This may be the author’s version of a work that was submitted/acceptedfor publication in the following source:
Oviedo Trespalacios, Oscar & Scott-Parker, Bridie(2018)The sex disparity in risky driving: A survey of Colombian young drivers.Traffic Injury Prevention, 19(1), pp. 9-17.
This file was downloaded from: https://eprints.qut.edu.au/107119/
c© Consult author(s) regarding copyright matters
This work is covered by copyright. Unless the document is being made available under aCreative Commons Licence, you must assume that re-use is limited to personal use andthat permission from the copyright owner must be obtained for all other uses. If the docu-ment is available under a Creative Commons License (or other specified license) then referto the Licence for details of permitted re-use. It is a condition of access that users recog-nise and abide by the legal requirements associated with these rights. If you believe thatthis work infringes copyright please provide details by email to [email protected]
Notice: Please note that this document may not be the Version of Record(i.e. published version) of the work. Author manuscript versions (as Sub-mitted for peer review or as Accepted for publication after peer review) canbe identified by an absence of publisher branding and/or typeset appear-ance. If there is any doubt, please refer to the published source.
https://doi.org/10.1080/15389588.2017.1333606
1
1
Title
The sex disparity in risky driving: A survey of Colombian young drivers
Authors
Oscar Oviedo-Trespalacios1,2,3, & Bridie Scott-Parker3,4,5,6
Affiliation 1 Queensland University of Technology (QUT), Centre for Accident Research and Road Safety
– Queensland (CARRS-Q), Institute of Health and Biomedical Innovation (IHBI), Australia
² Department of Industrial Engineering, Universidad del Norte, Colombia
3 Consortium of Adolescent Road Safety (cadrosa.org), Australia
4 Adolescent Risk Research Unit (ARRU), Sunshine Coast Mind and Neuroscience - Thompson
Institute, University of the Sunshine Coast (USC), Australia
5 Sustainability Research Centre (SRC), Faculty of Arts, Business and Law, University of the
Sunshine Coast (USC), Australia
6 School of Social Sciences, Faculty of Arts, Business and Law, University of the Sunshine
Coast (USC), Australia
Corresponding Author
Oscar Oviedo-Trespalacios
Queensland University of Technology (QUT)
Brisbane
Queensland, Australia, 4059
Tel: +61 7 3138 4967
Fax: + 61 7 3138 0111
Email: [email protected], [email protected]
2
2
Abstract
Objective: The overrepresentation of young drivers in poor road safety outcomes has long been
recognised as a global road safety issue. In addition, the overrepresentation of males in crash
statistics has also been recognised as a pervasive young driver problem. Whilst progress in road
safety evidenced as a stabilisation and/or reduction in poor road safety outcomes has been made
in developed nations, less-developed nations contribute the greatest road safety trauma, and
developing nations such as Colombia continue to experience increasing trends in fatality rates.
The aim of the research was to explore sex differences in self-reported risky driving behaviours
of young drivers, including the associations with crash-involvement, in a sample of young
drivers attending university in Colombia.
Methods: The Spanish version of the Behaviour of Young Novice Drivers Scale (BYNDS-Sp)
was applied to a sample of 392 students (225 males) aged 16-24 years attending a major
university in an online survey. Appropriate comparative statistics and logistic regression
modelling were used when analysing the data.
Results: Males reported consistently more risky driving behaviours, with approximately one-
quarter of all participants reporting risky driving exposure. Males reported greater crash-
involvement, with violations such as speeding associated with crash-involvement for both
males and females.
Conclusion: Young drivers in Colombia appear to engage in the same risky driving behaviours
as young drivers in developed nations. In addition, young male drivers in Colombia reported
greater engagement in risky driving behaviour than young female drivers; a finding again
consistent with the behaviours of young male drivers in developed nations. As such, the
research findings suggest that general interventions such as education, engineering, and
enforcement should target transient rule violations such as speeding and using a handheld
mobile phone while driving for young drivers in Colombia. Future research should investigate
how these interventions could be tailored specifically for the Colombian cultural context,
including how their effects can be evaluated, prior to implementation.
Keywords
Sex differences; young drivers; risk taking; BYNDS; crashes; risk exposure
3
3
The sex disparity in risky driving: A survey of Colombian young drivers
INTRODUCTION
Road trauma is a major public health concern with epidemic occurrence of fatalities and
injuries arising from road crashes. The World Health Organization (WHO) (2015) estimates
that every year, 1.24 million people die worldwide from road trauma. Additionally, on a daily
basis, at least 140,000 people are injured on the world's roads, and 15,000 of these are disabled
for life (Sharma 2008). The incidence of road trauma clearly differs based on each country’s
level of economic development. The road crash burden is disproportionately high in low- and
middle-income countries which register 91% of all fatalities (World Health Organization
2013). Within Latin America specifically, more than 130,000 fatalities and 6 million serious
injuries result from road crashes every year (Bezerra et al 2015). Despite a plethora of
education, engineering, and enforcement-based interventions, it is projected globally that
deaths attributable to road trauma in 2030 will be 2.4 million fatalities (World Health
Organization 2013).
While the prioritisation of vulnerable road user groups such as young drivers is vital for
their reduction in the road crash burden more generally, it is noteworthy that a breadth of social
and financial resource limitations particularly in developing nations impacts upon the nature -
and thus the efficacy - of education, engineering, and enforcement-based interventions.
Although road safety is a complex problem involving consideration of infrastructure,
engineering, economic, social and political elements which frequently balance competing
demands of mobility and safety (Canoquena & King 2015, Scott-Parker et al 2015a), the
overrepresentation of young drivers in road crashes is both consistent and persistent worldwide
(Chen et al 2012). Road trauma is the leading cause of death among young people aged 15–29
years (World Health Organization 2015), with 7,550 young passenger vehicle drivers aged 16
to 24 years old fatally injured in the U.S. in 2013 alone, and 75% of teen passenger deaths
arising from crashes in which they were driven by another young driver (Insurance Institute for
Highway Safety (IIHS) 2013). The increased crash risk experienced by all young drivers is
attributed more generally to age-related factors (e.g., neurobiological immaturity, see
(Steinberg 2005)), and the influence of psychosocial factors (Miller & Taubman - Ben-Ari
2010, Scott-Parker et al 2009), which can increase risk-taking behaviours) and driving
inexperience-related factors (e.g., underdeveloped hazard perception and reaction skills, see
(Summala 1987)). Despite a plethora of interventions, there has been no consistent progress in
this issue around the world and the need for urgent action has been emphasised (Jones et al
2015). Moreover, the majority of peer-reviewed young driver-focused research has been
undertaken within ‘WEIRD’ (western, educated, industrialised, rich, and democratic) countries
(Scott-Parker & Senserrick 2013, Scott-Parker & Senserrick 2016). Given the heterogeneity of
road traffic systems worldwide, it is essential that cultural differences are considered in order
to maximize any intervention benefits.
In Latin America, and more specifically Colombia, little attention has been paid to risky
sub-groups of drivers, such as young drivers. Although at the time of this research project, no
studies have been conducted specifically exploring young drivers’ risky behaviours in
Colombia, it is well known in Colombia that traffic laws are rarely enforced or followed,
4
4
contributing to risky driver habits (Jeon et al 2006, Norza-Céspedes et al 2014). In Colombia
in 2012 the most common causes of fatal crashes among drivers of all ages, as documented
within police reports (in descending order) were speeding (25%), drink-driving (13%), lane
excursion (12%), road rule violations (6%), and unsafe following distances (5%) (Norza-
Céspedes et al 2014). While these statistics are for drivers of all ages, it is reasonable to
conclude that young drivers may also engage in these behaviours. This supposition is supported
by numerous studies that have linked these behaviours with increased crash risk in young
drivers around the world, not only in developed countries such as Australia (Harbeck &
Glendon 2013, Scott-Parker et al 2012b) and the United States (Simons-Morton et al 2005,
Simons-Morton et al 2015) but also in middle-income countries such as Guatemala (Rodríguez-
Guzmán et al 2014), and Romania (Gheorghiu & Felonneau 2013). Most importantly, as noted
previously, the lack of research regarding young driver behaviours in developing countries
such as Colombia prevents researchers and practitioners from designing and implementing
culturally-appropriate countermeasures.
