Transportation Research Part F 55 (2018) 78–89

Contents lists available at ScienceDirect

Transportation Research Part F

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

Measuring risky driving behaviours among young drivers:

Development of a scale for the Oman setting
Hamed Al Reesi a,b,⇑, James Freeman c, Jeremy Davey c, Samir Al Adawi d, Abdullah Al Maniri e
a
Department of Family Medicine and Public Health, College of Medicine and Health Sciences, Sultan Qaboos University, Oman
b
North Batinah Governorate, Directorate General of Planning and Studies, Ministry of Health, Oman
c
Centre of Accident Research and Road Safety – Queensland (CARRS-Q), Queensland University of Technology, Australia
d
Department of Behavioural Medicine, College of Medicine and Health Sciences, Sultan Qaboos University, Oman
e
Research and Studies Department, Oman Medical Specialty Board, Oman

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

Article history: Background: A large body of road safety research has focused on developing self-report
Received 14 April 2016 measurement scales that identify the type and frequency of risk driving behaviours that
Received in revised form 19 February 2018 lead to crash involvement. With the dearth of Arabic measurement tools, the aim of the
Accepted 26 February 2018
study was to develop a modified, valid and reliable measurement tool that can be utilized
Available online 15 March 2018
among young drivers within the Oman context.
Methods: A total of 1319 (27.1% female) young drivers aged 17–25 years completed a
Keywords:
questionnaire that was distributed through a snowballing sampling technique across
Measurement tool
Risky driving behaviours
Oman. The survey included a range of demographic information and driving behaviours,
Self-report and utilized aspects of the Driver Behaviour Questionnaire (DBQ) and the Behaviour of
Crash involvement Novice Young Drivers Scale (BNYDS). An exploratory factor analysis (EFA) was undertaken
Young drivers to examine the factor structure of the modified 40-items tool.
Oman Results: A maximum likelihood and varimax rotation factor analysis revealed seven
behavioural dimension comprising 39 items, which explained 49.28% of the variance in
the behavioural scale of young drivers. These factors were transient violations (20.12% of
the variance), mood driving (7.03% of the variance), speeding (6.59% of the variance),
fatigue driving (4.36% of the variance), distracted driving (4.12% of the variance), seatbelt
usage (3.55% of the variance) and close following (3.51% of the variance). The composite
behavioural scale (39-items) showed an excellent internal consistency (a = 0.939) with
transient violations exhibiting the highest internal consistency (a = 0.927) and close
following showed the lowest internal consistency (a = 0.700). Crash predictability of the
seven behavioural dimensions was investigated (as 39.6% of the sample reported crash
involvement). Conducting logistic models between each behavioural dimensions and crash
involvement adjusted for drivers’ characteristics found that mood driving, fatigue driving
and distracted driving were strong predictors of crash involvement among young drivers.
However, consistent with previous research, the full model was not an efficient predictor of
crash involvement among the sample of young Omani drivers, as distracted driving was the
only significant predictor in the model.
Conclusions: The modified risky driving behaviours scale exhibited appropriate psychome-
tric properties and key aberrant driving behaviours were associated with crash involve-

⇑ Corresponding author at: Department of Family Medicine and Public Health, College of Medicine and Health Sciences, Sultan Qaboos University, Oman.
E-mail addresses: abdullah.a@omsb.org, adawi@squ.edu.om (H. Al Reesi).

https://doi.org/10.1016/j.trf.2018.02.030
1369-8478/Ó 2018 Elsevier Ltd. All rights reserved.
 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89 79

ment. This paper further outlines the key study findings and provides suggestions for
future research that aims to develop effective self-report methods to identify ‘‘at risk”
drivers.
Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction

The origin and nature of traffic crashes continue to be empirically and theoretically investigated in order to reduce the
road toll. An accumulating body of road safety literature indicates that risky driving behaviours are the leading cause of traf-
fic crashes, and thus, represent a major public safety concern (Houston, Harris, & Norman, 2003; Iversen, 2004). Additionally,
driving violations are the most common type of behaviour identified in young driver crash investigations, including exceed-
ing the speed limit (Scott-Parker, Hyde, Watson, & King, 2013), following too close to the vehicle in front (Fu & Wilmot,
2008), and dangerous overtaking (Fernandes, Hatfield, & Job, 2010). Driving violations and errors have also been found to
be common among young Omani drivers and they predict their crash involvement (Al Reesi et al., 2013). More specifically,
official statistics reveal driving violations (e.g. speeding and dangerous overtaking) to be the main contributing factors to
crash involvement in Oman (Royal Oman Police, 2014). Driver inattention has also been found to be a significant contributor
in young drivers’ crashes especially in the form of driver distraction (Neyens & Boyle, 2007). Mobile phones, inbuilt technol-
ogy in vehicles, eating and drinking are all common sources of distraction that young drivers tend to engage in while driving
(Shope, 2006). Furthermore, carrying many passengers particularly teens is another significant source of drivers distraction
(Fu & Wilmot, 2008).
In Oman, the overall contribution of risky driving behaviours to the occurrence of crashes has been estimated to be up to
95% of the total crashes (Royal Oman Police, 2014), which is in line with the global proportion of contribution. It has also
been revealed in the literature that individuals involved in one type of risky driving behaviour are likely to also engage in
other types (Begg & Langley, 2004). Furthermore, Iversen (2004) suggested that crash involvement in a specified period
reflects the amount of engagement in more risky driving behaviour in the same period.
As a result of the above, researchers have attempted to investigate risky driving behaviours through a range of self-report
instrument tools developed to provide a better understanding of the most common dimensions of such behaviours, partic-
ularly among ‘‘at risk” populations (Scott-Parker, Watson, & King, 2010). A large body of road safety research has focused on
developing self-report measurement scales that identify the frequency and type of risky driving behaviours that predict
crashes. Such endeavors over the past 20 years have produced an array of measures including the: Manchester Driver
Behaviour Questionnaire (DBQ) (Reason, Manstead, Stradling, Baxter, & Campbell, 1990), Driver Anger Scale
(Deffenbacher, Oetting, & Lynch, 1994), Driver Attitude Questionnaire (DAQ) (Parker, Stradling, & Manstead, 1996), Driving
Skill Inventory (Lajunen & Summala, 1997), the Safety Climate Questionnaire-MD (SCQ-MD) (Glendon & Stanton, 2000) and
the Behaviour of Novice Young Drivers scale (BNYDS) (Scott-Parker et al., 2010).
Self-report data is heavily relied upon because it offers a number of advantages associated with economy and simplicity
of use (af Wåhlberg, Dorn, & Kline, 2011). From the aspect that risky driving behaviours have unequal proportion of contri-
bution to the occurrence of crashes (Laapotti & Keskinen, 2004), it is crucial to investigate which types of behaviours are
most frequently exhibited by motorists and how much they contribute to crash involvement. However, a range of problems
have been demonstrated to exist with self-report tools, such as their lack of predictive power (af Wåhlberg et al., 2011;
Wishart, Freeman, Davey, Wilson, & Rowland, 2012), and a lack of ability to be able to distinguish between several forms
of risky driving behaviours (Scott-Parker et al., 2010).
The most well-known self-report tool developed to measure risky driving behaviour is the Driver Behaviour Question-
naire (DBQ) which was originally developed by Reason et al. (1990). It was designed mainly to measure two dimensions
of self-reported behaviours: ‘‘violations” and ‘‘errors”. ‘‘Violations are operationalized into two types- ‘ordinary violations’
(e.g. Speeding and crossing red lights) and ‘aggressive violation’ (e.g. honking to display anger or aggressive stance). ‘Errors’
refer to misjudgment such as failing to notice a cyclist or a pedestrian. The DBQ has been widely utilized within a range of
cross-cultural populations (de Winter & Dodou, 2010; Lajunen, Parker, & Summala, 2004; O’zkan, Lajunen, Chliaoutakis,
Summala, & Parker, 2006) including Arab communities such as the UAE, Qatar (Bener, Ozkan, } & Lajunen, 2008) and Oman
(Al Reesi et al., 2013). Although, the DBQ was mainly developed to investigate driving behaviours among more experienced
drivers, it has also been used among less experienced drivers and young drivers (e.g. Al Reesi et al., 2013; Bianchi & Summala,
2004). The internal consistency of the tool has been found to be good (Xie & Parker, 2002) or acceptable (Al Reesi et al., 2013;
Bener et al., 2008; Davey, Wishart, Freeman, & Watson, 2007).
DBQ is one of the most widely researched instruments in traffic and behaviour. The semantics employed in DBQ still dom-
inate the field but increasing loosing its relevancy due evolving factors that trigger traffic mishaps. The DBQ has attracted
increasing criticism in recent times, particularly due to its poor predictive ability (af Wahlberg, Dorn, de Winter, Dodou,
& Freeman, 2012), and omission of other aberrant driving behaviours that directly contribute to crashes. In regards to the
former, a recent meta-analytic review of the Driver Behaviour Questionnaire involving 174 studies estimated that the
‘‘violations” construct (which is a sub-factor of the scale) predicted crashes with correlations of 0.07 reported in multivariate
80 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89

