displays
each of the items from this study on a separate axis. Those results are different from the results obtained Granié et al.
(2013) and Yildirim (2007). Granié et al. (2013) displayed violations and errors as one group of transgressions, because par-
ticipants in France did not distinguish between these two groups. On the other hand, the participants in Turkey (Yildirim,
2007) did not differentiate between errors and lapses. Displaying each item on a separate axis represents an improvement
compared to the previously disclosed results.
All dangerous pedestrian behaviours in traffic are risky (Granié et al., 2013), but it is necessary to make difference
between the intentional exposure to risk (violations), unintentional exposure to risk (errors) and making wrong decisions
due to the lack of concentration on the task (lapses). Pedestrians who are prone to violations will make them in almost
all circumstances (Lawton, Parker, Manstead, & Stradling, 1997), while the risk exposure as a consequence of errors and
lapses should not be expected in all situations. The results obtained from the Serbian version of the Pedestrian Behaviour
176 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178

Table 6
The association of age and pedestrians’ behaviours.

Age Violation Error Lapse Aggressive behaviour Positive behaviour

15–25 Median 0.24 0.17 0.73 0.00 0.19
Perc. 25 0.30 0.78 1.09 0.85 0.67
Perc. 75 0.84 1.01 0.28 0.52 0.48
26–35 Median 0.23 0.30 0.12 0.05 0.27
Perc. 25 0.62 0.64 0.69 0.49 0.01
Perc. 75 0.14 0.98 0.72 1.25 0.79
36–45 Median 0.03 0.01 0.15 0.01 0.36
Perc. 25 0.65 0.74 0.39 1.06 0.76
Perc. 75 0.77 1.19 1.09 0.54 0.91
46–55 Median 0.54 0.53 0.38 0.12 0.13
Perc. 25 0.93 0.89 0.86 0.67 0.28
Perc. 75 0.14 0.70 0.68 0.67 0.67
>56 Median 0.75 1.11 1.02 0.23 0.58
Perc. 25 1.01 1.23 0.67 0.87 0.22
Perc. 75 0.64 0.97 1.52 0.25 1.08
Spearman rho test rho 0.37 0.23 0.25 0.01 0.16
p <0.001 <0.001 <0.001 0.869 0.002

Table 7
Reasons for walking and the behaviour of pedestrians.

Reasons for walking Violation Error Lapse Aggressive behaviour Positive behaviour
Necessity Median 0.21 0.17 0.29 0.16 0.22
Perc. 25 0.65 0.78 0.97 0.68 0.44
Perc. 75 0.51 0.96 0.74 0.69 0.74
Pleasure Median 0.26 0.61 0.00 0.21 0.27
Perc. 25 0.66 0.94 0.82 0.91 0.54
Perc. 75 0.60 0.69 0.84 0.41 0.70
Mann–Whitney U test ZU 0.256 2.811 1.547 3.138 0.575
p 0.789 0.005 0.122 0.002 0.565

Table 8
The association of daily distances walked and the behaviour of pedestrians.

Daily distances walked Violation Error Lapse Aggressive behaviour Positive behaviour
<300 m Median 0.25 0.38 0.19 0.33 0.21
Perc. 25 0.75 0.61 0.05 0.87 0.70
Perc. 75 0.01 1.04 0.68 1.52 0.71
300–800 m Median 0.26 0.58 0.68 0.07 0.14
Perc. 25 0.63 0.99 0.94 0.60 0.77
Perc. 75 0.48 0.60 0.27 0.55 0.76
800–1300 m Median 0.21 0.01 0.00 0.24 0.28
Perc. 25 0.65 0.73 0.65 1.06 0.26
Perc. 75 0.84 0.97 1.29 0.52 0.80
>1300 m Median 0.06 0.71 0.10 0.23 0.22
Perc. 25 0.42 0.12 0.98 0.34 0.05
Perc. 75 0.29 1.13 0.96 0.62 0.65
Spearman rho test rho 0.05 0.13 0.09 0.04 0.13
p 0.327 0.007 0.06 0.371 0.007

Scale (PBS) show that the Serbian participants distinguish between these groups of behaviour, which is extremely important
for understanding their behaviour.
The conducted Pedestrian Behaviour Scales – PBS in Serbia examined the association of gender, age, reasons for walking
and distances daily walked and the behaviour of pedestrians.