Likewise, sex differences in young drivers’ risky behaviour in Colombia are also
seldom reported in the scientific literature. Other jurisdictions have confirmed sex differences
in driving practices related directly to safety, such as greater proportions of young female
drivers using seat belts and driving with valid licences; and greater proportions of young male
drivers using alcohol, driving in the presence of peer passengers, violent behaviour while
driving, and driving whilst fatigued (Hennessy & Wiesenthal 2001, McKnight & McKnight
2003, Tsai et al 2008). In Latin America, males present consistently higher death rates due to
traffic injuries than females in every age group, including as young drivers (Diez-Roux &
Bahlla 2012). Likewise, studies in Colombia have confirmed that young male drivers are
overrepresented in drink-driving (Bonilla-Escobar et al 2015, Posada et al 2000). It can be seen
from Figure 1, which illustrates the total and sex-distribution of all driver deaths aged 14 to 26
years from 2004 to 2012 in Colombia, that the proportion of males remains consistently around
93% throughout this period. These findings further support the supposition that sex-related
factors play an important role in the road trauma of young drivers. While greater driving
exposure, as indicated by duration and/or frequency of driving, is likely to explain partially
these sex differences, there is strong evidence suggesting that the sex gap in risky driving,
including poor driving outcomes such as crashes, is narrowing worldwide (Sivak 2013).
However, there remains a dearth of information regarding risky driving behaviour differences
according to sex, which may contribute to the findings depicted in Figure 1, in Colombia.
[Insert Figure 1 here]
Study Aim
The aim of the research was to explore sex differences in self-reported risky driving
behaviours of young drivers, including the associations with crash-involvement, in a sample
of young drivers attending university in Colombia.
METHOD
Participants
5
5
Three hundred and ninety-two participants accessed the anonymous questionnaire. The age of
participants was: n = 39 aged 16-18 years, 12 males; n = 348 aged 19-24 years, 211 males; M
= 21.09, SD = 2.10, median = 21 years. The age of males and females did differ significantly
(t-test, p < .05). The mean age for females was M = 20.8, SD = 2.2 years and for males was M
= 21.4, SD = 2.0 years. Overall the period the driver’s licence had been held was M = 3.5, SD
= 2.2 years, with a range of less than 1 year to 9.5 years. The duration licensed differed
significantly according to sex (t(384)= 5.7, p < .05). The mean time with licence for females
was M = 2.8, SD = 1.9 years and for males was M = 4.1, SD = 2.2 years. Participants reported
driving M = 95.8, SD = 98.4 kilometres per week. No significant differences were found by
gender in the self-reported kilometres driven per week (t(350.1)= 1.79, p > .05).
Note that the small proportion of young drivers aged 16-18 years is not unexpected
since this age-group of drivers in Colombia is small. Inclusion criteria were (a) native Spanish
speaking and (b) have driven a motor vehicle in the past month. Similar to other low- and
middle-income countries, vehicle ownership is low in Colombia and vehicles are accessible
typically only to persons from middle- (usually limited to one vehicle per family) and high-
income families (Roque & Masoumi 2016), suggesting that young drivers are a particularly
privileged group. For this reason, participants were recruited in universities and other higher
education centres which were more likely to have students from middle- and high-income
families (Melguizo et al 2016), thus the participants were more likely to represent the typical
Colombian young driver.
Materials
The Spanish version (BYNDS-Sp) of the Behaviour of Young Novice Drivers Scale (BYNDS)
(Scott-Parker et al 2012b) was operationalised in this study. The Spanish version was obtained
and adapted to the Colombian transport system using a back-translation methodology. Six road
safety practitioners, transport engineers and survey design experts participated in the process
(additional information regarding the psychometric properties of the scale are reported in
Oviedo-Trespalacios and Scott-Parker (Under review). The BYNDS was developed
specifically to investigate the risky behaviour of young and inexperienced drivers, and the
instrument was developed in the Queensland, Australia, driving context. The BYNDS explores
transient and fixed driving violations, risky driving exposure, driving misjudgements, and
driving whilst influenced by mood (1 = never, 5 = almost always) (see Appendix Table A1 for
example items). In order to include particularities of the Colombia’s road transport system,
items MS3, MS5, TR8 and TR12 were modified from ‘right’ to ‘left’. For example, “You
travelled using the right lane” was changed to “You travelled using the left lane”. One self-
reported question assessed if they have been involved in a crash as a driver (0 = no, 1 = yes);
if yes, they were asked the number of times. Participants also responded to items asking if they
had their own vehicle and/or have access to a vehicle (no, yes); how many journeys they made
per week (free response) and the approximate number of kilometres and duration of time spent
driving on the road each week (free response).
Procedure
6
6
We implemented a cross-sectional internet-based survey in the Colombian-Caribbean Region.
The advertisement provided a brief description of the study and a hyperlink to an online
informed consent form. The Google Forms survey was disseminated using electronic mail
through university mailing lists from two universities, and required 25-35 minutes to complete.
The emails were sent from the student’s affairs offices. Ethics approval was obtained from the
Universidad del Norte’s Human Research Ethics Committee (123-3-MAR-2015) and the
University of the Sunshine Coast Human Research Ethics Committee (E/15/027).
Statistical analysis
The analysis of self-reported driving behaviours was conducted according to sex. BYNDS-Sp
responses were grouped into three categories: "Never", "Occasionally or sometimes", and
"Usually or almost always", with the frequencies and percentages recorded. Next, the original
values were averaged per item and the effect sizes of sex differences calculated per item. To
study sex differences in the BYNDS’s items, T-test with Bonferroni correction (p < 0.001) was
selected over its non-parametric counterpart following the arguments of Norman (2010) (with
correction for unequal variances where needed). In scale level analysis, internal consistency
Cronbach's alpha (α) was used to investigate the reliability of the BYNDS-Sp subscales and
composite scale. Similarly, descriptive statistics—including mean, standard deviation,
minimum, and maximum — were calculated for each subscale for the total and by sex. To
examine the associations between BYNDS-Sp subscale scores and self-reported crashes, two
stepwise binary logistic regression models were tested using these outcomes as the dependent
variable: “none” (0) and “at least one” (1); and the average score for each subscale (subscale
items summed and divided by the number of items in the subscale) as the independent
variables: transient rule violation, fixed rule violations, misjudgements, risky driving exposure,
driver mood, and risky driving behaviour (composite BYNDS-Sp). We performed stratified
analyses, with separate models for males and females, that were found to give the best fit to
the data using penalized log likelihood criteria (see Guo and Trivedi (2002)). To investigate
potential bias in maximum likelihood estimation caused by low self-reported crash
occurrences, models were re-fitted and compared using Firth's penalised likelihood procedure
(Firth 1993, Heinze et al 2013). Analyses were undertaken in Statistical Package for the Social
Sciences (SPSS) version 22 and STATA 13.1.