analysis. Similarly, ‘‘errors” had similar correlations of 0.06 in regards to the same outcome variable (de Winter & Dodou,
2010). In fact, research with fleet drivers has demonstrated that exposure to the road (e.g., kilometers driven per year) is
more effective at predicting self-reported crashes than measurement scales (Freeman, Davey, & Wishart, 2008), even after
modifying the measures to meet cultural sensitivities (Davey et al., 2007). In regards to the latter, it may also be argued that
popular driving assessment scales are somewhat simplistic and are becoming increasingly antiquated (Wishart et al., 2012).
For example, popular scales do not measure a range of contemporary issues that affect driving performance such as fatigue,
and time pressure. Today, mobile phone usage is a significant source of distraction and a reason for crash involvement among
youth (Al Reesi et al., 2013; Poysti, Rajalin, & Summala, 2005; Riquelme, Al-Sammak, & Rios, 2010). Within the current con-
text, Al Reesi et al. (2013) found a high prevalence of mobile phones use while driving among a sample of younger Omani
drivers.
One of the more recently developed tools designed to address risky driving behaviours (and consider a range of contempo-
rary driving issues) for young drivers is the Behaviour of Novice Young Drivers Scale (BNYDS). Scott-Parker et al. (2010) developed
the measurement tool to examine a target population of young (i.e. novice) drivers in the three stages of the graduate driving
licensing (GDL) system in Queensland, Australia. BNYDS was developed to be a comprehensive measurement scale, as its items
include a range of driving related behaviours as well as items related to the requirements of the GDL system. For the develop-
ment of the BNYDS, sixty-three items were identified to assess a variety of driving behaviours, however, the corresponding fac-
tor analysis revealed a five factor solution comprising 44 items which accounted for 55% of the variance in risky driving
behaviours among a sample of young novice drivers. These five factors are transient rule violations (decisions undertaken dur-
ing the journey), fixed rule violations (decisions undertaken before the journey), misjudgment (e.g., misjudging the speed of an
oncoming vehicle), risky driving exposure (e.g., driving at night) and driver mood (measuring driver emotions or mood). Testing
and validating the questionnaire among young drivers aged 17–25 years resulted in a very high internal consistency for the
composite scale (a = 0.947) as well as for the five subscales (a = 0.923, 0.846, 0.870, 0.869, and 0.891 respectively) (Scott-
Parker et al., 2010). The five factors, particularly driver mood (correlation (r) ranged from 0.10 to 0.15), and their composite
scale (r = 0.016, p < 0.01) were found to correlate significantly with crash involvement. The scales were also found to correlate
significantly with self-reported intentions to bend the road rules among participants, and transient rule violations was the
strongest among the five factors (Scott-Parker et al., 2010). However, the BNYDS may not be easily transferrable to other inter-
national driving environments and samples as it contains items related to Queensland’s GDL system that do not exist in other
parts of the world. Another problematic issue may be the inclusion of sensitive questions such as ‘‘You drove after taking an illicit
drug such as marijuana or ecstasy”, which hold legal consequences as per Oman’s laws.
Building upon the BNYDS and several versions of the DBQ, the current study aimed to develop a modified, valid and a reliable
tool to measure a range of contemporary risky driving behaviours among young Arabic speaking drivers. There remains a strong
need to design a culturally-specific instrument tapping contemporary risky driving behaviours in the Omani context, as Oman
is one of the Arab countries which consistently accumulates one of the highest global traffic related crash records. The aim of the
current study was to develop a modified tool that incorporates the BYNDS and DBQ concepts, and which has acceptable psy-
chometric properties in regards to validity and reliability. The second aim was to investigate the capability of the newly devel-
oped tool to predict self-reported crash involvement among a sample of young Omani drivers.

2. Methods

2.1. Study design and participants

A cross-sectional study was conducted to measure risky driving behaviours among a sample of Omani young drivers aged
between 17 and 25 years across all regions of Oman. Sultan Qaboos University has student body from all corners of the coun-
try, that is, 11 governorates or Muhafazah. Representative students from each Muhafazah were contacted by the research
assistant who detailed them the scope of the study. The regional representatives were requested to distribute questionnaires
through youth assembly places (i.e. sport clubs, summer centers etc.). The returned questionnaires appear to have captured
the diversity of geography and societies of Oman as well as different socio-demographic characteristics (i.e. Gender, working
status, place of residence, and years of licensing).
This study solicited more than 1600 cases. A total of 1319 were complete cases while the remaining were deleted because
of exclusion criteria (unlicensed, out of age range) and missing values (75 cases). Thus, the present quest for the complete
cases with no missing values.
The study was conducted from January to March 2015 and targeted Omani drivers who held a full driving licence (i.e. as
Oman has no graduated driving licensing system, all drivers hold non-restricted licences issued after successfully passing
driving tests). Since they account for approximately 17% of licence holders in Oman (ROP, 2014), female drivers were over
represented in the current study with a proportion of 27.1% (n = 357).

2.2. Sampling technique

To increase the accessibility to female drivers in the general population and to ensure the participation of young drivers
with varied driving experiences and socio-demographic backgrounds, a snowballing sampling technique was used to
 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89 81

approach young drivers from all governorates of Oman. Students from Sultan Qaboos University (SQU) were employed to
distribute the survey within all states in their governorates through personal contact. This technique assisted in recruiting
357 female drivers (27.1%) and 628 employee drivers (47.5%) with an average driving experience of 2.69 years (S.D = 1.73).

