• Vol 9, No 3 (2018)
  • Civil Engineering

The Relationship between Self-reported Traffic Crashes and Driver Behavior in the Road Transportation of Goods and Freight in Bali

Dewa Made Priyantha Wedagama, Darren Wishart


Cite this article as:
Wedagama, D.M.P., Wishart, D., 2018. The Relationship between Self-reported Traffic Crashes and Driver Behavior in the Road Transportation of Goods and Freight in Bali . International Journal of Technology. Volume 9(3), pp.558-567
77
Downloads
Dewa Made Priyantha Wedagama - Department of Civil Engineering, Udayana University
-
Darren Wishart Centre for Accident Research and Road Safety (CARRS-Q) Queensland University of Technology
Email to Corresponding Author

Abstract
image

Road safety stakeholders in Bali have highlighted the need to investigate and better understand road user behavior as a means of reducing the high proportion of crashes in goods and freight distribution sector. This study aimed to analyze the relationship between self-reported driver behaviors and traffic crashes reported by drivers transporting goods and freight in Denpasar, Bali. A driving safety survey was distributed to 350 goods and freight transport drivers to obtaine a range of demographic information, self-reported crash and offence data, along with psychosocial factors data associated with driving safety. Factor analysis identified a four factor solution with distinctions between di?erent driving practices within the sample consisting of aggresive driving behaviors, driving errors, traffic code violations and normlessness. The results of this study demonstrated that normlessness consisting of behaviors such as taking a slight risk when overtaking and ignoring speed limits were the most common form of unsafe behaviors reported by the goods vehicle transport drivers. The study indicated that speeding behaviors influenced more self-reported driving errors, aggresive behaviors and traffic rule violations and traffic crashes. In addition, drink driving was found to be positively related to traffic crashes. The implications of these results are discussed.

Driver behaviors, Goods vehicle, Road safety, Traffic crashes

Introduction

Annually, there are estimated to be more than 38,000 road traffic fatalities in Indonesia (The World Health Organization, 2015). This number includes both passengers and goods transport. Goods transport is the conveyance of materials or products from one place to another using different modes and networks, and is usually expressed in ton-kilometers, while passenger transport is the carrying of people and is expressed in passenger-kilometers.

As Indonesia is a major distribution economy in the South Pacific region, goods distribution and transport are becoming an increasing focus of road safety stakeholders. Within the goods distribution and transport sector in Indonesia, overloading has been identified as a major contributor to traffic crashes (Indonesian National Police, 2014). Indonesian crash data over a five year period (2009-2013) showed that this is a highly relevant issue, with the number of violations by freight transport firms reported as representing 8.5% of the total number of traffic violations.  In  addition, there  was  an  average  of  21.2%  of  freight traffic crashes of the total number of traffic crashes per year over the 5 year period from 2009 to 2013 (Indonesian National Police, 2014).

While much of the responsibility for driving safely ultimately rests with the driver of a vehicle (Zuraida et al., 2017), the load carrying capacity (overloading) of a vehicle can be influenced by factors that are both individual and also organizational in nature. For example, organizational factors could include circumstances whereby an organization is endeavoring to transfer freight and goods with the least amount of resources (eg. to save time or use of vehicles), which may contribute to the overloading of vehicles. In addition, an organizational culture may exist that encourages drivers to overload vehicles, particularly in developing countries, where the enforcement of legal load requirements may not be stringently enforced by the authorities. Furthermore, in developing countries such as Indonesia, utilisation of, or a lack of, resources such as vehicles fit for purpose, may be one of the main contributing factors to freight and goods distribution vehicle overloading. There also exists a multitude of individual factors that can influence freight and goods distribution vehicle transport safety. Despite the prevalence of crashes, little attention has been focused on the behaviors of goods vehicle drivers in Indonesia.

Meanwhile, goods and freight distribution vehicle transport movements in Bali play an important role in Indonesia, as they connect lines of freight between Java and the west, and the Nusa Tenggara islands to the east. In comparison with river and ferry services, the proportion of road transportation in Bali from 2006 to 2011 was 91.25% of total goods and freight transport. In other words, road transportation makes a significant contribution towards goods and freight transport in Bali. However, it should also be noted that within Bali various types of vehicles are permitted to deliver freight and goods, including vans and small sedans, light tray back trucks, and utility and heavy truck vehicles. In 2015, there were a total of 113,937 heavy transport vehicles registered in Bali, of which 78,720 were trucks and 35,217 were pick up vehicles. In addition, in Denpasar alone, there were a total of 45,115 heavy transport vehicles registered, consisting of 32,720 trucks and 12,395 pick-up vehicles (Statistics of Bali Province, 2015). Consequently, the goods and freight distribution transport sector comprises a large proportion of the registered vehicles in Denpasar, accounting for approximately 40% of all vehicles registered in Bali.