RESULTS
Self-reported driving behaviours, by sex
The participants’ responses to the individual BYNDS-Sp items were examined. Appendix
Table A2 summarises the participants’ self-reported driving behaviours, by sex and by
individual BYNDS-Sp item. As can be seen, for both male and female participants, one quarter
or more of the participants reported usually or almost always engaging in risky driving
exposure, including driving in the rain, on weekends, at night, during peak hour, dawn or dusk,
and carrying friends as passengers at night, and with friends who were carried as passengers
telling the young driver where to go. For male participants only, one quarter or more reported
usually or almost always speeding, including speeding in areas in which they were unlikely to
7
7
be detected, when overtaking, speeding up to 10km/hr above the posted limit, and racing out
of intersections on green lights.
T-tests were undertaken to identify any significant differences between the individual
self-reported driving behaviours of male and female participants. Appendix Table A2 also
summarises the descriptive statistics (mean, standard deviation) for each item, by sex. As can
be seen, in all instances male participants self-reported engaging in more of the risky driving
behaviours captured within the BYNDS-Sp. However, just a handful of items achieved a
statistically significant difference after Bonferroni correction for multiple tests: Transient rule
violations (TR1, TR2, TR3, TR5, TR6, TR7, TR8, and TR9), Fixed rule violations (FI6, FI8,
FI9, and FI10), Misjudgements (MS1), and Risky driving exposure (EX1, EX4, and EX5).
BYNDS-Sp subscales and composite scores, by sex
The descriptive statistics of the subscales and composite BYNDS-Sp were calculated for the
whole participant sample, and separately by sex. As can be seen in Table 1, the subscales
exhibited satisfactory internal consistency in all instances (Cronbach’s alpha ranging from 0.84
to 0.95). Similarly, the composite scale had a very high internal consistency (Cronbach’s alpha
= 0.95). Also, shown in Table 1 are the average total scores and standard deviations, number
of items, and the range for the subscales and composite by sex. Additionally, a t-test for
independent samples of males and females was conducted for the five subscales and the
composite scale, with males exhibiting consistently higher scores on all subscales and the
composite scale. Males, compared with females, reported significantly greater engagement in
risky driving as measured by the transient rule violations subscale and the BYNDS-Sp
composite scale.
[Insert Table 1 here]
Self-reported crash involvement, by sex
A total of 216 crashes were reported (57% by males) by the participants. In the total sample by
sex, approximately 38% (n = 85) of males were involved in road traffic crashes in contrast to
28% (n = 47) of females. No statistically differences by sex were found in the proportion of
drivers who self-reported at least one crash (Fisher's exact test, p = 0.052). Male participants
also reported being involved in more total road crashes (total crashes = 142, range 1 to 5, M =
1.71, mode = 1.0, median = 1.0) than female participants (total crashes = 74, range 1 to 5, M =
1.61 mode = 1.0, median = 1.0). Poisson regression was used (over dispersion was not an issue
for the data) but no statistical differences were found in the number of crashes reported by
males and females.
Given that differences in terms of age and duration licensed were found between males
and females, additional multivariable regressions were conducted focusing only on the
influence of demographic characteristics (sex, age, and duration licensed, and their
interactions) in risky behaviour. The results revealed that being male (β = 8.76; p = 0.002) and
a longer duration of licensure (β = 2.030; p = 0.001) were associated with greater self-reported
risky behaviour in young drivers (p < 0.001).
8
8
BYNDS-Sp associations with self-reported crash involvement, by sex
Separate logistic regression models for males and females were used to explore associations
between self-reported crash involvement (as the dependent variable) and the five subscales:
transient rule violations, fixed rule violations, misjudgements, risky driving exposure, and
driver mood (as independent variables). For their theoretical importance, age, time with driving
licence, and kilometres driven per week were included. A forward stepwise logistic regression
analysis using maximum likelihood estimation method was operationalised, and the
coefficients of the independent variables and the constant were evaluated using the Wald test.
The goodness of fit of the regression models was verified using the method proposed by Agresti
(2013), which considers comparisons of groups of observed and fitted values though X2 and
G2 statistics. Neither males nor females models showed lack of fit with (p > 0.1). This method
has been suggested due to the presence of continuous or nearly continuous predictors that do
not have approximate chi-squared distributions. The significance values are also shown in
Table 2. The percentage of the cases correctly classified by the model was 62.2% for males
and 72.4% for females.
[Insert Table 2 here]
The regression model that resulted from the stepwise selection, and the odds ratio estimates for
self-reported crashes, are shown in Table 2. The results revealed that for male participants (n
= 223), self-reported crashes were associated with transient rule violations and time with
licence, whilst for female participants (n = 164), self-reported crashes were associated with
transient rule violations.
Male drivers who self-reported on average recurrent involvement in transient rule
violations had higher odds of being involved in a collision. When the transient rule violation
increased by one unit, with time driving with licence remaining constant, the odds of self-
reporting a crash are approximately 48% higher (OR = 1.48, 95% CI = 1.01–2.17). The time
driving with licence increased the odds of self-reporting a collision. Specifically, for every
additional year of having a valid licence, young male drivers had 1.23 times the odds of
reporting a collision (OR = 1.23, 95% CI = 1.07–1.39). Similarly, female drivers who self-
reported on average recurrent involvement in transient rule violations had higher odds of being
involved in a collision. Results suggest that for every extra unit in the transient violations score,
the odds of self-reporting a collision increased by 75% (OR = 1.75, 95% CI = 1.08–2.85).
Associated probabilities of self-reported crashes were calculated for males and females.
The results summarised in Figure 2 shows that at any level of transient rule violations, if time
driving with licence is not considered, males are more likely to report a crash compared to
females. The impact of years driving with a licence influences the probability of reporting
crashes in young male drivers. As observed in Figure 3, it is clear that the greater exposure
afforded by having a valid driving licence increases the probability of reporting a crash at any
frequency of self-reported transient rule violations. Based on the logistic regression, associated
probabilities of self-reported crashes were calculated in the lower, median and upper quartile
(see Table 3). Poisson regression was used (over dispersion was not an issue) to study the
9
9
impact of BYNDS-Sp’s subscales upon the number of crashes self- reported by the young
driver participants. However, no significant associations were identified.
[Insert Table 3 here]
[Insert Figure 2-3 here]
DISCUSSION
The current study investigated sex differences in risky driving behaviours in young drivers in
the Colombian Caribbean Region. Risky driving behaviours were collected using the Spanish
version of the BYNDS (BYNDS-Sp). While in general, young drivers in this Colombian
sample reported high levels of involvement in risky driving behaviour, young male drivers
consistently reported greater engagement in risky driving behaviours, and transient rule
violations including speeding in particular. In the current study, 65% of participants mentioned
speeding 10–20 km/h over the speed limit at least occasionally and 62% of participants stated
they drive more than 20 km/h over the speed limit at least occasionally. As will be discussed
below, the great engagement in speeding is likely a product of factors related to enforcement
and social influences. In addition, it is important to keep in mind that the level of sophistication
of roads in Colombia is not comparable to developed countries, and the interaction of speeding
and under-developed and inadequate infrastructure contributes to more severe consequences
(Huicho et al 2012). Young male drivers consistently reported a higher frequency of speeding
(68%-84%) compared to females (55%-59%). Although the differences in speeding between
males and females are significant in this sample, the high prevalence of this behaviour is of
considerable importance because of its link with greater injury severity (Abu-Zidan & Eid
2015) and must be part of road safety initiatives for both groups.
