Accident Analysis and Prevention 40 (2008) 14111417

Contents lists available at ScienceDirect

Accident Analysis and Prevention

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

The Driver Behaviour Questionnaire in Arab Gulf countries: Qatar and United Arab Emirates
Abdulbari Bener a,b, , Turker Ozkan c , Timo Lajunen c
a

Department of Medical Statistics & Epidemiology, Hamad General Hospital, Hamad Medical Corporation & University of Qatar, Doha, Qatar Department of Evidence for Population Health Unit, School of Epidemiology and Health Sciences, The University of Manchester, Manchester, UK c Safety Research Unit, Department of Psychology, Middle East Technical University, Ankara Turkey
b

a r t i c l e

i n f o

a b s t r a c t
Manchester Driver Behaviour Questionnaire (DBQ) is one of the most widely used instruments for measuring self-reported driving style and investigating the relationship between driving behaviour and accident involvement. In spite of the fact that Arab Gulf countries have a higher road accident fatality rate compared to European countries and USA, the DBQ has not been used in Arab countries so far. The aim of the present study was to investigate the factor structure of the DBQ, then to examine the relationships between the factors of the DBQ and accident involvement, and nally to compare DBQ scores between the two gulf countries: Qatar and United Arab Emirates (UAE). In this study, 1110 Qatari (263 females and 847 males) and 1286 UAE drivers (294 females and 992 males) lled a survey questionnaire including the DBQ and background information. The results showed that UAE drivers scored higher on almost all DBQ items than Qatari drivers. Surprisingly, only very small differences between men and women on the DBQ item scores were found in UAE. Factor analysis resulted in four factors, which were named as errors, pushing-speeding violations, lapses, and aggression-speeding violations. However, there were a number of differences in the factor structure of the DBQ in UAE and Qatar when compared to the theoretical four-factor structure of the DBQ. Reliabilities of some subscales were also questionably lower than in the original British data. Logistic regression analyses showed that errors, lapses, and aggression-speeding violations predicted accident involvement in Qatar but not in UAE after controlling the effect of the demographic variables (age, sex, and annual mileage). 2008 Elsevier Ltd. All rights reserved.

Article history: Received 19 July 2007 Received in revised form 10 March 2008 Accepted 13 March 2008 Keywords: Driver Behaviour Questionnaire (DBQ) Trafc accidents Qatar United Arab Emirates (UAE)

1. Introduction Compared to European countries and USA, Arab countries have a very high road accident fatality rate. In 2000, 14.8, 11.7, and 7.3 persons per 10,000 vehicles were killed in Saudi, United Arab Emirati and Qatari road trafc, respectively (Bener et al., 2003). The same gures (1999 statistics) were, for example, approximately 1.8 for Finland, 2.4 for France, 1.5 for UK, and 1.9 for USA (IRF, 2003). Analyses of trafc accidents indicate that human factors are a sole or a contributory factor in road trafc accidents (Lewin, 1982). Human factors in driving can be seen as being composed of two separate components, driving skills and driving style (Elander et al., 1993). Driving skills include those information processing and motor skills, which improve with practice and training, i.e. with driving experience. Driving style concerns individual driving habits,

Corresponding author at: Department of Medical Statistics & Epidemiology, Hamad General Hospital & Hamad Medical Corporation, Weill Cornell Medical College, P.O. Box 3050, Doha, Qatar. Tel.: +974 439 3765/6; fax: +974 439 3769. E-mail addresses: abener@hmc.org.qa, abaribener@hotmail.com (A. Bener). 0001-4575/$ see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.aap.2008.03.003