As a means of addressing the trauma associated with road crashes, particularly in developing countries, there is a need to have a better understanding of the factors associated with unsafe and risky driving behaviors, particularly in heavily congested and urban roads along typical transport and goods distribution corridors. Previous research on driving safety within the work sector in Indonesia is scarce; however, research in other countries has utilized a variety of self-reporting measures to better understand the influential factors within the work driving setting. For example, research in the Australian setting has identified several contributing factors to fleet vehicle crashes, including age, disobeying traffic rules, alcohol and drugs, speeding, inexperience, inattention, fatigue, negligence and road conditions (Wishart & Davey, 2004). Other research has also identified higher crash rates and less engagement in vehicle checking practices in work vehicles, in comparison to personal vehicles (Newnam et al., 2002). Similarly, fleet drivers have been shown to have the tendency to engage in speeding behaviors and considered speeding to be acceptable, in contrast to other factors such as drink driving, tailgating or risky overtaking manoeuvres (Davey et al., 2006; Freeman et al., 2008). Other factors such as attitude, behavior, knowledge and hazard perception have also been shown to be highly correlated with the self-reported collisions of fleet drivers.

More specifically, attitude and behavior scores, distance travelled, driver age and personality have been shown to have a statistically signi?cant relationship with the involvement in collisions of fleet drivers (Darby et al., 2009). Within the heavy vehicle freight distribution sector, law abiding driving behavior in trucks has been found to be more related to attitudes, subjective norms and intentions than perceived behavioral control (Poulter et al., 2008). However, other research within this area has identified that driver and company perceptions differ in regards to fatigue and fatigue management, particularly in relation to the causes and level of fatigue, and the strategies that should be used to manage it (Arnold et al., 1997). This study therefore aims to investigate the effect of individual factors on goods and freight distribution vehicle transport safety, using Denpasar, the capital city of Bali Province, Indonesia, as the case study area. More specifically, the study investigates the relationship between self-reported traffic crashes and the behaviors of goods and freight distribution vehicle transport drivers, highlighting the influence of driver behavior and other contributing factors to traffic crashes in the road transportation of goods and freight. It is therefore essential to place the behavioral factors within a relevant theoretical structure to comprehend, foresee and deal with goods and freight-related driver safety.




Experimental Methods

2.1.   Participants and Procedure

The researchers approached goods and freight distribution vehicle transport drivers at various organizational distribution locations, such as shops and wholesale/manufacturing premises, in Denpasar. The participants were advised of the aim of the survey and their cooperation to take part was requested. A total of 350 individuals participated in the study, who were all male drivers of vehicles used to transport goods and freight. The data indicated that on average respondents were around 35 years of age (varying between 20 and 63 years old). The type of vehicles reported as being used for freight and goods distribution by the participants consisted of light vehicles such as sedans, hatchbacks, station wagons and utility vehicles (21.4%), four-wheel drives (26%), trucks (36.3%) and others (16.3%).

The data indicated that the majority of participants had not been involved in a traffic crash in the previous 12 months, with 70.6% indicating having had no non-work-related crashes, and 85.4% having had no work-related crashes. Of the 103 (29.4%) drivers who reported having been involved in a work crash in the previous 12 months, 77 stated that they had had only one crash. A total of 51 (14.6%) drivers reported that they had been involved in a non-work crash in the previous 12 months, with 26 of these stating that only one crash had occurred. In brief, a total sample of 350 respondents was used for the analysis. Of these, 103 had been involved in traffic crashes (code = 1), while the remainder had not (code = 0).

Self-reporting of traffic offences indicated that 70.6% of the participants had not committed a traffic offence during work hours, and 74.9% had not done so outside of work hours. The participants also reported that 21.4% of them did most of their driving on urban roads, while 26% travelled mostly on a mixture of both urban and rural roads. A total of 36.3% of the sample drove only on rural roads, and 16.3% mostly off road.

With regard to driving experience and exposure, on average the participants had been driving a work vehicle for about 14 years (varying between 2 and 40 years). The highest proportion of the sample, 30%, drove between 1 and 10 hours per week, with 27.7% estimating that they drove between 30,000 and 40,000 km per year.