The significantly greater engagement by males in self-reported risky behaviour is
consistent with police records in Colombia showing that males comprised nearly 80% of
detected traffic offenders (Norza-Céspedes et al 2014). The overrepresentation of males in both
archival records and our self-reported findings suggests that countermeasures should target
more precisely the riskier (male) group. While the idea of implementing sex based
interventions is intuitively sound, efficient in reaching the most vulnerable group, and appears
to be based in evidence (e.g., Scott-Parker et al (2015b) in addition to our findings), few sex-
based campaigns have been developed in the realm of road safety generally, and in young
driver road safety specifically. One young driver road safety-specific example in the Australian
context is the ‘Pinkie’ advertising campaign in the Australian state of New South Wales. This
campaign targeted speeding young male drivers through asserting the link between this risky
behaviour and small genitalia (Watsford 2008). Pinkie was a highly popular campaign, and it
appears that sex-based interventions merit further consideration particularly as recent
Australian research by Kaye et al (2015) found considerable differences in the perception of
road safety messages between young male and female drivers. These differences in perception
in turn compromise the effectiveness of road initiatives.
10
10
Approximately half of the participants in the Colombian sample reported at least
occasionally committing driving misjudgements (46%-64%). Drivers address driving task
demands using their capability, which is a function of competence and personal human factors
(Fuller et al 2008). This is particularly manifest in young drivers who may lack a fully
developed set of control or manoeuvring skills (Summala 1987). In the current study, we found
a similar proportion of misjudgements in young male and female drivers, including misjudging
stopping distances, gaps when turning across traffic, and the speed of oncoming vehicles. These
results are perhaps not unexpected since all young drivers in Colombia progress through the
same licensing (and therefore, training) program. Furthermore, serious concerns regarding
irregularities in the implementation and lack of governmental control in this process have been
raised (Martínez Beltran 2013).
There were also statistically significant differences in the self-reported engagement in
drink driving amongst young female and male drivers. In Colombia, in 2012, a total of 496
deaths were attributed to drink driving (Norza-Céspedes et al 2014), despite Colombia having
strict drink-driving laws that specify three levels of offence and sanctions. In the current study
one third of participants reported driving over the legal alcohol limit. These rates of drink
driving are well in excess of findings in Australia (14% of young drivers aged 18-20 years in
South-east Queensland, see Scott-Parker et al (2014b)) and New Zealand (23% of young
drivers aged 16-25 years, see Scott-Parker and Oviedo-Trespalacios (2017)), both jurisdictions
in which alcohol limits are strictly enforced. Previous research in Colombia has stated the need
for strong leadership and coordination to address the drink-driving problem (Castano 2012),
and our results confirm the relevance of this road safety problem.
The findings in this study supported other research that has found that young male
drivers engage more frequently in risky driving behaviours compared to young female drivers.
A caution must be given regarding this finding in the Colombian context, however, noting the
considerable differences in the sex of our participants according to their age. To date, little
research has attempted to understand the behaviours or the characteristics of young drivers in
Colombia, and future research regarding the personal characteristics of Colombian drivers
(noting that a substantially larger proportion of our younger participants were female, while a
substantially larger proportion of our older participants were male) merits consideration.
Furthermore, country-specific characteristics such as military service could influence the age
to start driving. Specifically, military service of at least one year duration is compulsory (if not
enrolled in formal education immediately after high-school) for young males aged 18 years,
while it is not compulsory for young females. Future research could further delineate any
influence of military service by exploring the young driver’s military experience and their
driving behaviour.
Regarding self-reported crashes, no significant differences in self-reported rates of
crash involvement between young male and female drivers were detected. While not able to
be compared to other young driver-focused studies in Colombia due to a dearth of information
regarding Colombian young drivers, this result is consistent with previous general population
studies in Colombia (O'Bryant 2008). Perhaps, given that males and females showed similar
risk of being involved in a crash, the considerable extent and variety of sex differences found
in a breadth of self-reported risky driving behaviours may explain the increased chance of
incurring a fatal injury for males, as noted in Figure 1. Further research is needed to confirm
11
11
this association. The current study also explored the BYNDS-Sp’ subscales’ associations with
self-reported crashes by sex. Transient rule violations were associated with self-reported
crashes by young male and female drivers. This is not surprising given that the principal cause
of road crashes in Colombia is due to speeding (a risky behaviour captured within the transient
violations subscale) (Norza-Céspedes et al 2014). However, the association between duration
of licensure and crashes is seen only in males. A potential explanation is that there were more
males who had a licence for a longer period and therefore there may have been more power to
detect an effect. In theory, this finding has considerable practical implications; particularly
when significant sex differences were found in respect to the duration of licensure in the current
sample, suggesting that multifaceted intervention is required. Specifically, future research
should explore the relationship between duration of licensure and the nature and prevalence of
risky driving behaviours in Colombia and in other jurisdictions, with previous research
suggesting that young drivers become more risky in their driving behaviour over time due to
factors such a fewer parental restrictions (see Scott-Parker et al (2012a) and McCartt et al
(2003)).
More generally, the findings regarding the considerable engagement by a notable
proportion of Colombian young driver participants in risky driving behaviours, ranging from
speeding to not wearing seatbelts, can be explained to some extent by two main theoretical
perspectives used to investigate factors influencing risky behaviour of young drivers:
Deterrence theory and Akers’ social learning theory. Firstly, according to deterrence theory,
the individuals' decision to commit offences is based on utilitarian calculations of the benefits
of performing the crime (utility) and punishments or sanctions for the crime (disutility), which
can be transitioned to enforcement from a road safety perspective (Watson & Freeman 2007).
As assessed by the WHO (2013), using a scale of 0 (none) to 10 (good), the enforcement of
road use legislation in Australia (M = 7.4; SD = 0.9) and New Zealand (M = 8.2; SD = 0.8) is
rated more highly than such enforcement in Colombia (M = 3.8; SD = 1.9). Moreover, it has
been observed in Colombia that police officers in official vehicles do not follow road rules, a
study of driving behaviour and road safety in Colombia finding that the general perception in
the country is that transit police officers are corrupt and rarely enforce road regulations (IPSOS
Napoleon Franco 2012). The influence of such differential police enforcement upon the risky
driving behaviour of males and females in Colombia remains unknown at this time.
Secondly, the highly validated Akers’ social learning theory asserts that behaviour is
learned via imitation and then influenced by differential enforcement, with the attitudes and
behaviours of the individual influenced by the attitudes and behaviours of those with whom the
individual interacts (Akers et al 1979). Applied to road safety, social and non-social sources of
reward, attitudes and behaviours are likely to increase the frequency of risky driving behaviours
(Fleiter et al 2010, Scott-Parker et al 2012c). Such mechanisms influencing road user behaviour
arguably play an important role in Colombia, a country in which one of every two drivers
reports committing traffic infractions, and two out of every three drivers disagrees with the
traffic rules (IPSOS Napoleon Franco 2012). Furthermore, 56% of Colombians believe they
are extremely unlikely to be involved in a road crash during the next year (Corporación Fondo
de Prevención Vial 2012). Similarly, ineffective legislation and low and haphazard
enforcement translate to pragmatic driving practices and greater exposure to risk factors.
Anecdotally, Colombian drivers generally do not indicate or yield to oncoming traffic, with
12
12
drivers tending to edge out into traffic and change lanes constantly. While the differential
influence of such variables upon the driving behaviour of young males and young females in
Colombia remains unknown, it seems reasonable to conclude that the road safety problem –
and the young driver road safety problem specifically – in Colombia has both bureaucratic and
psychosocial elements that require further exploration for the development of countermeasures.