i.e. the way a driver chooses to drive. Driving style becomes established over a period of years, but does not necessarily get safer with driving experience (Elander et al., 1993). Manchester Driver Behaviour Questionnaire (the DBQ) is one of the most widely used instruments for measuring driving style. The DBQ is based on a theoretical taxonomy of aberrant behaviours divided into violations and errors (see Reason, 1990; DBQ by Reason et al., 1990). Violations refer to deliberate deviations from those practices believed necessary to maintain the safe operation of a potentially hazardous system. Violations can be classied as aggressive and ordinary ones (Lawton et al., 1997). The aggressive ones involve overtly aggressive acts whereas the ordinary ones consist of deliberately breaking the Highway codes and/or law without aggressive motives. Errors were dened as a failure of planned actions to achieve their intended consequences that can involve the unwitting deviation of action from intention (slips and lapses) or departure of planned actions from some satisfactory path toward a desired goal (mistakes). In spite of small differences in the theoretical four-factor structure of the DBQ, the overall factor structure has been conrmed in different samples (for detailed description see Mesken et al., 2002;

1412

A. Bener et al. / Accident Analysis and Prevention 40 (2008) 14111417

Lajunen et al., 2004; Sullman et al., 2002; Bener et al., 2007; Gras et al., 2006) also obtained the four-factor structure of the DBQ among Spanish drivers. However, there were a number of differences to the theoretical four-factor structure of the DBQ. Results of the factor analysis showed that lapses items loaded mostly on errors factors, which was a mixture of lapses and errors and one aggressive violations item. Lapses factor was relatively unreliable and composed of only three original lapses items. The Spanish factor structure did not include the original aggressive violations factor either. Rather, the results conrmed the interpersonal violations factor as a separate entity from the other aggressive violations. Besides, a strong violations factor including both ordinary and aggressive violations was obtained and the content of the factor was based on mainly hurrying up in trafc. In the study by Gras et al. (2006) the alpha reliabilities for errors, lapses, interpersonal violations, and violations were 0.82, 0.46, 0.59, and 0.81, respectively. Except for lapses factor, reliabilities of the scales were at about the same level as in the original British data and other previous data (e.g., Lajunen et al., 2004; Parker et al., 1995b). The DBQ scores have also been found to be reliable over time (Parker et al., 1995a; Ozkan et al., 2006a). It has been, in general, reported that women and older drivers tend to commit violations less frequently than men and young drivers; female and older drivers, on the other hand, commit more errors than male and young drivers; it has also been found that the more drivers drive, the more often they tend to violate traf c rules (Aberg and Rimmo, 1998; Blockey and Hartley, 1995; Reason et al., 1990). Besides, Ozkan et al. (2006b) reported that Western/Northern European drivers scored lower on errors and violations items than Southern European/Middle Eastern drivers. It can be concluded that both external factors (e.g., trafc culture) and internal factors (e.g., age and gender) inuence the frequency of different driver behaviours (see Ozkan, 2006). The ndings of the previous studies have also shown that selfreported driving violations are associated with both active (a driver hits another car or road user) and passive (a driver is hit by another vehicle) accidents. Besides, violations predicted accident involvement, both retrospectively and prospectively (Parker et al., 1995a,b). In particular, violations have been reported to be associated with active loss-of-control and passive right-of-way accidents (Parker et al., 1995b), as well as with speeding and parking offences (Mesken et al., 2002). Lapses have rarely predicted accident involvement in previous studies (Mesken et al., 2002), but errors seemed to be main predictor of involvement in active accidents among elderly drivers (Parker et al., 2000). However, empirical research about the DBQ factor structure and its relationship with accident involvement has not been studied in Arabic countries before. The aim of the present study was to investigate the factor structure of the DBQ, then to examine the relationships between the factors of the DBQ and accident involvement, and nally to compare DBQ scores between the two gulf countries: Qatar and United Arab Emirates (UAE).