2.2.   Materials

A work driving safety questionnaire previously utilized within the Australian work driving setting (Freeman et al., 2008) was used in this study, adapted and translated into the Indonesian language. The work questionnaire contained modified versions of the Driver Behavior Questionnaire (DBQ) (Reason et al., 1990); the Driver Attitude Questionnaire (DAQ) (Parker et al., 1996); the Safety Climate Questionnaire (SCQ-MD) (Glendon & Stanton, 2000); the Thrill Seeking Subscale from the Driver Stress Inventory (Matthews et al., 1997); and several contemporary issues specifically connected with work driving safety (e.g. fatigue).

Behavior and attitude variables were derived from two types of question. The first type consisted of a total of 44 questions related to driver experiences (expressed by 1 = never to 7 = always) when driving for work over the previous 6 months. The second type comprised a total of 20 questions on perceptions of and feelings towards work driving, and the readiness to change risky driving behavior (expressed from 1 = strongly disagree to 7 = strongly agree). All these variables were subjected to Principal Component Analysis (PCA). With regard to the results, three behavior, attitude and thrill seeking variables were specified in subscales 4-6 in Table 1.

 

Table 1 Questionnaire items

No.

Items

1

Demographic information (age, driving license, gender)

2

Exposure (hours per week driving, km travelled per year)

3

Crashes and offences (in the last 12 months)

4

Driver Behavior items (DBQ plus extra contemporary items)

5

Driver Attitude Questionnaire

6

Readiness to change (derived from Prochaska & DiClemente, 1984 )

 

Table 1 shows the questionnaire items, which comprised six sections consisting of a total of 73 items. The study employed the self-reporting method because it specifies similar participants with various risky behaviors and is documented more suitably (Rhodes & Pivik, 2011). Prejudice towards reacting in a socially desirable way was also discovered to be fairly insignificant in the driver behavior answers (Lajunen & Summala, 2003). Self-report measures have also been shown to identify “at risk” drivers in professional drivers’ involvement with aberrant driving behaviors in fleet based settings (Freeman et al., 2009).

2.3.   Data Analysis Techniques

Factor analysis was used to examine the behavior and perception variables included in the model construction. Principal Component Analysis (PCA) was employed for each behavior, attitude and thrill seeking variable (subscales 4-6 in Table 1). The maximum variance method (MVM) was utilized to examine the principal components. Statistically, Cronbach’s alpha was used to measure the internal consistency between these behavior and perception variables. The variables resulting from the factor analysis, together with demographic, exposure and crash variables, were entered into the model construction.

The modelling were carried out with logistic regression models to deal with the binary nature of the dependent variables, such as involvement in traffic crashes (code = 1) or no involvement in traffic crashes (code = 0). Two logistic regression models were constructed to examine the predictive ability of driver behavior and attitudes to work and non-work crashes, above and beyond simple driving exposure. Driving exposure consisted of hours of driving/week and km travelled/year. In order to identify more contributing factors on traffic crashes, beyond those involving the demographic, behavior and perception variables, self-reported crashes by the participants were classified into crashes during working and non-working hours.

Results and Discussion

3.1.   Principal Component Analysis (PCA)

Table 2 shows the loading factors of the seven behavior items in a group of driving experiences in the previous six months, consisting of items no 4.14 (failure to check rear view mirror before pulling out or changing lanes); 4.18 (racing away from tra?c lights with the intention of beating the driver next to you); 4.20 (driving even though you suspect you may be over the legal blood-alcohol limit); 4.21 (disregarding the speed limit on a residential road); 4.22 (exceeding the speed limit on a residential road without realising it); 4.23 (becoming angered by another driver and giving chase); and 4.32 (having one or two alcoholic drinks before driving for work). Statistically, these items explained variances greater than 20% and Cronbach's Alpha was used to measure loading factors more than 0.7 (Hooper et al., 2008).

In addition, items relating to driver attitude were significant and classified into two groups. The first group consisted of items 5.9 (it’s ok to have a few alcoholic drinks before driving home after work at the end of the weekend) and 5.12 (speed limits are often set too low, with the result that many drivers ignore them), while the second group contained items 5.3 (it is quite acceptable to take a slight risk when overtaking) and 5.6 (some people can drive perfectly, even when they only leave a small gap between the vehicle in front)Item no 6 were classified into a factor which justified less than 20% of variance, which is statistically insignificant (Hooper et al., 2008). As a result, these items are not included in the model development. The variables and factors employed in the model construction are shown in Table 2.

 

Table 2 Behavior and perception variable selection using PCA

No

Factors

Variance Explained

(> 20%)

Question No.

Cronbach's Alpha

(> 0.7)

 

1.

Driving experience in the past 6 months

20.381%

4.14

4.18

4.20

4.21

4.22

4.23

4.32

0.707

0.816

0.884

0.871

0.706

0.720

0.761

2.

Attitude towards alcohol & speed violations

22.326

5.9

5.12

0.817

0.782

3.