Limitations and recommendations
Several important limitations need to be considered. Firstly, self-report data are frequently
subject to potential influence of social desirability or distorted memories. However, recent
research suggests that self-report data is consistent with actual police records in Australia (Ivers
et al 2009) and objective driving observations (Zhao et al 2012). Participants completed the
questionnaires anonymously and had nothing to gain by giving biased responses. Moreover,
the consistency of the findings with the findings reported in the handful of published
Colombian studies gives us confidence regarding the data validity. Notwithstanding these
assertions, further research in this area is needed including observational and/or experimental
studies. Secondly, the analyses regarding risky behaviour was focused upon measures in the
BYNDS-Sp. There are a multitude of other factors that could exert a significant effect upon
risk taking behaviour and were not included in the questionnaire, including vehicle
characteristics of the drivers (Horswill & Coster 2002), the purpose of the trip (Chliaoutakis et
al 1999, Scott-Parker et al 2015b), and driver characteristics (such as sensation seeking
propensity (Prato et al 2010) and capacity for self-regulation (Oviedo-Trespalacios et al 2017a,
Oviedo-Trespalacios et al 2017b)). Nevertheless, further studies must be conducted to validate
self-reported crashes with objective records (af Wåhlberg et al 2011). Thirdly, the present study
was undertaken to test the specific hypothesis that the BYNDS-Sp is associated with self-
reported crashes. Longitudinal studies are necessary to study causation and risk. Fourthly, the
driving context in Colombia lacks extensive scholarly research and documentation which make
it difficult to generalise the findings beyond the participant population. Fifthly, the relatively
small sample size and characteristics of the participants (all of whom attended a university)
further prevent us from generalising the findings to all Colombian young drivers. However,
most of these students are likely from middle and high income families, allowing access to a
vehicle and therefore access to driving, a phenomenon which is relatively uncommon in low
income Colombian families. Additionally, we could distinguish sex differences which differ
by a practically-small extent, such as in driving exposure, suggesting that interventions should
more generally target these risky driving behaviours, rather than simply the risky driving
behaviour of males in these circumstances. Sixthly, to discuss our data, we used Akers’ social
learning theory and deterrence theory but other theoretical models should be considered as
well. Finally, the contribution of risk factors to all crashes (including fatal crashes) remains
unknown as only information regarding involvement in non-fatal cases is reported. In addition,
while best practice learner driver training tools such as the GDE framework (Hatakka et al
2002) have been proffered, much remains unknown about the capacity to increase skills and
therefore reduce risks through learner driver training more generally. Interventions such as
graduated driving licensing (GDL) reduce risk for new drivers primarily through managing
their exposure to risk (e.g., longer training periods, driving restrictions during the earliest
13
13
period of independent licensure) (Scott-Parker et al 2014a). The understanding of injury
causation in the Colombian context could be enhanced by examining the impact of GDE-
framed training within a GDL program, with young driver behaviour, including driving
misjudgements, as an indicator of the impact of such training.
Practical implications
To the authors’ knowledge, this is the first investigation of the self-reported risky behaviours
of young drivers in Colombia. The current research is long overdue, given that we are more
than halfway through the Decade of Action for Road Safety, and that developing nations
continue to carry the greatest road safety death and injury burden worldwide. The findings
regarding the self-reported risky behaviour of young drivers in the developing nation of
Colombia echo the findings regarding the self-reported risky behaviour of young drivers in
developed nations. That is, Colombian young drivers engage in a breadth of risky driving
behaviours some of which were found to be associated with greater on-road crash risk.
Transient rule violations such as speeding by young drivers were associated with involvement
in self-reported crashes. Therefore, it appears that young drivers around the world are relatively
similar in their road safety risks, suggesting that effective interventions which have been
implemented in developed nations (such as traffic law enforcement) are likely to be effective
in the Colombian context. Notwithstanding this assertion, however, cultural considerations
unique to nations such as Colombia and which may impact upon the implementation and
evaluation of such interventions should be considered and addressed, prior to implementation.
In addition, the findings regarding the self-reported risky behaviours of young male drivers in
comparison to young female drivers in Colombia similarly is consistent with findings from
developed nations. That is, there was an increased propensity for males to take risks and to
drive in a riskier manner, compared to females. As such, it appears that sex-based interventions
merit further consideration given the robust nature of these phenomena in both developed and
developing nations.
14
14
REFERENCES
Abu-Zidan FM, Eid HO. 2015. Factors affecting injury severity of vehicle occupants following road traffic collisions. Injury 46: 136-41
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 Agresti A. 2013. Categorical Data Introduction: Distributions and Inference for Categorical Data. pp.
165-197. John Wiley & Sons, Inc. Akers RL, Krohn MD, Lanza-Kaduce L, Radosevich M. 1979. Social learning and deviant behavior: A
specific test of a general theory. American Sociological Review 44: 636-55
Bezerra BS, Kaiser IM, Battistelle RAG. 2015. Road safety-implications for sustainable development in Latin America. Latin American Journal of Management for Sustainable Development 2: 1-18
Bonilla-Escobar FJ, Herrera-López ML, Ortega-Lenis D, Medina-Murillo JJ, Fandiño-Losada A, et al. 2015. Driving under the influence of alcohol in Cali, Colombia: prevalence and consumption patterns,
2013. International journal of injury control and safety promotion 23: 1-10 Canoquena J, King M. 2015. Adaptive nature of coordinated road traffic injury prevention responses in
some OECD countries. Journal of Transport Geography 46: 81-88
Castano R. 2012. The drink driving situation in Colombia. Traffic injury prevention 13: 120-25 Chen H, Jan S, Boufous S, Martiniuk A, Ivers R, et al. 2012. Variations in car crash-related hospitalization
costs amongst young adults in New South Wales, Australia. Injury 43: 1593-99 Chliaoutakis JE, Darviri C, Demakakos PT. 1999. The impact of young drivers’ lifestyle on their road
traffic accident risk in greater Athens area. Accident Analysis & Prevention 31: 771-80
Corporación Fondo de Prevención Vial C. 2012. Caracterización de la prestación del servicio de transporte en motocicleta. In Boletin Prensa No. 153
Diez-Roux E, Bahlla K. 2012. Cost of Traffic Injuries in Latin America, Inter-American Development Bank, Washington
Firth D. 1993. Bias reduction of maximum likelihood estimates. Biometrika 80: 27-38 Fleiter JJ, Lennon A, Watson B. 2010. How do other people influence your driving speed? Exploring the
‘who’ and the ‘how’ of social influences on speeding from a qualitative perspective.
Transportation Research Part F: Traffic Psychology and Behaviour 13: 49-62 Fuller R, McHugh C, Pender S. 2008. Task difficulty and risk in the determination of driver behaviour.
Revue Européenne de Psychologie Appliquée/European Review of Applied Psychology 58: 13-21
Gheorghiu A, Felonneau ML. 2013. Young drivers and self-reported driving under the influence of
alcohol: A Comparison between Romanian and French drivers. Analele Ştiinţifice ale Universităţii" Alexandru Ioan Cuza" din Iaşi. Psihologie 22: 19-40
Guo JQ, Trivedi PK. 2002. Flexible Parametric Models for Long‐tailed Patent Count Distributions. Oxford Bulletin of Economics and Statistics 64: 63-82
Harbeck EL, Glendon AI. 2013. How reinforcement sensitivity and perceived risk influence young drivers’
reported engagement in risky driving behaviors. Accident Analysis & Prevention 54: 73-80 Hatakka M, Keskinen E, Gregersen NP, Glad A, Hernetkoski K. 2002. From control of the vehicle to
personal self-control; broadening the perspectives to driver education. Transportation Research Part F: Traffic Psychology and Behaviour 5: 201-15
Heinze G, Ploner M, Dunkler D, Southworth H. 2013. logistf: Firth’s bias reduced logistic regression. In
R package version Hennessy DA, Wiesenthal DL. 2001. Gender, driver aggression, and driver violence: An applied
evaluation. Sex Roles 44: 661-76 Horswill MS, Coster ME. 2002. The effect of vehicle characteristics on drivers' risk-taking behaviour.