represent over 70% of total visits per year. Qualied nurses and health educators were instructed to structurally interview and complete a questionnaire for randomly selected Qatari men and women during the period from October 2004 to March 2005. One thousand one hundred and ten drivers (263 female and 847 male) participated in this study. 80.5% of drivers reported that they had been involved in at least one trafc accident. 2.1.2. United Arab Emirati drivers A cross-sectional survey was conducted the period from January to July 2002 in the Al Ain City, Abu Dhabi and Dubai cities in the UAE. A multi-stage stratied cluster sampling design was developed using an administrative division of the UAE into three cities of approximately equal size in terms of number of inhabitants. In order to ensure a representative sample of the study population, the sampling plan was stratied with proportional allocation according to stratum size. Stratication was based upon geographical location. With 2.5% error bound and 99% condence limit the required sample size was estimated as 1800 drivers. The participants were sampled from each region so that the sample size in each region was proportional to its share of total population in the Emirates. A health educator and social workers recorded the data of each subject on a standardized questionnaire. A representative sample of 1800 UAE drivers was selected and approached while renewal of car registration. The sample included males and females aged 18 years and above. A total number of 1286 Arabian Emirati drivers (294 females and 992 males) took part in the study (Bener et al., 2007). 55.8% of drivers reported that they had been involved in at least one trafc accident. All participants had driving licenses and were assured of anonymity and condentiality. The participants lled out the DBQ and items related to drivers driving records and demographic variables. Although the sampling strategy was different in the two studied countries, the Driver Behaviour Questionnaire was the tool used to investigate the relationship between self-reported driving questionnaire and crash involvement in both countries. 2.2. Measures 2.2.1. Aberrant driver behaviours DBQ with extended violations was used to measure aberrant driver behaviours (Lajunen et al., 2004; Lawton et al., 1997). An Arabic version of the DBQ was revised by the bilingual coinvestigator and back translated by another bilingual expert. The translators made together the necessary corrections, modications and rewording after considering the minor differences and discrepancies. The extended version of the DBQ includes aggressive and ordinary violations (10 items, e.g. disregard speed limit on motorways), lapses (8 items, e.g. forget where you left your car in the car park, and errors (8 items, e.g. misjudging the speed of another vehicle when overtaking). Two violations items were dropped because of cultural or structural factors (e.g., drinking and driving, and pulling out of, force your way). Gras et al. (2006) also suggested that pulling out of, force your way item might be omitted from future research because of its instability across the DBQ factors. Two violations items were also revised after getting feedback from drivers. Disregard speed limits on a residential road was revised as disregard the speed limits late at night or early in the morning. Drive close to the car in front, making it difcult to stop in an emergency was revised as Drive especially close to the car in front as a signal to its driver to go faster or get out of the way. Participants were asked to indicate how often they committed each of the 26 behaviours in the previous year on a six-point scale (0 = never, 1 = hardly ever, 2 = occasionally, 3 = quite often, 4 = frequently, and 5 = nearly all the time).

2. Method 2.1. Participants 2.1.1. Qatari driver sample A multi-stage stratied cluster sampling was applied by using the administrative division of the Qatar into twenty-one Primary Health Care (PHC) Clinics of the State of Qatar. PHCs are approximately equal sized in terms of number of inhabitants. The participants were selected among patients registered and attending eleven PHC Centres (8 urban and 3 semi-urban), which

A. Bener et al. / Accident Analysis and Prevention 40 (2008) 14111417

1413

2.2.2. Demographic variables Participants were asked to indicate their age, gender, marital status, educational level, occupation, place of living, housing conditions, driving experience, type of car, frequency of seat belt use, reasons for not wearing a seat belt, speed choice on different roads, annual mileage, trafc offences, history of accident and injury involvement.