4.1. Comparison by gender

The comparison by gender showed some important differences in behaviour of men and women. Men are more prone to
violations, errors and aggressive behaviour. These results confirm the results obtained in other studies (Granié et al., 2013;
 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178 177

Moyano Díaz, 1997, 2002) that men are more prone to violations and express more aggressive behaviour than women. These
results were expected bearing in mind the gender differences in the society and the structure of killed pedestrians and
pedestrians punished for the committed violations in traffic. In both cases men are more represented than women (Road
Traffic Safety Agency, 2015). These data tell us that women are more careful, they respect regulations, they care about other
persons, while men are prone to competing and proving themselves, which is reflected in their behaviour.

4.2. Comparison by age groups

The results showed that age was statistically significantly positively associated with lapses and positive behaviours, while
negatively with violations and errors. These results were confirmed in the study of Torquato and Bianchi (2010), where
younger pedestrians (17–25) showed more violations than older pedestrians (25–49). Yildirim (2007) showed that the
increase in age led to the reduction in violations, for the drivers and pedestrians. The same result was confirmed in the stud-
ies conducted by Rosenbloom et al. (2004) and Moyano Díaz (2002). As mentioned above, the participants were divided into
five age groups in order to understand their attitudes and behaviour better and possibly recognize the differences between
these groups. The most striking differences were noticed between the youngest and oldest pedestrians. These differences
concerned violations, especially the use of mobile phones while crossing the street. Naturally, the young (age group 15–
25) used mobile phones while crossing the street most, and the oldest persons (>56 years of age) least. The results showed
that the oldest pedestrians made the smallest number of errors and violations, but the biggest number of lapses. This result
was expected because the oldest persons have reduced psycho-physical abilities, they are aware of it, and are therefore less
exposed to risk when crossing the street. But on the other side, these reduced psycho-physical abilities lead directly to the
reduction of concentration, which is reflected in the increased number of lapses when crossing the street. The oldest group of
respondents in Serbia also showed the greatest number of positive behaviours towards other traffic participants, while in the
study in France (Granié et al., 2013) the age group of 36–45 years showed the biggest number of positive behaviours. This
result could be the consequence of cultural and sociological differences between the two countries. Namely, the standards of
living are not the same in Serbia and France. A great number of middle-aged persons in Serbia are unemployed which leads
to their daily discontent. These persons do not have the option of using the means of transport so they often participate in
traffic as pedestrians. Their everyday life discontent is reflected in the reduction of the number of positive behaviours to
other road users.

4.3. Comparison by the reason for walking

The pedestrians who walked by necessity, i.e. who do not have a possibility of using other means of transport to satisfy
their needs, stated that they made more errors. On the other hand, persons who walked for pleasure were prone to lapses. As
for the behaviour to other road users, persons who walked by necessity were inclined to aggressive behaviour. The results of
the research conducted by Granié et al. (2013) are similar or identical to ours. The effects of the reason for walking have
already been shown for the perception of the environment, in terms of safety and security (Miaux, 2008). The pedestrians
who walk by necessity do it very often, almost every day (to go to the office, school, to the shops). Because of their obliga-
tions, they hurry, choose the shortest way and focus on the goal of their trip. This relation has the effect of making errors and
expressing aggressive behaviours to other road users.

4.4. Comparison by the daily distances walked by pedestrians

The daily distances walked are positively associated with errors and positive behaviour. Errors were most present in the
groups that walked the longest distances in a day (800–1300 m and >1300 m), which was expected due to the fact that the
pedestrians’ concentration declines with the longer distance walked. Daily distances walked were positively associated with
errors in the study by Granié et al. (2013). Positive behaviours were positively connected to the increase in daily distances
walked, because persons who walked the longest distances in a day (800–1300 m and >1300 m) are those who walk for plea-
sure. These results can be explained by the fact that the increase in daily distances walked leads to the increase in the pos-
sibility of deviant pedestrian behaviours due to pedestrians’ fatigue and declined concentration.