Ergonomics 45: 85-104
Huicho L, Adam T, Rosales E, Paca-Palao A, López L, et al. 2012. Evaluation of interventions on road traffic injuries in Peru: a qualitative approach. BMC public health 12: 1-13
Insurance Institute for Highway Safety (IIHS). 2013. Teenagers: Driving Carries Extra Risk for Them. IPSOS Napoleon Franco I. 2012. Segundo estudio de Comportamientos y Seguridad Vial. Estudio No. -
JB:11-0.16902-03
Ivers R, Senserrick T, Boufous S, Stevenson M, Chen H-Y, et al. 2009. Novice drivers' risky driving behavior, risk perception, and crash risk: findings from the DRIVE study. American journal of public health 99: 1638
15
15
Jeon CM, Amekudzi AA, Vanegas J. 2006. Transportation system sustainability issues in high-, middle-, and low-income economies: Case studies from Georgia (US), South Korea, Colombia, and
Ghana. Journal of urban planning and development 123: 172-86 Jones S, McKenna F, Stradling S, Christie N, Mullarkey T, et al. 2015. Crashes involving young drivers.
BMJ 350: h659
Kaye S-A, White MJ, Lewis IM. 2015. Young drivers' perceptions of road safety messages and a high performance vehicle advertisement: a qualitative exploration. Journal of the Australasian College of Road Safety 26: 14-25
Martínez Beltran DC. 2013. Influencia de los centros de enseñanza automovilisticos en el conocimiento de los conductores en las vias en cuanto a normatización. B.A. Industrial Engineer thesis. Escuela Colombiana de Ingeniería Julio Garavito, Bogota
McCartt AT, Shabanova VI, Leaf WA. 2003. Driving experience, crashes and traffic citations of teenage
beginning drivers. Accident Analysis & Prevention 35: 311-20 McKnight AJ, McKnight AS. 2003. Young novice drivers: careless or clueless? Accident Analysis &
Prevention 35: 921-25 Melguizo T, Sanchez F, Velasco T. 2016. Credit for Low-Income Students and Access to Higher
Education in Colombia: A Regression Discontinuity Approach. World Development 80: 61-77
Miller G, Taubman - Ben-Ari O. 2010. Driving styles among young novice drivers—The contribution of parental driving styles and personal characteristics. Accident Analysis & Prevention 42: 558-70
Norman G. 2010. Likert scales, levels of measurement and the “laws” of statistics. Advances in Health Sciences Education 15: 625-32
Norza-Céspedes EH, Useche-Hernández SA, Rodríguez JM, Granados-León EL, Hernández MR. 2014. Componentes descriptivos y explicativos de la accidentalidad vial en Colombia: incidencia del
factor humano. Rev. crim 56: 31-31
O'Bryant AL. 2008. Factors Associated with Traffic Crashes in Pasto, Colombia: 2005-2006. Master of Public Health (MPH) thesis. Georgia State University
Oviedo-Trespalacios O, Haque MM, King M, Washington S. 2017a. Effects of road infrastructure and traffic complexity in speed adaptation behaviour of distracted drivers. Accident Analysis & Prevention 101: 67-77
Oviedo-Trespalacios O, Haque MM, King M, Washington S. 2017b. Self-regulation of driving speed among distracted drivers: An application of driver behavioral adaptation theory. Traffic Injury Prevention: 1-7
Oviedo-Trespalacios O, Scott-Parker B. Under review. Transcultural validation and reliability of the
Spanish version of the Behaviour of Young Novice Drivers Scale (BYNDS) in a Colombian young
driver population Transportation Research Part F: Traffic Psychology and Behaviour Posada J, Ben-Michael E, Herman A, Kahan E, Richter E. 2000. Death and injury from motor vehicle
crashes in Colombia. Revista Panamericana de Salud Pública 7: 88-91 Prato CG, Toledo T, Lotan T, Taubman - Ben-Ari O. 2010. Modeling the behavior of novice young drivers
during the first year after licensure. Accident Analysis & Prevention 42: 480-86 Rodríguez-Guzmán S, Jiménez-Mejías E, Martínez-Ruiz V, Lupiáñez-Tapia F, Lardelli-Claret P, Jiménez-
Moleón JJ. 2014. Mobility, traffic accidents, and associated factors among Guatemalan
university students. Cadernos de Saúde Pública 30: 735-45 Roque D, Masoumi HE. 2016. An Analysis of Car Ownership in Latin American Cities: a Perspective for
Future Research. Periodica Polytechnica. Transportation Engineering 44: 5-12 Scott-Parker B, Goode N, Salmon P. 2015a. The driver, the road, the rules… and the rest? A systems-
based approach to young driver road safety. Accident Analysis & Prevention 74: 297-305
Scott-Parker B, King MJ, Watson B. 2015b. The psychosocial purpose of driving and its relationship with the risky driving behaviour of young novice drivers. Transportation research part F: traffic psychology and behaviour 33: 16-26
Scott-Parker B, Oviedo-Trespalacios O. 2017. Young driver risky behaviour and predictors of crash risk
in Australia, New Zealand and Colombia: Same but different? Accident Analysis & Prevention 99, Part A: 30-38
Scott-Parker B, Senserrick T. 2013. Methodology and broader implications of young driver research published in Traffic Injury Prevention in the past five years. Presented at Australasian College of Road Safety Conference, Adelaide, Australia
Scott-Parker B, Senserrick T. 2016. A call to improve sampling methodology and reporting in young novice driver research. Injury Prevention 23: 8-9
16
16
Scott-Parker B, Senserrick T, Simons-Morton B, Jones C. 2014a. Higher-order instruction by professional driving instructors: a naturalistic pilot study. Presented at Australasian Road Safety Research
Policing Education Conference, 2014, Melbourne, Victoria, Australia, Scott-Parker B, Watson B, King M, Hyde M. 2012a. Young, inexperienced, and on the road: do novice
drivers comply with road rules? Transportation Research Record: Journal of the Transportation Research Board: 98-106
Scott-Parker B, Watson B, King MJ. 2009. Understanding the psychosocial factors influencing the risky
behaviour of young drivers. Transportation Research Part F: Traffic Psychology and Behaviour 12: 470-82
Scott-Parker B, Watson B, King MJ, Hyde MK. 2012b. Confirmatory factor analysis of the Behaviour of Young Novice Drivers Scale (BYNDS). Accident Analysis & Prevention 49: 385-91
Scott-Parker B, Watson B, King MJ, Hyde MK. 2012c. “They’re lunatics on the road”: Exploring the
normative influences of parents, friends, and police on young novices’ risky driving decisions. Safety science 50: 1917-28
Scott-Parker B, Watson B, King MJ, Hyde MK. 2014b. "i drove after drinking alcohol" and other risky driving behaviours reported by young novice drivers. Accident Analysis and Prevention 70: 65-
73
Sharma B. 2008. Road traffic injuries: a major global public health crisis. Public health 122: 1399-406 Simons-Morton B, Lerner N, Singer J. 2005. The observed effects of teenage passengers on the risky
driving behavior of teenage drivers. Accident Analysis & Prevention 37: 973-82 Simons-Morton BG, Klauer SG, Ouimet MC, Guo F, Albert PS, et al. 2015. Naturalistic teenage driving
study: Findings and lessons learned. Journal of Safety Research 54: 29-44 Sivak M. 2013. Female Drivers in the United States, 1963–2010: From a Minority to a Majority? Traffic
injury prevention 14: 259-60
Steinberg L. 2005. Cognitive and affective development in adolescence. Trends in Cognitive Sciences 9: 69-74
Summala H. 1987. Young driver accidents: Risk taking or failure of skills? Alcohol, Drugs & Driving 3: 3-4
Tsai VW, Anderson CL, Vaca FE. 2008. Young female drivers in fatal crashes: recent trends, 1995–
2004. Traffic injury prevention 9: 65-69 Watsford R. 2008. The success of the ‘Pinkie’campaign-Speeding. No one thinks big of you: A new
approach to road safety marketing. . Presented at 2008 Joint ACRS-Travelsafe National Conference., Brisbane, Australia
Watson B, Freeman J. 2007. Perceptions and experiences of random breath testing in Queensland and
the self-reported deterrent impact on drunk driving. Traffic Injury Prevention 8: 11-19 World Health Organization. 2013. Global status report on road safety 2013: supporting a decade of
action, World Health Organization World Health Organization. 2015. Global status report on road safety 2015
Zhao N, Mehler B, Reimer B, D’Ambrosio LA, Mehler A, Coughlin JF. 2012. An investigation of the relationship between the driving behavior questionnaire and objective measures of highway
driving behavior. Transportation Research Part F: Traffic Psychology and Behaviour 15: 676-85
17
17
Figure 1. Trends of drivers’ deaths aged 14-26 years by sex – Colombia 2004-2012 Data from: National Institute of Legal Medicine and Forensic Sciences, 2015.