Table 1 Socio-demographic characteristics of drivers surveyed in Qatar and UAE Variables Age group <20 2029 3039 4049 50 Gender Male Female Education Illiterate Primary Intermediate Secondary University Occupation Not working Sedentary/professional Manual Business man Housewife Army/police Student Driving experience <1 year 25 years 59 years 10 years Car type 4WD Small car Seat belt use Yes No Qatar, n = 1110 (%) 51 (4.6) 375 (33.8) 361 (32.5) 225 (20.3) 98 (8.8) 847 (76.3) 263 (23.7) 148 (13.3) 254 (22.9) 148 (13.3) 382 (34.4) 178 (16.0) 77 (6.9) 392 (35.3) 204 (18.4) 158 (14.2) 92 (8.3) 114 (10.3) 73 (6.6) 147 (13.2) 314 (28.3) 252 (22.7) 397 (35.8) 436 (39.3) 674 (60.7) 501 (45.1) 609 (54.9) UAE, n = 1286 (%) 129 (10.0) 386 (30.0) 382 (29.7) 288 (22.4) 101 (7.9) 992 (77.1) 294 (22.9) 233 (18.1) 364 (28.3) 332 (25.8) 196 (15.2) 161 (12.5) 49 (3.8) 198 (15.7) 295 (22.9) 297 (23.1) 127 (9.9) 192 (14.9) 128 (10.0) 217 (16.9) 329 (25.6) 303 (23.6) 437 (34.0) 513 (39.9) 773 (60.1) 517 (40.2) 769 (59.8) P value

<0.001

2.3. Statistical analyses The analyses were performed using the Statistical Package for Social Sciences (SPSS) (Norusis, 1998) for the statistical analysis. One-way analysis of variance (ANOVA) was run to study whether there were signicant differences between Qatari and United Arab Emirati drivers and between male and female drivers on the DBQ item scores. Principal component analysis (PCA) with varimax rotation was run to examine the factor structure of the DBQ in Qatar and UAE. Cronbachs alpha reliability coefcients were also calculated for assessing the internal consistency of the DBQ scale scores. Direct logistic regression was performed with accident involvement as dependent and four DBQ factors as independent variables. Age, sex, and annual mileage were forced into the model to control their effect.

0.631

<0.001

<0.001

0.057

3. Results In Qatar, the age distribution of the participants ranged from 18 to 70 years with the mean age of 33.72 years (S.D. 0.21 years) and the mean of annual mileage was 25,900 km (S.D. 38,820 km). In UAE, the age distribution of the participants ranged from 19 to 70 years with the mean age of 33.51 years (S.D. 10.02) and the mean of annual mileage was 21,730 km (S.D. 6790).

0.760

0.014

3.1. Socio-demographic characteristics of the surveyed drivers in Qatar and UAE Signicant differences were found between Qatar and UAE in driver age, education, occupation and seat belt use Table 1.

3.3. The factor structure of the DBQ in Qatar and UAE The criteria used to determine the number of factors were the Kaiser criterion of eigenvalues over 1.0, the Cattell Scree plot, parallel analysis, and the interpretability of factors. Initially, eight and nine factors had eigenvalues over 1.0 in Qatar and UAE samples, respectively. However, the Scree plot and parallel analysis suggested the four-factor solution to be the most interpretable one in both samples. As presented in Table 3, the rst component included 10 items in the Qatar sample and seven items in the UAE sample. It seems to reect mostly errors on the road, and thus, was labelled as errors. However, the factor included two lapses (misreading the signs, exit from a roundabout on the wrong road and getting into the wrong lane approaching a roundabout or a junction) and one ordinary violations item (shooting lights) in the Qatar sample while the factor included one lapses (getting into the wrong lane approaching a roundabout or a junction) and one ordinary violations item (shooting lights) in UAE sample. The second component consisted of the same ve items both in the Qatar and UAE samples. This component seems to describe a drivers speeding related activities and a kind of pushing strategies on the car in front, and thus, was labelled as pushing-speeding. The third component included ve items in the Qatar sample and the same four items in the UAE sample. It seems to reect clear lapses factor. The fourth component consisted of four items in the Qatar sample and ve items in the UAE sample. This component seems to describe a drivers interpersonal aggression and speeding related competition, and thus, was labelled as aggression-speeding.