2.3. Instrument

To achieve the objectives of the study, a self-reported questionnaire was developed as a combination of BNYDS and DBQ,
plus other relevant items from the litatures. The draft instrument was then translated into Arabic using well-established for-
ward translation-back translation protocols. The questionnaire consisted of the following sections:

Socio-demographic characteristics: including age, gender, marital status, working status and vehicle ownership.
Driving history: including information regarding the initial driving instructors (the first driver under whom the respon-
dent started to learn driving), unsupervised on-road driving before licensing, driving experience in years (years of holding
driving licence) and driving exposure in terms of weekly average driving distances in kilometers (km) and driving time in
hours.
Risky driving behaviour: the main dependent variable in the study is the risky driving behaviours among young drivers
in Oman. A list of items was developed to measure risky driving behaviours known to be common among young drivers
(ROP, 2014) i.e. speed, aggressive driving, overtaking, inattention and distracted driving. These items were drawn from
past research using the DBQ in Oman (Al Reesi et al., 2013) and BNYDS (Scott-Parker et al., 2010) as well as items guided
by the literature review and Oman’s traffic regulations. In total, only seven items were taken from the DBQ and BNYDS
without any modification (Items # 7, 8, 11, 19, 21, 22 and 37 as shown in Table 1) while other items were either modified
from the DBQ and BNYDS or developed to satisfy Oman’s traffic regulation environment. Participants were required to
provide a response to the question ‘‘In the last twelve months, how often have you done the following behaviours while driv-
ing?” in a 5-point Likert-type scale ranging from 1 (never) to 5 (Always).
Crash and offence history: comprising questions related to the history of involvement in crashes including at-fault
crashes, and the history of offences committed by the participants during the last twelve months of regular driving.

2.4. Instrument validity

During the process of checking the content and clarity of the instrument, the questionnaire was sent to 10 experts in the
study field from various backgrounds, academics (n = 8) and police officers (n = 2). They were asked to provide professional
opinions regarding each item in terms of its ability to measure risky driving behaviours suitable for the Oman context, with
particular attention to items clarity. The feedback received from the experts was useful for further improvement of the
instruments. Only 15% of the items received suggested amendments, while 9% of the items received comments regarding
similarities with other items and were subsequently deleted. After receiving and revising the feedback, a modified version
of the instrument was piloted.

2.5. Pilot study

A pilot study was undertaken among a small sample of the target group (n = 42) to further examine the content validity of
the scale. Participants were asked to evaluate the readability and clarity of the items. Further modifications to the language
(in order to increase readability) were undertaken to a number of items. A final version of the scale was prepared consisting
of 40 items, and was then distributed to the target population.

2.6. Statistical analysis

The Statistical Package for the Social Sciences (SPSS) version 21 was utilized in the analysis of the data. Initially, an
exploratory factor analysis (EFA) was undertaken to identify the best factor structure for the 40-items behavioural scale. Fac-
tors were extracted through the Maximum Likelihood method with eigenvalues greater than 1.00 and a varimax rotation
method with Kaiser Normalization. Factor loading values less than 0.40 were suppressed from inclusion in the output. Sam-
pling adequacy and sphericity of the extraction was tested using the Kaiser-Meyer-Olkin (KMO), the recommended value is
at least 0.6, and the Bartlett’s test (presence of significance is recommended). Hair, Anderson, Tatham, and Black (1998) rec-
ommended a minimum sample size of 5 observations per variable for an accurate EFA. Since the requirement of minimum
sample size was met with 1319 participants, the EFA was then conducted.
Internal consistency of extracted factors was evaluated through Cronbach’s alpha. An alpha coefficient of 0.7 or above is
considered acceptable (Field, 2005). Descriptive statistics of extracted factors were calculated including mean, standard devi-
ation, range, skewness and Kurtosis. Pearson’s moment correlations were examined to measure the correlation between
extracted factors. Chi-Square (v2) tests of independence were used to assess the statistical differences between crash
involvement and socio-demographic characteristics. The strength of the Chi-square associations was assessed using the
phi coefficient for 2 ⁄ 2 tables and Cramer’s V coefficient for greater than 2 ⁄ 2 tables (Field, 2005). Finally, logistic models
82 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89

Table 1
Factor loadings of the 39 items in the risky driving behaviour scale (RDBS).

Sr. Items Transient Mood Speeding Fatigue Distracted Seatbelt Close

Violations Driving Driving Usage Following
F1 F2 F3 F4 F5 F6 F7
24 Attempt turning without ensuring road is devoid of 0.742
pedestrians or cyclists (DBQ-modified)
19 Cross a junction knowing that the traffic lights have already 0.735
turned red (DBQ)
23 Turn right/left into the path of another vehicle putting it at a 0.730
risk or making it breaks suddenly (blind spot)
22 Turn using an illegal U-turn (BYNDS) 0.709
21 On entering a roundabout or intersection, you pay such close 0.680
attention to the mainstream of traffic that you nearly hit the
car in front (DBQ)
12 Attempt to overtake a row of cars in a traffic jam from right 0.678
hand side
9 Get involved in ‘drifting’ 0.675
20 Enter the road in front of another vehicle which forces it to 0.667
break suddenly
13 Attempt to overtake another car in an area where overtaking 0.644
prohibited
8 Get involved with unofficial ‘races’ with other drivers on the 0.629
roads (DBQ)
11 Attempt to overtake a car that you hadn’t noticed to be 0.587
signaling a left/right turn (DBQ)
14 Attempt to overtake a row of cars, stopped on roads, for any 0.566
reason
6 Exceed the posted speed limit when you drive in a bad road 0.551
conditions (i.e. Working zone, slippery roads)
17 Misjudge the stopping distance you needed which forces you 0.501
to suddenly use the breaks (DBQ & BYNDS-modified)
18 Cross a junction knowing that the traffic lights have already 0.485
turned yellow
25 Turn right/left, without signaling the turn 0.451
7 Drive close to the car in front as a signal to its driver to go 0.409
faster or get out of the way (DBQ)
39 Get angered by other slow drivers (DBQ-modified) 0.573
34 Watching views or events happened on roads while driving 0.525
33 Joking with my friends while driving 0.497
40 Using horn to indicate my anger from anther driver’s 0.493
behaviour (DBQ-modified)
32 Listening to a specific radio programs while driving 0.487
37 Your driving affected by negative emotions like anger or 0.451
frustration (BYNDS)
38 Drive faster if you were in a bad mood 0.4260
5 Exceed the posted speed limit when you drive on open roads 0.749
or roads with low traffic
4 Exceed the posted speed limit when you drive in areas where 0.748
it was unlikely there was a radar or speed camera (BYNDS-
modified)
3 Exceed the posted speed limit by more than 15 km/h (e.g. 120 0.528
km/h – I drive with 135 km/h or more) (BYNDS-modified)
2 Exceed the posted speed limit by less than 15 km/h (e.g. 120 0.424
km/h – I drive with 121–134 km/h) (BYNDS-modified)
10 Attempt to overtake a car in front even when it keeps 0.415
appropriate speed
27 Keep driving while you feel tired 0.869
28 Keep driving while you feel sleepy 0.547
26 Driving for long distance without taking breaks 0.494
30 Using a hand held mobile phone (Call or reply) while driving 0.702
(BYNDS-modified)
31 Using mobile phone for texting or chatting while driving 0.611
29 Ingestion while driving 0.512
36 Putting seatbelt on only in the presence of traffic police 0.785
35 Driving without putting seatbelt on (BYNDS-modified) 0.629
15 Drive close to the car in front which forces you to use the 0.611
breaks many times
 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89 83