0
10
20
30
40
50
60
70
80
90
100
2004 2005 2006 2007 2008 2009 2010 2011 2012
Fa
tali
ties
Females Males Total
18
18
Figure 2. Predicted probabilities for self-reported collisions in relation to the average score of
the BYNDS-Sp’s Transient Violations subscale, by sex
0
0.1
0.2
0.3
0.4
0.5
0.6
1 1.5 2 2.5 3 3.5 4
Pre
dic
ted
pro
ba
bil
ity
of
self
-rep
ort
ed
cra
shes
BYNDS-Sp Transient violations subscale average score
Females Males
19
19
Figure 3. Males predicted probabilities for self-reported collisions in relation to the average
score of the BYNDS-Sp’s Transient Violations subscale, by time driving with licence
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Pre
dic
ted
pro
ba
bil
ity
of
self
-rep
ort
ed c
oll
isio
ns
ma
les
BYNDS-Sp Transient Violations subscale average score
Duration of licensure ≥5 years
Duration of licensure <5 years
20
20
Table 1. BYNDS-Sp scale and subscale reliability, and mean (SD), minimum and maximum, by sex±
Alpha
Cronbach
No. of
items
Total Males Females T-test
(significance) Factors SD Min. Max. SD Min. Max. SD Min. Max.
Transient rule
violation 0.89 13 29.03 9.79 13 65 31.16 9.72 13 65 26.17 9.15 13 50 p< 0.001
Fixed rule
violations 0.85 10 17.82 6.85 10 44 18.84 7.18 10 44 16.43 6.13 10 37 p= 0.001
Misjudgements 0.88 9 16.51 6.10 9 41 17.03 6.51 9 41 15.80 5.44 9 31 p= 0.074
Risky driving
exposure 0.88 9 26.59 7.75 9 43 27.70 7.14 9 43 25.10 8.29 9 42 p= 0.002
Driver mood 0.84 3 6.60 2.95 3 15 6.84 2.98 3 15 6.27 2.88 3 15 p= 0.074
Risky driving
behaviour
(BYNDS-Sp)
0.95 44 96.54 27.44 44 192 101.56 27.31 44 192 89.77 26.20 44 153 p< 0.001
Significant differences after Bonferroni correction, p < 0.01. ± Items in the scales were summed, and the summed scores were averaged across participants.
21
21
Table 2. Logistic regression analyses of history of at least one self-reported crash using BYNDS-Sp subscales, by sex
Males (n=223) Females (n=164)
β Std.
error p_value
Exp
(β)
95% C.I.for Exp
(β) β Std.
error p_value
Exp
(β)
95% C.I.for Exp
(β)
Lower Upper Lower Upper
Step 6 (Final)
Transient rule
violation 0.40 0.20 0.04 1.48 1.01 2.17 0.56 0.25 0.02 1.75 1.08 2.85
Time driving
with licence 0.20 0.07 < 0.01 1.23 1.07 1.39 - - - - - -
Constant -2.29 0.57 < 0.01 0.10 - - -2.12 0.56 < 0.01 0.12 - -
22
22
Table 3. Predicted probabilities of self-reported crash at the quartiles for the three predictor variables
Males Females
Time with driving licence† Transient rule violation ɸ Transient rule violation
Percentile (Value) Predicted
probability Percentile (Value)
Predicted
probability Percentile (Value) Predicted probability
25% (2.33) 0.30 25% (1.85) 0.32 25% (1.46) 0.21
50% (4.00) 0.37 50% (2.38) 0.37 50% (1.92) 0.26
75% (5.33) 0.44 75% (2.92) 0.42 75% (2.38) 0.31 † The average transient rule violation, M = 2.41, was included in this calculation. ɸ Average time with driving licence, M = 4.05, was included in this calculation.
23
23
Appendix
Table A1. BYNDS-Sp’s subscales definitions and examples
Subscale Definition Example
Transient rule
violations
Risky driving behaviours that are
able to be performed multiple
times during the journey
Driving 10
km/hr over the
speed limit
Fixed rule
violations
Risky driving behaviours that are
more stable in nature across the
journey
Driving after
drinking
alcohol
Misjudgements Common driving errors that place
young novice drivers at greater
risk
Misjudging the
speed of an
oncoming
vehicle
Risky driving
exposure
Driving at times that place young
novice drivers at greater risk
Driving at night
Driver mood Risky driving behaviours in
response to emotions experienced
by the young novice driver
Driving faster
if in a bad
mood
24
24
Table A2. Proportions and comparison of young drivers reporting risky driving behaviours, by sex
Items
Male
Female
Males
Females
Cohen
’s D t_test
Never (%)
Occasionally
or
Sometimes
(%)
Usually
or
always
(%)
Never
(%)
Occasionally
or
Sometimes
(%)
Usually
or
always
(%)
M SD M SD
TR1. You drove over
the speed limit in areas
where it was unlikely
there was a radar or
speed camera
24% 50% 26%
46% 41% 14%
2.59 1.28
1.97 1.13 0.51 t(378)= 5.04, p< 0.001
TR2. You went 10–20
km/h over the speed
limit (e.g. 72 km/h in a
60 km/h zone, 112
km/h in a 100 km/h
zone)
29% 48% 24%
44% 47% 9%
2.42 1.24
1.94 1.07 0.42 t(381)= 4.1, p< 0.001
TR3. You deliberately
sped when overtaking
16% 51% 33%
29% 53% 18%
2.88 1.29
2.34 1.19 0.43 t(390)= 4.21, p< 0.001
TR4. You sped at night
on roads that were not
well lit
30% 57% 13%
41% 50% 9%
2.23 1.10
1.97 1.04 0.24 t(390)= 2.33, p= 0.02
TR5. You went up to
10 km/h over the speed
limit (e.g. 65 km/h in a
60 km/h zone, 105
km/h in a 100 km/h
zone)
23% 52% 25%
41% 51% 8%
2.58 1.23
1.95 1.03 0.55 t(384)= 5.48, p< 0.001
TR6. You went more
than 20 km/h over the
speed limit (e.g. 60
km/h in a 40 km/h
zone, 120 km/h in a
100 km/h zone)
32% 48% 19%
45% 45% 10%
2.29 1.19
1.90 1.06 0.35 t(376)= 3.41, p< 0.001
TR7. You raced out of
an intersection when