Attitude towards risky overtaking and following too closely

20.039

5.3

5.6

0.831

0.775


With reference to a three factor solution shown in Table 2, the first factor contains seven items associated with a mixture of aggressive driving behaviors (questions 4.18 and 4.23), driving errors (questions 4.14 and 4.22) and traffic code violations (questions 4.20, 4.21 and 4.32). All the items in the second and third factors represent attitudes towards unsafe and risky driving behaviors, associated with alcohol, speeding, risky overtaking and distance between other vehicles. The analysis indicates that four types of behavior, traffic code violations, driving errors, aggressive driving behaviors and driver attitudes, are significant in influencing goods vehicle driver behavior in Denpasar, Bali. These findings are very similar to those of a study by Davey et al. (2007), which examined the self-reported driving behaviors of Australian ?eet drivers and found a three significant factor, identified as aggressive driving violations, errors and highway code violations.

3.2.   Prediction of Traffic Crashes 

The variables of interest were entered into the logistic regression model using SPSS version 15  with the exposure factors first, followed by the behaviors. Both chi squares (steps 1 and 2 in Table 3) for the traffic crashes during working hours model were 66.81 (p < 0.001); for the H-L (Hosmer-Lemeshow) test less than 0.05 and 168.92 (p < 0.001) and H-L test greater than 0.05, while both chi squares (steps 1 and 2 in Table 4) for the non-working hours model were 126.29 (p < 0.001 and H-L Test less than 0.05) and 327.76 (p < 0.001 and H-L Test greater than 0.05). Considering these results, steps 2 in Tables 3 and 4 for both traffic crashes during working and non-working hours models were statistically significant.

As shown in Table 3 (step 2), hours of driving per week and age of driver were significantly and negatively related to traffic crashes during working hours. In addition, the goods vehicle drivers’ responses to the items ‘race away from tra?c lights with the intention of beating the driver next to you’, ‘become angered by another driver and give chase’, and ‘speed limits are often set too low with the result that many drivers ignore them’ were significantly and negatively related to traffic crashes during working hours. In contrast, the goods vehicle drivers’ responses to the statementsdrive even though you suspect you may be over the legal blood-alcohol limit’, ‘it is quite acceptable to take a slight risk when overtaking’ and ‘it’s ok to have a few alcoholic drinks before driving home after work at the end of the weekend’ were significantly and positively linked to traffic crashes during working hours. In other words, driving experiences and normlessness were generally significant and both positively and negatively influenced goods vehicle drivers’ involvement in traffic crashes during working hours in Denpasar.

 

Table 3 Working hours logistic regression model

Step 1

S.E.

Wald

Sig.

Exp(B)

95% C.I.for EXP( )

Lower

Upper

Hours of driving/week

-.368

.103

12.745

.000

.692

.565

.847

Km travelled/year

.088

.077

1.297

.255

1.092

.939

1.270

Step 2

S.E.

Wald

Sig.

Exp(B)

95% C.I.for EXP( )

Lower

Upper

Hours of driving/week

-.556

.142

15.238

.000***

.573

.434

.758

Km travelled/year

.096<

Conclusion

The results of the study indicate that drink driving is positively related to traffic crashes during working hours. In addition, a greater number of items of normlessness influenced more than driving errors, aggresive behaviors and traffic rule violations on traffic crashes. These normlessness factors (i.e speeding behavior and taking risks when overtaking), however, may also re?ect traffic rule violations, but do not necessarily reflect the situation that self-reported behaviors are only applicable to normlessness.

Further behavioral studies are required to analyze the contributing factors to traffic crashes involving drivers transporting goods and those transporting freight. In addition, a further comparative study is required to analyze the traffic crash involvement of drivers of passenger cars and buses and drivers transporting goods and freight.

Acknowledgement

Funding for thiswork was provided by the University of Udayana, Bali, Indonesia.

Supplementary Material
FilenameDescription
R3-CVE-960-20180427143717.JPG ---
References

Arnold, P.K., Hartley, L.R., Corry, A., Hochstadt, D., Penna, F., Feyer, A.M., 1997. Hours of Work and Perceptions of Fatigue among Truck Drivers.  Accident Analysis and Prevention, Volume 29(4), pp. 471–477

Bjerre, B., Kostela. J., 2008. Primary Prevention of Drink Driving by the Large-scale Use of Alcolocks in Commercial Vehicles. Accident Analysis and Prevention, Volume 40(4), pp. 1294–1299

Darby, P., Murray, W., Raeside, R., 2009. Applying Online Fleet Driver Assessment to Help Identify, Target and Reduce Occupational Road Safety Risks. Safety Science, Volume 47(3), pp. 436–442