Table 1 (continued)

Sr. Items Transient Mood Speeding Fatigue Distracted Seatbelt Close

Violations Driving Driving Usage Following
F1 F2 F3 F4 F5 F6 F7
16 Drive close to the car in front in traffic jam 0.460
Eigenvalues 7.85 2.74 2.57 1.70 1.61 1.39 1.37
% variance explained 20.12 7.03 6.59 4.36 4.12 3.55 3.51
Cronbach’s a 0.927 0.751 0.783 0.790 0.819 0.797 0.700
# of items 17 7 5 3 3 2 2

were implemented to investigate self-reported behavioural factors predictive of crash involvement adjusted for socio-
demographic characteristics.

3. Results

3.1. Factorial validity

An exploratory factor analysis (EFA) was undertaken to examine the factor structure of the 40-items driving scale. The
Maximum Likelihood method, incorporating a varimax rotation method with Kaiser Normalization, was used in the extrac-
tion. The most interpretable solution factors emerged from using 7 factors, as depicted from the eigenvalues and the scree
plot. The Kaiser-Meyer-Olkin (KMO) with a value of 0.946, above the recommended value of 0.6, showed good sampling ade-
quacy and Bartlett’s test of sphericity was significant at p < 0.01 (v2 = 24054.88, p < .01). Items that loaded with less than
0.40 upon any factor and items that loaded significantly upon more than one factor were excluded. All items loaded in
one of the seven interpretable factors except for one item. After the removal of the unloaded item, the EFA was implemented
with the remaining 39 items, and revealed 7 interpretable solution factors that explained 49.28% of the variance in the beha-
vioural scale of young drivers and all items loaded successfully into one of the seven factors. Table 1 presents the result of the
EFA including the factor loadings of the 39 items into the seven interpretable factors of the risky driving behaviour scale
(RDBS).
The first factor contained 17 items, accounted for 20.12% of variance, and appeared to measure regular violations charac-
terized primarily by being transient in nature. Such behaviours can be practiced by drivers multiple times during a journey
other than speeding, although it included one speeding item (item # 6), as well as dangerous overtaking, dangerous turning
and misjudgments. Therefore, it was given the name ‘‘Transient Violations”. Factor loadings ranged from 0.409 to 0.742, the
majority of which were greater than 0.500. The second factor contained 7 items, accounted for 7.03% of variance, and
appeared to measure driving behaviours influenced by the mood of drivers. It was given the name ‘‘Mood driving”. Factor
loadings ranged from 0.426 to 0.573, the majority of which were greater than 0.450. The third factor contained 5 items,
accounted for 6.59% of variance, and measured speeding behaviours in various road conditions, and thus it was called ‘‘Speed-
ing”. Factor loadings ranged from 0.415 to 0.749, and the majority of which were greater than 0.500. The fourth factor con-
tained 3 items, accounted for 4.36% of variance, and measured driving behaviours leading to fatigue such as continuing
driving even after feeling tired or sleepy. It was called ‘‘Fatigue”. Factor loadings ranged from 0.494 to 0.869, and all were
greater than 0.450. The fifth factor contained 3 items, accounted for 4.12% of variance, and appeared to measure driving
behaviours related to distraction, including mobile phone use (calling, replaying, texting or chatting) and consumption (eat-
ing or drinking), and thus it was called ‘‘Distracted behaviour”. Factor loadings ranged from 0.512 to 0.702, and all were
greater than 0.500. The sixth factor contained 2 items, accounted for 3.55% of variance, and appeared to measure the use
of seatbelts while driving and thus it was called ‘‘Seatbelt Usage”. Factor loadings ranged from 0.629 to 0.785, all of them were
greater than 0.600. Finally, the seventh factor contained 2 items and accounted for 3.51% of variance and appeared to mea-
sure driving behaviours related to headway (e.g., close following of vehicles) and thus it was called ‘‘close following”. Factor
loadings ranged from 0.460 to 0.611, and all were greater than 0.450.
Table 2 depicts the number of items, Cronbach’s alpha, means, standard deviations, factor means and range for the seven
factors for the composite RDBS. The composite risky driving behaviour scale (RDBS) with 39 items showed a very high inter-
nal consistency with a Cronbach’s alpha value of (0.939). Among the seven factors of the scale, the transient violations factor
exhibited the highest internal consistency (0.927), followed by the distracted driving factor (0.819). The other factors (mood
driving, speeding, fatigue, seatbelt usage and close following) also showed good internal consistency with at least 0.700
Cronbach’s alpha values. The sample reported an overall risky driving behaviour exposure of 2.41 per items. The highest
amount of risky driving behaviour was reported for speeding behaviour (M = 2.95) followed by mood driving (M = 2.91)
while the least likely behaviour committed was transient violations (M = 1.95). The distribution of the composite RDB scale
was approximately symmetrical to slightly positively skew with skewness of 0.58 and mesokurtic (normal) in shape with
kurtosis of 0.57. For the RDBS subscales, the distributions were approximately symmetrical to slightly skew as skewness ran-
ged from 0.11 to 1.19 and mesokurtic (normal) as kurtosis ranged from 1.07 to 1.24.
84 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89

Table 2
Risky driving behaviours scale (RDBS) psychometric properties.

Behaviour N. Items Cronbach’s alpha Mean SD Mean per item Range Skewness Kurtosis
Transient Violations 17 0.927 33.18 12.54 1.95 68 1.19 1.24
Mood Driving 7 0.751 20.34 5.45 2.91 28 0.01 0.15
Speeding 5 0.783 14.74 4.30 2.95 20 0.11 0.46
Fatigue 3 0.790 7.21 2.94 2.40 12 0.54 0.24
Distracted driving 3 0.819 8.39 3.10 2.80 12 0.21 0.61
Seatbelt usage 2 0.797 5.41 2.55 2.71 8 0.21 1.07
Close following 2 0.700 4.70 1.99 2.35 8 0.54 0.28
Composite RDBS 39 0.939 93.98 24.77 2.41 156 0.58 0.57

Pearson’s product moment correlation coefficients, values and significance for the RDBS subscales and the composite
RDBS are shown in Table 3. Overall, RDBS subscales correlated positively and moderately with each other with correlation
coefficients ranging between 0.33 and 0.62. The highest correlation coefficient (r = 0.627) was between transient violations
and close following followed by the correlation coefficient between mood driving and distracted driving behaviour (r =
0.551).