3.2. Descriptive statistics for the DBQ item scores: Qatari and United Arab Emirati drivers The means and standard deviations of the DBQ items for Qatar and UAE can be found in Table 2. Get into the wrong lane approaching a roundabout or a junction and aversion, indicate hostility were the least frequently reported driver behaviours whereas the two most frequently reported driver behaviours were overtake a slow driver on the inside and sound horn to indicate your annoyance in Qatar and cross a junction when the trafc lights have turned red and disregard the speed limits on motorway in UAE, respectively. Qatari male drivers scored higher on all DBQ items than Qatari female drivers except on miss give way signs. The most striking nding of the present study was that there was no signicant difference between male and female Emirati drivers on the DBQ items except hitting something when reversing. United Arab Emirati male drivers scored signicantly higher on the following six DBQ items than female drivers: cross a junction when the lights turned red, angered by another drivers behaviour, attempt to overtake someone, queuing to enter to main road and nearly hit the car in front, hit a two wheeler, and fail to check your rear-view mirror.

1414

Table 2 Means and standard deviations of DBQ items separately for male and female drivers in Qatar and UAE Variables Violations Drive especially close to the car in front as a signal to its driver to go faster or get out of the way Cross a junction knowing that the trafc lights have already turned red Disregard the speed limits late at night or early in the morning Disregard the speed limits on a motorway Have an aversion to a particular class of road user and indicate your hostility by whatever means you can Become impatient with a slow driver in the outer lane and overtake on the inside (right) lane Get involved with unofcial races with other drivers Angered by another drivers behaviour, you give chase with the intention of giving him/her a piece of your mind Sound your horn to indicate your annoyance to another driver Stay in a motorway that you know will be closed ahead until the last minute before forcing you way into the other lane Errors Attempt to overtake someone that you had not noticed to be signalling a left/right turn Miss give way signs and narrowly avoid colliding with trafc having right of way Fail to notice that pedestrians are crossing when turning into a side street from a main road Queuing to turn right/left onto a main road, you pay such close attention to the mainstream of trafc that you nearly hit the car in front On turning right/left nearly hit a two wheeler who has come up on your inside Fail to check your rear-view mirror before pulling out or changing lanes, etc. Underestimate the speed of an oncoming vehicle when overtaking Apply sudden brakes on a slippery road, or steer wrong way in a skid Lapses Get into the wrong lane when approaching a roundabout or a junction Misread the signs and exit from the roundabout on the wrong road Forget where you left your car in the car park Hit something when reversing that you had not previously seen Attempt to drive away from the trafc lights Switch on one thing, such as headlights, when you meant to switch on something else, such as wipers Intending to drive to destination A and, you wake up to nd yourself in destination B, because the latter is your more usual destination Realize you have no clear recollection of the road along which you have been travelling
* ** ***

Qatar total 1.48 1.59 1.28 1.57 1.79 1.71 1.41 1.61 1.29 1.52 1.94 1.81 1.59 1.67 1.46 1.52 1.91 1.71 1.34 1.54

UAE total 1.71 1.73 2.69 1.85 2.67 1.86 1.73 1.87 1.39 1.61 1.98 1.72 2.25 1.82 1.83 1.67 2.09 1.69 1.40 1.63

F-values 10.95*** 397.23 141.72*** 19.85*** 2.48 .26 86.02 31.12*** 6.95** .96
*** ***

Qatar male 1.53 1.60 1.36 1.56 1.87 1.68 1.49 1.62 1.31 1.52 1.97 1.79 1.60 1.66 1.50 1.51 1.93 1.70 1.40 1.55

Qatar female 1.35 1.56 1.05 1.57 1.56 1.79 1.15 1.54 1.20 1.50 1.87 1.88 1.56 1.69 1.35 1.55 1.83 1.73 1.16 1.50