5. Conclusion

The behaviour measured by this questionnaire can represent a good approach to objective measuring of the risky beha-
viours which could lead to pedestrian injuries in traffic (Granié et al., 2013). Serbian version of a Pedestrian Behaviour Scale
(PBS) differs from the so far validated version of the Pedestrian Behaviour Scale (PBS) in its results, because the participants
in Serbia distinguish between violations, errors and lapses. Each item had a separate axis in order to make the best descrip-
tion of deviant behaviours and behaviour to other road users. These two items (violations and errors), which represent inten-
tional and unintentional risk exposure, have shown different associations while comparing pedestrians by gender, age,
reasons for walking and distances daily walked. The questions for the first time introduced in the Serbian version of the
Pedestrian Behaviour Scale (PBS) have certainly assisted the obtaining of this result, which greatly helps in understanding
178 B. Antić et al. / Transportation Research Part F 41 (2016) 170–178

the behaviour of pedestrians. These are the questions that relate to the use of mobile phones while crossing the street. Put-
ting the issues in the group of violations helped the pedestrians to distinguish between the groups of violations and errors
and lapses. The reason for the introduction of these questions is the increased use of mobile phones while crossing the street
in Serbia, which Pešić, Antić, Glavić, and Milenković (2015) confirmed in their research.
The results obtained in Serbia have some similarities with the results obtained in the previous studies (Granié, 2008;
Granié et al., 2013; Moyano Díaz, 1997; Yildirim, 2007). The similarities can be found in almost every comparison made
between the groups (by gender, age, reasons for walking, daily distances walked). The significant differences were found
in the participants’ answers in terms of percentage. Namely, the participants in Serbia had considerably bigger mean values
of answers than the participants in France regarding aggressive behaviour, and a lot smaller mean values of the answers
regarding positive behaviour towards other road users. This result can be explained as a consequence of social and cultural
differences between the participants in Serbia and France. Living standards in Serbia are lower than in France. A great num-
ber of persons in Serbia are unemployed which leads to their daily discontent. Their discontent is reflected in the reduction of
the number of positive behaviours in everyday life, in this case in traffic. In addition, this result can also be justified by lower
traffic and cultural education of the participants in Serbia.
The future studies whose objective is having safer pedestrians in traffic should be aimed at improving the questionnaires
and forming clearer and more specific questions about the mobile phone use while crossing the street. A separate direction of
this kind of research should be the comparison of the results obtained through a questionnaire and the results obtained
through field observation of pedestrians’ behaviour. That is, on the same field, in this case the territory of Serbia, it is nec-
essary to conduct the observation of pedestrians’ behaviour (using the method of scientific observation), to note the results
which would be later compared to the results obtained in the questionnaire. This approach would give comprehensive data
about the behaviour of pedestrians in traffic, as well as their vulnerability in traffic.