3.2. Risky driving behaviours and crash involvement

3.2.1. Crash involvement by young drivers’ characteristics

The portion of young drivers involved in road traffic crashes since holding their driving licence was 39.6% (n = 522) with a
total of 1108 crashes. Among those involved in crashes (n = 522), 60.7% (n = 317) of them were at-fault in at least one crashes
in which they were involved. In addition, more than half of the participants (56.9%, n = 751) reported at least one offence
while driving during the last twelve months.
Table 4 presents the association between crash involvement and the young drivers’ characteristics. Young drivers differed
significantly in their crash involvement by gender and the likelihood of crash involvement was higher among males. The
likelihood of crash involvement increased with age. Crash involvement differed significantly according to marital status,
with the likelihood being higher among married drivers. Crash involvement also differed significantly according to vehicle
ownership, with those who regularly drive their own vehicles reporting a higher likelihood of crash involvement.
Young drivers differed significantly in crash involvement according to their history of unsupervised driving prior to
licensing, and the likelihood was higher among those who experienced unsupervised driving prior to the licensing period.
The likelihood of crash involvement increased during the early years of driving and then started to decrease gradually.
The likelihood of crash involvement increased with distances driven and with driving hours.

3.2.2. Crash involvement by risky driving behaviours subscales

Initially, a simple logistic regression analysis was conducted with each behavioural subscale and crash involvement
among young drivers. Table 5 shows that the score in each of the seven risky behaviour types was greater among those
who reported involvement in a crash compared to those who did not report such involvement. The mood driving, speeding,
distracted driving, fatigue and close following behavioural subscales significantly predicted crash involvement among young
drivers, as well as the composite scale, except for the transient violations factor which was not statistically significant.
Secondly, a multiple logistic regression analysis was conducted between each behavioural subscale and crash involve-
ment adjusted for the influence of young drivers’ characteristics: age, gender, marital status, car ownership, unsupervised
driving, driving experience, driving distance and driving hours. Only mood driving, fatigue and distracted driving beha-
vioural subscales, as well as the composite scale, were significant in predicting crash involvement. Each one of the adjusted
multiple logistic models was significant and showed good levels of fit (P(Hosmer-Lemeshow test) > 0.05), except for speed-
ing and fatigue.

Table 3
Correlations between behavioural factors and the composite RDBS.

F1 F2 F3 F4 F5 F6 F7 RDBS
F1: Transient Violation 1
F2: Mood 0.415** 1
F3: Speeding 0.450** 0.435** 1
F4: Fatigue 0.516** 0.433** 0.366** 1
F5: Distraction 0.402** 0.572** 0.491** 0.529** 1
F6: Seatbelt 0.391** 0.469** 0.377** 0.365** 0.483** 1
F7: Unsafe Distance 0.622** 0.414** 0.437** 0.422** 0.386** 0.330** 1
Composite RDBS 0.877** 0.710** 0.675** 0.676** 0.683** 0.600** 0.694** 1
**
Pearson Correlation is significant at the 0.01 level.
 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89 85

Table 4
Association between young drivers’ characteristics and crash involvement.

Characteristics Crash Involvement Chi-square p-value Odds Ratio 95% C.I For OR
No (n = 797) Yes (n = 522)
N % N %
Gender 5.97 0.015
Male 562 70.5 400 76.6 Phi = 0.067 1.37 (1.06, 1.77)*
Female 235 29.5 122 23.4 1
Age 16.76 0.001
17–19 Years 150 18.8 55 10.5 V = 0.113 1
20–22 Years 366 45.9 257 49.2 1.92 (1.35, 2.71)*
23–25 Years 281 35.3 210 40.2 2.04 (1.43, 2.91)*
Marital status 7.76 0.005
Single 644 80.8 388 74.3 Phi = 0.077 1
Married 153 19.2 134 25.7 1.45 (1.12, 1.89)*
Working status 1.975 0.160
Student 430 54.0 261 50.0 1
Employee 367 46.0 261 50.0 1.17 (0.94, 1.46)
Vehicle ownership 19.29 0.001
Own vehicle 449 56.3 357 68.4 Phi = 0.121 1.68 (1.33, 2.11)
Others’ vehicle 348 43.7 165 31.6 1
Prior unsupervised driving 5.50 0.019
Yes 340 42.7 257 49.2 Phi = 0.065 1.30 (1.04, 1.63)
No 457 57.3 265 50.8 1
Driving experience (Years) 23.4 0.001
(0–1.99) Years 356 44.7 156 29.9 V = 0.157 1
(2.00–3.99) Years 277 34.8 207 39.7 1.71 (1.31, 2.21)
(>=4.00) Years 164 20.6 159 30.5 2.21 (1.66, 2.95)
Weekly driving distances (km) 14.71 0.001
<400 km 522 65.5 287 55.0 Phi = 0.126 1
>=400 km 275 34.5 235 45.0 1.55 (1.24, 1.95)
Weekly driving hours 13.06 0.001
<=7 h 452 56.7 243 46.6 Phi = 0.100 1
>7 h 345 43.3 279 53.4 1.50 (1.21, 1.88)
*
= statistical significant.

Table 5
Logistic model of the individual association between each behavioural subscale and crash involvement adjusted for the influence of structural characteristics.

Crash Not-Adjusted Adjusted Nagelkerke R Square Sig. (Hosmer-Lemeshow)

Involvement
(Mean)
Yes No Wald Sig. Wald Sig.
Transient violations 33.5 33.0 0.66 0.418 0.04 0.835 0.063 0.080
Mood driving 21.0 19.9 12.29 0.000 10.12 0.001 0.073 0.292
Speeding 15.2 14.5 8.80 0.003 3.51 0.061 0.066 0.021
Fatigue 7.6 7.0 15.72 0.000 8.61 0.003 0.071 0.015
Distracted driving 9.0 8.0 27.03 0.000 16.55 0.001 0.079 0.319
Seatbelt usage 5.6 5.3 4.36 0.037 2.19 0.139 0.065 0.200
Close following 4.8 4.6 3.86 0.049 1.18 0.277 0.064 0.197
RDBS 96.7 92.2 10.16 0.001 5.22 0.022 0.068 0.051

 Adjusted for: Governorate, Age, Gender, Marital status, working status, car ownership, Experience, Driving Distance.
 All Models were significant at p < 0.05, All models showed good fit with H&l (p > 0.05).
 Initially, the model run with driver characteristics only: N R square = 0.095; H&L (p > 0.05).

Finally, a full logistic model was conducted involving the seven behavioural factors and adjusted for the socio-
demographic characteristics of age, gender, marital status, car ownership, unsupervised driving, driving experience, driving
distance and driving hours (Table 6). The full model revealed that distracted driving (p = 0.019) was the only statistically
significant behaviour in predicting crash involvement among young drivers.
The full model was significant at 0.01 level indicating that the model predicted crash involvement significantly better
than with only the constant (chi-square (v2 = 86.17, p < 0.001). Driving experience and marital status were significant
predictors among the socio-demographic characteristics. Moreover, the Hosmer-Lemeshow test (v2 = 8.910, p = 0.350)
indicated a good level of fit.
86 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89

Table 6
The full logistic model of risky driving behaviours subscales as predictors of crash involvement adjusted for the influence of other characteristics.