the light went green
11% 55% 34%
25% 54% 20%
2.93 1.24
2.45 1.28 0.38 t(390)= 3.73, p< 0.001
TR8. You travelled in
the left lane on multi-
lane highways
34% 52% 14%
50% 47% 3%
2.15 1.10
1.76 0.89 0.39 t(387)= 3.88, p< 0.001
25
25
TR9. You sped up
when the lights went
yellow
18% 59% 24%
28% 61% 11%
2.59 1.14
2.22 1.04 0.34 t(374)= 3.39, p< 0.001
TR10. You went too
fast around a corner
33% 57% 10%
41% 56% 4%
2.10 1.00
1.86 0.88 0.26 t(390)= 2.48, p= 0.01
TR11. You did an
illegal U-turn
44% 45% 10%
54% 41% 5%
1.96 1.08
1.68 0.90 0.28 t(390)= 2.69, p= 0.01
TR12. You overtook
someone on the right
42% 48% 9%
50% 42% 8%
2.02 1.05
1.78 1.01 0.23 t(390)= 2.26, p= 0.02
TR13. You spoke on a
mobile that you held in
your hands
20% 65% 15%
25% 56% 19%
2.42 1.08
2.35 1.15 0.07 t(390)= 0.65, p= 0.51
FI1. Your passengers
did not wear seatbelts
35% 51% 14%
45% 48% 7%
2.18 1.15
1.89 1.04 0.26 t(390)= 2.53, p= 0.01
FI2. You drove after
taking an illicit drug
such as marijuana or
ecstasy
78% 17% 4%
81% 17% 2%
1.38 0.85
1.30 0.72 0.10 t(390)= 0.96, p= 0.34
FI3. You carried more
passengers than could
legally fit in your car
45% 43% 12%
47% 45% 8%
1.98 1.11
1.84 1.00 0.13 t(390)= 1.26, p= 0.21
FI4. You did not
always wear your
seatbelt
55% 29% 16%
63% 28% 8%
1.97 1.31
1.71 1.14 0.21 t(380)= 2.11, p= 0.04
FI5. You drove without
a valid licence as
because you had not
applied for one yet or it
had been suspended
68% 21% 11%
72% 22% 6%
1.62 1.10
1.51 0.97 0.11 t(390)= 1.05, p= 0.29
FI6. You did not wear a
seatbelt if it was only
for a short trip
54% 33% 13%
63% 31% 6%
1.89 1.19
1.58 0.95 0.29 t(388)= 2.85, p< 0.001
FI7. If there was no red
light camera, you drove
through intersections
on a red light
52% 40% 8%
61% 35% 4%
1.82 1.03
1.61 0.92 0.21 t(390)= 2.06, p= 0.04
FI8. You carried more
passengers than there
were seatbelts for in
your car
39% 47% 14%
51% 42% 7%
2.09 1.12
1.78 0.97 0.30 t(390)= 2.9, p< 0.001
FI9. You drove when
you thought you may
63% 30% 7%
75% 24% 1%
1.61 0.97
1.32 0.62 0.36 t(383)= 3.54, p< 0.001
26
26
have been over the
legal alcohol limit
FI10. You drove a
high-powered vehicle
33% 48% 19%
49% 43% 8%
2.31 1.20
1.89 1.06 0.37 t(379)= 3.67, p< 0.001
MS1. You misjudged
the speed when you
were exiting a main
road
45% 47% 8%
50% 47% 3%
1.94 1.05
1.68 0.82 0.27 t(389)= 2.69, p< 0.001
MS2. You misjudged
the speed of an
oncoming vehicle
39% 54% 7%
40% 55% 5%
1.93 0.94
1.86 0.89 0.07 t(390)= 0.7, p= 0.48
MS3. You misjudged
the gap when you were
turning left
51% 42% 7%
51% 46% 3%
1.75 0.97
1.68 0.82 0.08 t(390)= 0.73, p= 0.46
MS4. You misjudged
the stopping distance
you needed
45% 48% 7%
46% 52% 2%
1.86 0.95
1.72 0.79 0.16 t(384)= 1.58, p= 0.11
MS5. You turned left
into the path of another
vehicle
50% 42% 8%
59% 37% 4%
1.81 1.00
1.63 0.93 0.18 t(390)= 1.75, p= 0.08
MS6. You misjudged
the gap when you were
overtaking another
vehicle
47% 48% 5%
53% 40% 7%
1.83 0.95
1.72 0.96 0.11 t(390)= 1.11, p= 0.27
MS7. You missed your
exit or turn
39% 54% 7%
42% 54% 4%
1.96 0.98
1.84 0.87 0.13 t(390)= 1.27, p= 0.20
MS8. You entered the
road in front of another
vehicle
48% 46% 5%
54% 40% 6%
1.79 0.93
1.73 0.95 0.06 t(390)= 0.63, p= 0.53
MS9. You didn’t
always indicate when
you were changing
lanes
33% 55% 12%
41% 49% 10%
2.16 1.12
1.93 1.04 0.21 t(390)= 2.07, p= 0.04
EX1. You drove on the
weekend
4% 28% 68%
13% 38% 49%
3.83 1.22
3.29 1.42 0.41 t(325)= 3.94, p< 0.001
EX2. You drove in the
rain
6% 60% 34%
15% 56% 29%
3.09 1.04
2.80 1.22 0.26 t(322)= 2.5, p= 0.01
EX3. You drove at
peak times in the
morning and afternoon
20% 43% 37%
20% 49% 31%
2.89 1.30
2.75 1.31 0.11 t(390)= 1.05, p= 0.29
EX4. You drove at
night
4% 40% 56%
13% 45% 42%
3.58 1.12
3.19 1.36 0.31 t(316)= 2.99, p< 0.001
27
27
EX5. You drove at
dusk or dawn
5% 49% 46%
13% 51% 35%
3.27 1.07
2.95 1.19 0.28 t(390)= 2.79, p< 0.001
EX6. You carried your
friends as passengers at
night
11% 50% 39%
18% 50% 32%
3.04 1.22
2.78 1.24 0.21 t(390)= 2.07, p= 0.04
EX7. You drove when
you knew you were
tired
19% 60% 21%
24% 66% 10%
2.56 1.12
2.26 0.98 0.28 t(379)= 2.79, p= 0.01
EX8. Your car was full
of your friends as
passengers
12% 60% 28%
25% 53% 22%
2.80 1.15
2.50 1.23 0.25 t(390)= 2.46, p= 0.01
EX9. You went for a
drive with your mates
giving you directions to
where they wanted to
go
20% 57% 23%
24% 51% 25%
2.65 1.21
2.59 1.25 0.05 t(390)= 0.49, p= 0.62
DM1. Your driving
was affected by
negative emotions like
anger or frustration
33% 56% 12%
37% 50% 13%
2.17 1.06
2.11 1.10 0.06 t(390)= 0.59, p= 0.55
DM2. You allowed
your driving style to be
influenced by what
mood you were in
30% 54% 16%
41% 49% 10%
2.30 1.16
2.00 1.05 0.27 t(374)= 2.69, p= 0.01
DM3. You drove faster
if you were in a bad
mood
28% 53% 19%
37% 47% 16%
2.36 1.17
2.16 1.18 0.17 t(390)= 1.69, p= 0.09
TR = Transient violations; FR = Fixed violations; MS = Misjudgements; EX = Exposure; DM = Driver mood Significant differences after Bonferroni correction, p < 0.01.
28
28