F-values 2.53 7.84 6.74** 8.77** 1.17 .50 .12 2.02 .76 4.74*
**

UAE male 1.70 1.73 2.71 1.83 2.67 1.86 1.73 1.87 1.39 1.62 1.93 1.69 2.24 1.82 1.84 1.68 2.04 1.68 1.42 1.63

UAE female 1.76 1.70 2.63 1.89 2.67 1.85 1.73 1.88 1.39 1.57 2.15 1.77 2.28 1.78 1.80 1.65 2.26 1.68 1.34 1.60

F-values .26 .42 .00 .00 .00 3.66 > .056 A. Bener et al. / Accident Analysis and Prevention 40 (2008) 14111417 .09 .10 3.72 > .054 .65

1.35 1.67 1.43 1.64 1.40 1.52 1.66 1.71 1.57 1.59 1.37 1.44 1.51 1.52 1.39 1.32 1.11 1.43 1.56 1.78 1.14 1.33 1.47 1.40 1.49 1.45 1.85 1.75 1.39 1.49 1.44 1.63

1.68 1.96 1.83 1.63 1.65 1.59 2.55 1.81 2.22 1.62 1.96 1.69 1.93 1.57 1.85 1.56 1.70 1.79 1.99 1.82 1.49 1.43 1.66 1.51 1.62 1.56 2.32 1.81 1.68 1.51 1.49 1.66

19.94*** 36.36*** 15.74*** 149.41

***

1.36 1.67 1.42 1.62 1.42 1.54 1.70 1.70 1.61 1.59 1.39 1.42 1.52 1.50 1.39 1.31 1.12 1.41 1.59 1.79 1.19 1.35 1.55 1.44 1.52 1.45 1.83 1.75 1.43 1.51 1.44 1.61

1.30 1.67 1.45 1.70 1.34 1.48 1.54 1.74 1.44 1.60 1.31 1.49 1.48 1.57 1.37 1.34 1.08 1.50 1.48 1.73 0.98 1.25 1.23 1.23 1.40 1.44 1.88 1.77 1.27 1.40 1.45 1.67

.31 .07 .57 1.81 2.04 .66 .12 .04 .16 .71 5.19* 10.11** 1.26 .12 2.13 .01

1.71 1.98 1.78 1.63 1.65 1.60 2.58 1.80 2.23 1.62 2.00 1.69 1.92 1.55 1.84 1.55 1.67 1.76 2.01 1.83 1.52 1.44 1.73 1.53 1.64 1.56 2.31 1.82 1.70 1.52 1.44 1.63

1.58 1.87 1.99 1.62 1.67 1.52 2.41 1.83 2.21 1.61 1.85 1.69 1.97 1.62 1.88 1.58 1.79 1.86 1.92 1.75 1.38 1.37 1.44 1.40 1.55 1.54 2.35 1.76 1.62 1.48 1.69 1.72

1.01 3.52 .03 1.96 .03 1.62 .21 .18 .99 .55 2.26 8.63** .65 .12 .54 5.13*

98.50*** 82.78*** 43.91*** 59.18*** 77.94*** 32.87*** 37.37*** 10.36*** 4.27* 42.37*** 21.97*** .64

p < 0.05. p < 0.01. p < 0.001.