B S.E. Wald Df Sig. Exp(B)

Gender 2.074 1 0.15
Male 0.212 0.147 2.074 1 0.15 1.236
Female 1
Age 3.031 2 0.22
17–19 Yrs. 1
20–22 Yrs. 0.307 0.196 2.442 1 0.118 1.359
23–25 Yrs. 0.164 0.222 0.545 1 0.46 1.178
Marital Status 5.744 1 0.017
Single 0.355 0.148 5.744 1 0.017 0.701
Married 1
Vehicle Ownership 3.304 1 0.069
Own Vehicle 0.236 0.13 3.304 1 0.069 1.267
Others’ Vehicle 1
Unsupervised Driving 0.097 1 0.755
Yes 0.04 0.127 0.097 1 0.755 1.04
No 1
Driving Experience 11.074 2 0.004
(0–1.99) Years 1
(2.00–3.99) Years 0.319 0.147 4.738 1 0.03 1.376
(>=4.00) Years 0.585 0.178 10.827 1 0.001 1.795
Driving Distances 1.59 1 0.207
<400 km 1
>=400 km 0.165 0.131 1.59 1 0.207 1.179
Driving Hours 0.326 1 0.568
<=7 h 1
>7 h 0.074 0.129 0.326 1 0.568 1.077
Transient violations 0.013 0.007 3.676 1 0.055 0.987
Mood driving 0.019 0.014 1.689 1 0.194 1.019
Speeding 0.006 0.017 0.134 1 0.715 1.006
Fatigue 0.041 0.026 2.564 1 0.109 1.042
Distraction driving 0.063 0.027 5.52 1 0.019 1.065
Seatbelt usage 0.014 0.028 0.245 1 0.621 0.986
Close following 0.005 0.039 0.017 1 0.898 1.005

4. Discussion

The current study aims were to develop a culturally specific instrument to measure the risky driving behaviour of young
drivers in Oman. Many instruments are available in the literature to measure risky driving behaviours, however, most of the
measures have not been scrutinized in terms of their application to cross-cultural populations, particularly in Gulf countries.
In contrast, most of the available instruments have been validated in Western-Europe, North America and Australia. Studies
from non-western populations have employed these instruments without subjecting them to culturally-specific psychome-
tric properties i.e. the DBQ as an example (Al Reesi et al., 2013; Bener et al., 2008). The current study utilized two scales (DBQ
and BNYDS) as well as items guided by literature review and Oman’s traffic regulation.
Factor analysis identified seven scales that were transient violations, mood driving, speeding, fatigue, distracted driving,
seatbelt usage and close following, which appears to be a good spread of contemporary driving behaviours. The psychome-
tric properties and the factorability of the developed tool were assessed and found to be valid and reliable. The internal con-
sistency analyses for the composite risky driving behaviour scale and each of the seven subscales indicated a good sound
reliability (a > 0.70). The scale with the highest internal reliability was ‘‘Transient violations” (a = 0.93) and the scale with
the lowest internal reliability was ‘‘close following” (a = 0.70). The seven factors explained 49.28% of the total variance. In
regards to the behaviours most frequently engaged in by the respondents, speeding was exhibited most often on the roads
followed by mood driving.
Transient behaviours involved several types of risky driving behaviours that are exhibited during the journey, including
dangerous overtaking and error-based behaviours that included items from the DBQ (Al Reesi et al., 2013; Bener et al., 2008;
Gras et al., 2006; Lajunen et al., 2004). This scale is similar to a previously extracted factor in the BNYDS. While speeding was
part of ‘‘transient violations” in BNYDS (Scott-Parker et al., 2010), in the current version, speeding was extracted separately
from transient violations indicating the prevalence of speeding behaviour (as a separate construct) among Omani drivers.
This finding is also consistent with previous DBQ-related study in Oman which has demonstrated speeding as a separate fac-
tor (Al Reesi et al., 2013). Speeding has also been widely acknowledged to be an important contributory factor to crash
involvement, particularly among young drivers (Bates, Davey, Watson, King, & Armstrong, 2014; Scott-Parker et al.,
2013). In Oman, more than 50% of crashes are related to speeding as depicted in the official statistical reports (ROP, 2014).
 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89 87

Mood driving related behaviours, that involved driving instances that could be influenced by emotions, also emerged as
an important factor among young drivers in Oman. This result also supports the previous finding Scott-Parker et al. (2010),
where mood driving was identified as an individual factor in the BNYDS. Further, mood driving was found to be a significant
predictor of crash involvement, which also aligns with previous findings that indicate driving when angry significantly pre-
dicts crash involvement (Dahlen & White, 2006). This was also confirmed by the Omani version of the DBQ, where the high-
est ranked item was the one that indicated the drivers’ hostile mood ‘‘Sound your horn to indicate your annoyance to another
driver” (Al Reesi et al., 2013).
The distracted driving factor in the current study involved eating and drinking while driving as well as using a mobile
phone for calling, texting or chatting. This was found to be the only behavioural predictor for crash involvement in the mul-
tivariate analysis. Similarly, Neyens and Boyle (2007) identified distracted driving as a significant predictor of young drivers’
crashes. Researchers found that young drivers who are distracted while driving, underestimate the potentially hazardous
consequences of their distraction (Horrey, Lesch, & Garabet, 2008). This finding is consistent with a large body of empirical
evidence that has indicated that mobile phone usage is one of the most common sources for distraction, especially among
young drivers in Oman (Al Reesi et al., 2013) and globally (Poysti et al., 2005; Riquelme et al., 2010). More specifically, visual
distraction has been shown to increase the likelihood of crash involvement (Klauer, Dingus, Neale, Sudweeks, & Ramsey,
2006) as well as near crashes (Rosenbloom, 2006). However, it is a difficult task for the ROP to identify mobile phone usage
as a cause of crashes, particularly in cases where the at-fault driver is deceased.
Fatigue driving was also extracted as a separate driving behavioural factor. Fatigue was measured as driving for long time
periods without rest or continuing to drive while tired or while feeling sleepy. Sufficient evidence exists that indicates fati-
gue is a very risky for young drivers (Hutchens, Senserrick, Jamieson, Romer, & Winston, 2008) and increases the likelihood
of their crash involvement (Klauer et al., 2006). Specific risks associated with fatigued driving include night driving (Rice,
Peek-Asa, & Kraus, 2003) and sleepiness (Mohamed et al., 2012). Similar to distraction, it is difficult for the ROP to identify
fatigue as a cause of crashes. This problem is not unique to Oman, but rather, is widely acknowledged in the literature (Grigo
& Baldock, 2011). However, there are some indications in official statistics that fatigue can be linked to night crashes and
single-vehicle crashes (Al Reesi et al., 2015). Thus, it can be suggested that fatigue is masked by other causes in the official
statistics. Previous research has revealed that young drivers overestimate their ability to control the negative effects of fati-
gue on driving performance (Lucidi et al., 2006) and thus they have a tendency to underestimate the associated risky con-
sequences (Smith, Carrington, & Trinder, 2005).
Tailgating (i.e. following too closely) was also found among common risky driving behaviours of young Omani drivers in
the current study and it emerged as a separate factor. This is consistent with the official statistics that indicates almost 7% of
crashes in Oman were due to following another vehicle too closely (ROP, 2014). Although, it wasn’t a strong predictor of
crash involvement in the current study, the results indicate that it is a common driving behaviour on Omani roads, which
also aligns with international young driver research (Neyens & Boyle, 2007). Tailgating drivers often do not have enough
time to avoid crashes, and this particularly affects young drivers, as they may require more reaction time to respond
(Caird, Willness, Steel, & Scialfa, 2008).
Seatbelt usage is also an important driving safety measure indicating a wider compliance of drivers to road rules. It might
not be a direct contributory factor in crash involvement but may be a significant factor in regards to the severity of crash
outcomes (Calisir & Lehto, 2002). Within the current context, it is a common traffic violation in Oman to travel unrestrained
by seatbelts, with more than 85,000 seatbelt offences registered in Oman in 2013 (ROP, 2014).