References

Assum, T. (1997). Attitudes and road accident risk. Accident Analysis and Prevention, 29(2), 153–159. http://dx.doi.org/10.1016/s0001-4575(96)00071-1.
Bernhoft, I. M., & Carstensen, G. (2008). Preferences and behaviour of pedestrians and cyclists by age and gender. Transportation Research Part F: Traffic
Psychology and Behaviour, 11(2), 83–95. http://dx.doi.org/10.1016/j.trf.2007.08.004.
Bungum, T. J., Day, C., & Henry, L. J. (2005). The association of distraction and caution displayed by pedestrians at a lighted crosswalk. Journal of Community
Health, 30(4), 269–279. http://dx.doi.org/10.1007/s10900-005-3705-4.
Da Silva, M. P., Smith, J. D., & Najm, W. G. (2003). Analysis of pedestrian crashes.Washington, DC: National Highway Traffic Safety Administration.
Evans, D., & Norman, P. (2003). Predicting adolescent pedestrians’ road-crossing intentions: An application and extension of the Theory of Planned
Behaviour. Health Education Research, 18(3), 267–277. http://dx.doi.org/10.1093/her/cyf023.
Granié, M. A. (2008). Influence de l’adhésion aux stéréotypes de sexe sur la perception des comportements piétons chez l’adulte. Recherche Transports
Sécurité, 101, 253–264.
Granié, M. A., Pannetier, M., & Guého, L. (2013). Developing a self-reporting method to measure pedestrian behaviors at all ages. Accident Analysis and
Prevention, 50, 830–839. http://dx.doi.org/10.1016/j.aap.2012.07.009.
Hatfield, J., & Murphy, S. (2007). The effects of mobile phone use on pedestrian crossing behaviour at signalized and unsignalized intersections. Accident
Analysis and Prevention, 39(1), 197–205. http://dx.doi.org/10.1016/j.aap.2006.07.001.
Law on the Road Traffic Safety (2009). Official Gazette of the Republic of Serbia No. 41/09, 53/10, 101/11.
Lawton, R., Parker, D., Manstead, A. S. R., & Stradling, S. G. (1997). The role of affect in predicting social behaviours: The case of road traffic violations. Journal
of Applied Social Psychology, 27, 1258–1276.
Louka, G., Yannis, G., & Kanellaidis, G. (2004). Investigation of European drivers’ attitude towards speed. In 2nd International conference for transport research
in Greece, Dover, Athens, Greece.
Miaux, S. (2008). Comment la façon d’’envisager la marche conditionne la perception de l’’environnement urbain et le choix des itinéraires piétonniers.
L’’expérience de la marche dans deux quartiers de Montréal. Recherche Transports Sécurité, 25/101, 327–351. http://dx.doi.org/10.3166/rts.101.327-351.
Milenković, M., Maslać, M., & Trifunović, A. (2014). The effects of mobile phone use on pedestrian crossing behavior at unsignalised intersections. In 9th
International conference road safety in local community, Zajecar, Serbia.
Moyano Díaz, E. (1997). Teoría del comportamiento planificado e intención de infringir normas de transito en peatones. Estudos de Psicologia (Natal), 2(2),
335–348. http://dx.doi.org/10.1590/S1413-294X1997000200008.
Moyano Díaz, E. (2002). Theory of planned behavior and pedestrians’ intentions to violate traffic regulations. Transportation Research Part F: Traffic
Psychology and Behaviour, 5(3), 169–175. http://dx.doi.org/10.1016/S1369-8478(02)00015-3.
Nasar, J., Hecht, P., & Wener, R. (2008). Mobile telephones, distracted attention, and pedestrian safety. Accident Analysis and Prevention, 40(1), 69–75. http://
dx.doi.org/10.1016/j.aap.2007.04.005.
Papadimitriou, E., Theofilatos, A., & Yannis, G. (2013). Patterns of pedestrian attitudes, perceptions and behaviour in Europe. Safety Science, 53, 114–122.
http://dx.doi.org/10.1016/j.ssci.2012.09.008.
Pešić, D., Antić, B., Glavić, D., & Milenković, M. (2015). The effects of mobile phone use on pedestrian crossing behaviour at unsignalized intersections –
Models for predicting unsafe pedestrians behavior. Safety Science, 82(2016), 1–8. http://dx.doi.org/10.1016/j.ssci.2015.08.016.
Reason, J., Manstead, A., Stradling, S., Baxter, J., & Campbell, K. (2011). Errors and violations on the roads: a real distinction? Ergonomics, 33(10–11),
1315–1332. http://dx.doi.org/10.1080/00140139008925335.
Road Traffic Safety Agency (2015). Statistical report on the state of traffic safety in the Republic of Serbia for the year 2014.
Rosenbloom, T., Nemrodov, D., & Barkan, H. (2004). For heaven’s sake follow the rules: Pedestrians’ behavior in an ultra-orthodox and a non-orthodox city.
Transportation Research Part F: Traffic Psychology and Behaviour, 7(6), 395–404. http://dx.doi.org/10.1016/j.trf.2004.10.004.
Torquato, R. J., & Bianchi, A. S. A. (2010). Comportamento de Risco do Pedestre ao Atraversaar a Rua: Um Estudo com Universitarios. Transporte: Teoria e
Aplicacao, 2(1), 19–41.
Vanlaar, W., & Yannis, G. (2006). Perception of road accident causes. Accident Analysis and Prevention, 38(1), 155–161. http://dx.doi.org/10.1016/j.
aap.2005.08.007.
Yannis, G., Golias, J., & Papadimitriou, E. (2007). Modeling crossing behavior and accident risk of pedestrians. Journal of Transportation Engineering. http://dx.
doi.org/10.1061/(ASCE)0733-947X(2007)133:11(634).
Yildirim, Z. (2007). Religiousness, conservatism and their relationship with traffic behaviours. Middle East Technical University. <https://etd.lib.metu.edu.tr/
upload/12608820/index.pdf>.
Zhou, R., Horrey, W. J., & Yu, R. (2009). The effect of conformity tendency on pedestrians’ road-crossing intentions in China: An application of the theory of
planned behavior. Accident Analysis and Prevention, 41(3), 491–497. http://dx.doi.org/10.1016/j.aap.2009.01.007.