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 qut.copyright@qut.edu.au

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: ooviedot@gmail.com, oscar.oviedotrespalacios@qut.edu.co

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