5. Conclusions

Overall, the current modified Arabic version tool to measure risky driving behaviours among young drivers in Oman
seems to show fair psychometric properties. The composite scale had good internal consistency while some of the subscales
require further reliability investigation. The predictive validity of the modified tool showed that the composite scale was
somewhat predictive of self-reported crash involvement, however consistent with the broader field, self-report scales in iso-
lation do not identify all individuals who report crash involvement (af Wåhlberg et al., 2011). Thus, further research should
consider the merit of combining self-report with official data sources. Within the scale, the distraction subscale was the
strongest predictor of crash involvement. However, the other behavioural subscales required further predictive validity
investigation.

6. Limitations

There are a number of methodological limitations present in the study and need to be considered when interpreting the
results. The study represented young drivers who participated in the study and may not be reflective of the wider driving
population. A self-reported questionnaire was utilized to collect the data and thus can contain methodological issues such
as response and recall bias. The tool was used to predict self-reported crash involvement rather than official crash records,
and some disparity may exist between the two data sources. Finally, it is noted in the text above that the alphas for Factor 1
and overall are >0.90. High alphas may indicate narrowness of construct or item redundancy. Thus, Streiner and Norman
88 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89

(2012) has suggested against employment of alphas greater than 0.90. To circumvest this limitation, a study is needed to
decipher confirmatory factor analysis in order to explore redundancy.
Despite the mentioned limitations, there is a dearth in tools developed to examine risky driving behaviours among young
drivers within the Omani context, and thus the current modified tool contributes to scientific knowledge in the area and
could be a reference for the future development of more efficient tools.

7. Future research

Future research on the modified risky driving behaviours scale and its seven subscales is required for further validity and
reliability investigations, which should be extended to further explore the predictive ability of the scale to identify crash
involvement and/or risky drivers. Future research can also include a comparison with official crash records. In addition, fur-
ther psychometric evaluation of this modified version using alternative methods such as confirmatory factor analysis (CFA)
needs to be considered to check the validity of the scale.

Acknowledgements

The current study is part of a research program investigating young drivers in Oman. Support and funding of the research
program is acknowledged to the Research Council in Oman.

References

af Wahlberg, A., Dorn, L., de Winter, J. C. F., Dodou, D., & Freeman, J. (2012). Commentaries and responses to ‘‘The Driver Behaviour Questionnaire as a
Predictor of Accidents: A Meta-analysis”. Journal of Safety Research, 43(1), 83–99.
af Wåhlberg, A., Dorn, L., & Kline, T. (2011). The Manchester Driver Behaviour Questionnaire as a predictor of road traffic accidents. Theoretical Issues in
Ergonomics Science, 12, 66–86.
Al Reesi, H., Al Maniri, A., Al Adawi, S., Davey, J., Armstrong, K., & Edwards, J. (2015). Prevalence and characteristics of road traffic injuries among young
drivers in Oman, 2009–2011. Traffic Injury Prevention. https://doi.org/10.1080/15389588.2015.1107679.
Al Reesi, H., Al Maniri, A., Plankermann, K., Al Hinai, M., Al Adawi, S., Davey, J., & Freeman, J. (2013). Risky driving behaviour among university students and
staff in the Sultanate of Oman. Accident Analysis and Prevention, 58, 1–9.
Bates, L., Davey, J., Watson, B., King, M. J., & Armstrong, K. (2014). Factors contributing to crashes among young drivers. Sultan Qaboos University Medical
Journal, 14(3), 297–305.
Begg, D. J., & Langley, J. D. (2004). Identifying predictors of persistent non-alcohol or drug-related risky driving behaviours among a cohort of young adults.
Accident Analysis and Prevention, 26, 1067–1071. https://doi.org/10.1136/ip.2007. 015925.
}
Bener, A., Ozkan, T., & Lajunen, T. (2008). The Driver Behaviour Questionnaire in Arab Gulf countries: Qatar and United Arab Emirates. Accident Analysis and
Prevention, 40, 1411–1417.
Bianchi, A., & Summala, H. (2004). The ‘‘genetics” of driving behaviour: Parents’ driving style predicts their children’s driving style. Accident Analysis and
Prevention, 36, 655–659. https://doi.org/10.1016/S0001-4575(03)00087-3.
Caird, J. K., Willness, C. R., Steel, P., & Scialfa, C. (2008). A meta-analysis of the effects of cell phones on driver performance. Accident Analysis and Prevention,
40, 1282–1293. https://doi.org/10.1016/j.aap.2008.01.009.
Calisir, F., & Lehto, M. R. (2002). Young drivers’ decision making and safety belt use. Accident Analysis and Prevention, 34, 793–805. https://doi.org/10.1016/
S0001-4575(01)00079-3.
Dahlen, E., & White, R. (2006). The Big Five factors, sensation seeking, and driving anger in the prediction of unsafe driving. Personality and Individual
Differences, 41, 903–915.
Davey, J., Wishart, D., Freeman, J., & Watson, B. (2007). An application of the driver behaviour questionnaire in an Australian organisational fleet setting.
Transportation Research Part F, 10, 11–21. https://doi.org/10.1016/j.trf.2006.03.001.
de Winter, J. C. F., & Dodou, D. (2010). The driver behaviour questionnaire as a predictor of accidents: a meta-analysis. Journal of Safety Research, 41,
463–470.
Deffenbacher, J. L., Oetting, E. R., & Lynch, R. S. (1994). Development of a Driving Anger Scale. Psychological Reports, 74, 83–91.
Fernandes, R., Hatfield, J., & Job, R. F. S. (2010). A systematic investigation of the differential predictors for speeding, drink-driving, driving while fatigued,
and not wearing a seat belt, among young drivers. Transportation Research Part F: Traffic Psychology and Behaviour, 13, 179–196.
Field, A. (2005). Discovering statistics using SPSS (2nd ed.). London: Sage.
Freeman, J., Davey, J., & Wishart, D. (2008). A study of contemporary modifications to the Manchester Drivers Behaviour Questionnaire for organizational
fleet settings. In Lisa Dorn (Ed.), Driver behaviour and training. Human factors in road and rail safety (vol 3, pp. 201–214). Ashgate: Aldershot.
Fu, H., & Wilmot, C. G. (2008). Effect of passenger age and gender on fatal crash risks of young drivers. Transportation Research Record: Journal of the
Transportation Research Board, 2078, 33–40. https://doi.org/10.3141/2078-05.
Glendon, A. I., & Stanton, N. A. (2000). Perspectives on safety culture. Safety Science, 34, 193–214.
Gras, M. E., Sullman, M. J. M., Cunill, M., Planes, M., Aymerich, M., & Font-Mayolas, S. (2006). Spanish drivers and their aberrant driving behaviours.
Transportation Research Part F: Traffic Psychology and Behaviour, 9, 129–137.
Grigo, J. A. L., & Baldock, M. R. J. (2011). Sleepiness and road crashes: Challenges of definition and measurement. CARS Report Series CASR082. Adelaide: The
University of Adelaide.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.). New Jersey: Prentice-Hall International.
Horrey, W. J., Lesch, M. F., & Garabet, A. (2008). Assessing the awareness of performance decrements in distracted drivers. Accident Analysis and Prevention,
40, 675–682. https://doi.org/10.1016/j.aap.2007.09.004.
Houston, J., Harris, P., & Norman, M. (2003). The aggressive driving behaviour scale: Developing a self-report measure of unsafe driving practices. North
American Journal of Psychology, 5(2), 269–278.
Hutchens, L., Senserrick, T. M., Jamieson, P. E., Romer, D., & Winston, F. K. (2008). Teen driver crash risk and associations with smoking and drowsy driving.
Accident Analysis and Prevention, 40, 869–876. https://doi.org/10.1016/j.aap. 2007.10.001.
Iversen, H. (2004). Risk-taking attitudes and risky driving behaviour. Transportation Research Part F, 7, 135–150.
Klauer, S. G., Dingus, T. A., Neale, V. L., Sudweeks, J. D., & Ramsey, D. J. (2006). The impact of driver inattention on near-crash/crash risk: An analysis using the
100-car naturalistic driving study data. DOT HS 810 594.
Laapotti, S., & Keskinen, E. (2004). Has the difference in accident patterns between male and female drivers changed between 1984 and 2000? Accident
Analysis and Prevention, 36, 577–584. https://doi.org/10.1016/S0001-4575(03)00064-2.
 H. Al Reesi et al. / Transportation Research Part F 55 (2018) 78–89 89