A. Bener et al. / Accident Analysis and Prevention 40 (2008) 14111417

1415

One DBQ item, intending to drive to destination A, instead drive to B, did not load on any of the factors neither in the Qatar nor in the UAE sample. Disregarding speed limits late at night or early in the morning item did not load on any of the factors in the Qatar sample while getting angry, give chase, having no clear recollection of the road, missing Give Way signs, and misreading the signs, exit from a roundabout on the wrong road did not load on any of the factors in the UAE sample (see Table 3). 3.4. Reliability coefcients for the DBQ subscales in Qatar and United Arab Emirates Reliability analyses for the four-factor solution of the DBQ revealed that the alpha reliabilities for errors, pushing-speeding, lapses, and aggression-speeding factors were 0.70, 0.67, 0.60, and 0.50 in the Qatar sample, and 0.56, 0.64, 0.58, and 0.24 in the UAE sample, respectively. The reliabilities scores were, in general, very low and, for example, aggression-speeding factor was highly unreliable in the UAE sample. 3.5. The relationship between DBQ factors and accident involvement 3.5.1. In Qatari driver sample A test of the full model with all predictors against a constant-only model was statistically signicant, ( 2 = 176.89, 7,1110 p < 0.001), indicating that the set of the predictors showed signicant difference between drivers with accidents and without accidents. The variance in accident involvement was 15% with an overall rate of 86% of drivers correctly classied. As shown in Table 4, errors, lapses, and aggression-speeding factors predicted accident involvement. The odds ratio of 2.82 showed a relatively large change

in the likelihood of accident involvement on the basis of a oneunit change in errors. The ratio was 1.51 for lapses and 1.19 for aggression-speeding. 3.5.2. In United Emirati driver sample A test of the full model with all predictors against a constantonly model was not statistically signicant, ( 2 = 11.45, 7,1286 p > 0.05), indicating that the set of the predictors did not show any signicant difference between drivers with accidents and without accidents. Expectedly, the variance in accident involvement is very small (i.e., 1%) and success of prediction was unimpressive, with an overall rate of 56% of drivers correctly classied. As shown in Table 4, none of the variables predicted signicantly accident involvement. 4. Discussion Consistently with previous studies (Aberg and Rimmo, 1998; Blockey and Hartley, 1995; Mesken et al., 2002), the frequencies of the DBQ responses were, in general, between never to hardly ever and rarely occasionally both in Qatar and UAE. However, the scores of Qatari and United Arab Emirati DBQ items were, in general, higher than the scores of British (Reason et al., 1990), Australian (Blockey and Hartley, 1995), Finnish and Dutch (Lajunen et al., 2004), and Greek and Turkish (e.g., Ozkan et al., 2006b) drivers. United Arab Emirati drivers scored higher on all DBQ items than Qatari drivers, except for four DBQ items (see Table 2). However, the factor structure and reliability scores of the DBQ seem to require some culture-specic items as well, especially in United Arab Emirates. In addition, the results of the present study clearly support the idea (see Ozkan, 2006) that external factors (e.g., trafc context with the lack of enforcement and applied rules, and congestion) could sometimes be much more important than internal factors

Table 3 Four-factor solution of the DBQ items, eigenvalues and Cronbachs alpha coefcients, and variance of the DBQ subscales both in Qatar (Q) and United Arab Emirates (UAE) data Four-factor order in PCA Items (original type) Queuing, nearly hit car in front (E4) Fail to see pedestrians crossing (E3) Fail to check your rear-view mirror (E6) Brake too quickly on a slippery road (E8) Turning right nearly hit cyclist (E5) Misread the signs, exit from a roundabout on the wrong road (L2) Underestimate the speed of an oncoming vehicle (E7) Shooting lights (OV2) Get into the wrong lane approaching a roundabout or a junction (L1) Miss give way signs (E2) Disregard the speed limit on motorway (OV4) Close following as a signal (OV1) Attempt to overtake someone turning left (E1) Aversion, indicate hostility (AV5) Push in at last minute (AV10) Forget where you left your car (L3) Hit something when reversing (L4) Switch on one thing, meaning the other (L6) Attempt to drive away in third gear (L5) Have no clear recollection of the road (L8) Race from lights (AV7) Get angry, give chase (AV8) Overtake a slow driver on the inside (OV6) Sound horn to indicate your annoyance (AV9) Intending to drive to destination A, instead drive to B (L7) Disregard the speed limit morning-night (OV3) Eigenvalues Cronbachs alpha Variance (%) Note: Factor loadings below .30 were omitted for the sake of clarity. Errors Q 0.67 0.59 0.57 0.51 0.5 0.49 0.43 0.41 0.4 0.35 UAE 2 0.73 0.36 0.54 0.34 0.35 0.76 0.31 0.3 0.64 0.53 0.67 0.62 0.68 0.77 0.77 0.52 0.49 0.33 Pushing-speeding Q UAE 1 Lapses Q UAE 3 Aggression-speeding Q UAE 4