Davey, J., Freeman, J., Wishart, D., 2006. A Study Predicting Crashes among A Sample of Fleet Drivers. In: Proceedings Road Safety Research, Policing and Education Conference, Gold Coast, Queensland

Davey, J., Wishart, D., Freeman, J., Watson, B., 2007. An Application of the Driver Behavior Questionnaire in an Australian Organisational Fleet Setting. Transportation Research Part F, Volume 10(1), pp. 11–21

de Winter, J.C.F., Dodou, D., 2010. The Driver Behavior Questionnaire as a Predictor of Accidents: A Meta-analysis. Journal of Safety Research, Volume 41(6), pp. 463–470

Eksler, V., Janitzek, T., 2010. Drink Driving in Commercial Transport.  European Transport Safety Council

Freeman, J.E., Wishart, D.E., Davey, J.D., Rowland. B.D., 2008. Driving Safely for Work : A Study Investigating Aberrant Driving Behaviors within A Fleet Setting. In: Proceedings of CMRSC-XVIII, Whistler, Canada

Freeman, J.E., Wishart, D.E., Davey, J.D., Rowland, B.D., Williams, R., 2009. Utilising the Driver Behavior Questionnaire in an Australian Organisational Fleet Setting: Can It Identify Risky Drivers? Journal of the Australasian College of Road Safety, Volume 20(2), pp. 38–45

Glendon, A.I., Stanton, N.A., 2000.  Perspectives on Safety Culture. Safety Science, Volume 34, pp. 193–214

Hooper, D., Coughlan, J., Mullen, M., 2008. Structural Equation Modelling: Guidelines for Determining Model Fit. Electronic Journal of Business Research Methods, Volume 6(1), pp. 53–60

Indonesian National Police., 2014. Accident Data Report. Jakarta

Lajunen, T., Summala, H., 2003. Can We Trust Self-reports of Driving? E?ects of Impression Management on Driver Behavior Questionnaire Responses. Transportation Research Part F, Volume 6(2), pp. 97–107

 Matthews, G., Desmond, P.A., Joyner, L., Carcary, B., Kirby, G., 1997. A Comprehensive Questionnaire Measure of Driver Stress and Affect. In: Rothengatter, T., & Vaya E.C. (Eds.) Traffic and Transport Psychology: Theory and Application, pp. 317–324

Newnam, S., Watson, B., Murray W., 2002. A Comparison of the Factors Influencing the Safety of Workrelated Drivers in Work and Personal Vehicles. In: Proceedings of the Road Safety Research, Policing and Education Conference, Adelaide, Transport South Australia

Parker, D., Stradling, S.G., Manstead, A., 1996. Modifying Beliefs and Attitudes to Exceeding the Speed Limit: An Intervention Study Based on the Theory of Planned Behaviour. Journal of Applied Social Psychology, Volume 26(1), pp. 1–19

Poulter, D.R., Chapman, P., Bibby, P.A., Clarke, D.D., Crundall, D., 2008. An Application of the Theory of Planned Behavior to Truck Driving Behavior and Compliance with Regulations. Accident Analysis and Prevention, Volume 40(6), pp. 2058–2064

Prochaska, J.O., DiClemente, C.C., 1984. The Transtheoritical Approach: Towards a Systematic Eclectic Framework. Dow Jones Irwin, Homewood, IL, USA

Reason, J., Manstead, A., Stradling, S., Baxter, J., Campbell, K., 1990. Errors and Violations on the Roads: A Real Distinction? Ergonomics, Volume 33(10-11), pp. 1315–1332

Rhodes, N., Pivik, K., 2011. Age and Gender Differences in Risky Driving: The Roles of Positive Affect and Risk Perception. Accident Analysis and Prevention, Volume 43(3), pp. 923–931

Statistics of Bali Province., 2015. Bali in Figures. Bali

Sullman, M.J.M., Meadows, M.L., Pajo, K.B., 2002. Aberrant Driving Behaviors amongst New Zealand Truck Drivers. Transportation Research Part F, Volume 5(3), pp. 217–232

The World Health Organization, 2015. Global Status Report on Road Safety 2015. Switzerland

Wishart, D.E., Davey, J.D., 2004. A Research Based Case Study Approach to the Development of Fleet Safety Interventions in Large Vehicle Fleets. In: Proceedings Safety in Action Conference, Melbourne, Victoria

Zuraida, R., Iridiastadi, H., Sutalaksana, I.Z., 2017. Indonesian Drivers’ Characteristics Associated with Road Accidents. International Journal of Technology. Volume 8(2), pp. 311–319