Lajunen, T., Parker, D., & Summala, H. (2004). The Manchester Driver Behaviour Questionnaire: a cross-cultural study. Accident Analysis and Prevention, 36(2),
231–238.
Lajunen, T., & Summala, H. (1997). Effects of driving experience, personality, and driver’s skill and safety orientation on speed regulation and accident. In T.
Rothengatter & E. Carbonell Vaya (Eds.), Traffic and transport psychology, theory and application (pp. 283–294). United Kingdom: Emerald Group
Publishing Limited.
Lucidi, F., Russo, P. M., Mallia, L., Devoto, A., Lauriola, M., & Violani, C. (2006). Sleep-related car crashes: Risk perception and decision-making processes in
young drivers. Accident Analysis and Prevention, 38(2), 302–309. https://doi.org/10.1016/j.aap.2005.09.013.
Mohamed, N., Mohd-Yusoff, M. F., Othman, I., Zulkipli, Z. H., Osman, M. R., & Voon, W. S. (2012). Fatigue-related crashes involving express buses in Malaysia:
Will the proposed policy of banning the early-hour operation reduce fatigue-related crashes and benefit overall road safety? Accident Analysis
Prevention, 45, 45–49.
Neyens, D. M., & Boyle, L. N. (2007). The influence of driver distraction on the severity of injuries sustained by teenage drivers and their passengers. Accident
Analysis and Prevention, 40, 254–259. https://doi.org/10.1016/j.aap.2007. 06.005.
O’zkan, T., Lajunen, T., Chliaoutakis, J., Summala, H., & Parker, D. (2006). Crosscultural differences in driving behaviours: A comparison of six countries.
Transportation Research Part F: Traffic Psychology and Behaviour, 9(3), 227–242.
Parker, D., Stradling, S. G., & Manstead, A. (1996). Modifying beliefs and attitudes to exceeding the speed limit: An intervention study based on the theory of
planned behaviour. Journal of Applied Social Psychology, 26, 1–19.
Poysti, L., Rajalin, S., & Summala, H. (2005). Factors influencing the use of cellular (mobile) phone during driving and hazards while using it. Accident Analysis
and Prevention, 37, 47–51. https://doi.org/10.1016/j.aap.2004.06.003.
Reason, J. T., Manstead, A. S. R., Stradling, S. G., Baxter, J. S., & Campbell, K. (1990). Errors and violations on the road: A real distinction? Ergonomics, 33,
1315–1332.
Rice, T. M., Peek-Asa, C., & Kraus, J. F. (2003). Nighttime driving, passenger transport, and injury crash rates of young drivers. Injury Prevention, 9(3),
245–250.
Riquelme, H. E., Al-Sammak, F. S., & Rios, R. E. (2010). Social influences among young drivers on talking on the mobile phone while driving. Traffic Injury
Prevention, 11(2), 127–132. https://doi.org/10.1080/15389580903536712.
Royal Oman Police (2014). Facts and Figures 2013. Directorate General of Traffic (DGT). Royal Oman Police (ROP), Muscat, Sultanate of Oman.
Rosenbloom, T. (2006). Driving performance while using cell phones: An observational study. Transportation Research Part F, 37, 207–212. https://doi.org/
10.1016/j.jsr.2005.11.007.
Scott-Parker, B., Watson, B., & King, M. J. (2010). The risky behaviour of young drivers: Developing a measurement tool. In Proceedings of the 24th Canadian
Multidisciplinary Road Safety Conference, June 6–9, Niagara Falls, Canada.
Scott-Parker, B., Hyde, M. K., Watson, B., & King, M. J. (2013). Speeding by young novice drivers: What can personal characteristics and psychosocial theory
add to our understanding? Accident Analysis and Prevention, 50, 242–250. https://doi.org/10.1016/j.aap.2012.04.010.
Shope, J. T. (2006). Influences on youthful driving Behaviour and their potential for guiding interventions to reduce crashes. Injury Prevention, 12(SI), i9–i14.
https://doi.org/10.1136/ip.2006.011874.
Smith, S., Carrington, M., & Trinder, J. (2005). Subjective and predicted sleepiness while driving in young adults. Accident Analysis and Prevention, 37,
1066–1073. https://doi.org/10.1016/j.aap.2005.06.008.
Streiner, D. L., & Norman, G. R. (2012). Mine is bigger than yours: measures of effect size in research. Chest, 141(3), 595–598. https://doi.org/10.1378/
chest.11-2473.
Wishart, D. E., Freeman, J. E., Davey, J. D., Wilson, A., & Rowland, B. D. (2012). When non-significance may be significant: Lessons learned from a study of the
development, implementation and evaluation of a fleet risk assessment tool. In Lisa Dorn (Ed.), Driver behaviour and training (pp. 197–214). Surrey:
Ashgate.
Xie, C. Q., & Parker, D. (2002). A social psychological approach to driving violations in two Chinese cities. Transportation Research Part F, 5(4), 293–308.
https://doi.org/10.1016/S1369-8478(02)00034-7.