0.32 0.52 0.33 0.38

0.75 0.61 0.6 0.59 0.58

0.38

0.73 0.75 0.51 0.58 0.67 0.64 0.55 0.51 1.38 .50 5.31

4.25 .70 16.3

2.74 .56 10.55

2.04 .67 7.87

1.86 .64 7.17

1.42 .60 5.47

1.63 .58 6.29

0.5 0.32 0.36 0.34 1.44 .28 5.54

1416

A. Bener et al. / Accident Analysis and Prevention 40 (2008) 14111417

Table 4 Logistic analyses of accident involvement as a function of aberrant driver behaviour among Qatari and United Arab Emirati drivers Variables 95% condence interval for odds ratio B Qatar sample Age Sex Annual km Errors Pushing-speeding Lapses Aggression-speeding UAE sample Age Sex Annual km Errors Pushing-speeding Lapses Aggression-speeding Wald test (z-score) .39 3.38 6.65 60.05 1.17 14.12 4.27 1.17 7.91 .03 .00 .39 .23 .88 Odds ratio .99 .71 1.00 2.83 1.11 1.51 1.19 1.00 .68 1.00 .99 .96 .97 .94 Lower Upper

.00 .33 .00 1.04 .10 .41 .17 .00 .37 .00 .00 .03 .02 .06

1.01 .50 1.00 2.17 .91 1.21 1.00 .99 .52 1.00 .88 .87 .87 .82

.98 1.02 1.00 3.68 1.34 1.87 1.41 1.01 .89 1.00 1.13 1.07 1.08 1.07

errors than men (Aberg and Rimmo, 1998; Blockey and Hartley, 1995; Reason et al., 1990), there was no signicant difference between male and female Emirati drivers on DBQ item scores except on item hitting something when reversing. It seems that gender stereotypes or roles for drivers are not strictly dened in Arab Gulf trafc. Female drivers seem to commit as much violations and errors as male drivers. The reasons behind this freedom of expression for every one should be examined. Besides, none of the DBQ factors was related to accident involvement in the UAE. However, errors, lapses, and aggression-speeding violations factors predicted accident involvement in Qatar. In particular, errors appeared as the most critical factor in accident involvement. Therefore, the sources of the possible reasons of driver errors (e.g., the possible reasons for failure or departure from planned actions) should be investigated in Qatari trafc. Moreover, the problems of the applicability of the DBQ among Emirati drivers should be further examined by applying the DBQ among different driver groups. In addition, the reasons behind the differences in driver behaviour and trafc safety between countries should be investigated. 5. Methodological limitations The data were based solely on drivers self-reports of behaviour and no observations were made. However, several studies have indicated that self-reports of driving correspond well to actual driving behaviour. For instance, Ingham (1991) found high correlations between recorded driving on a 40 km test route and self-reported driver behaviour. It is also possible that some respondents embellished their answers about positive and aggressive driving. It should be noted, however, the respondents completed the questionnaires anonymously and could not gain anything by giving embellished responses. In their experimental study about the DBQ and social desirability bias, Lajunen and Summala (2003) concluded that the bias caused by socially desirable responding is very small in the DBQ responses. The measurement of accident involvement was based on a self-report of all past accidents. Simply because of forgetting to report some accidents (Maycock et al., 1991), some respondents may have underestimated the number of accidents in which they had been involved. The target group of the present study was local drivers both in Qatar and United Arab Emirates. Although expatriates might represent large proportion of the driver population in these multinational countries, the present study was restricted to Arab drivers because of difculties in data collection from about one hundred twenty different nationalities living in the area and because of other methodological issues (e.g., difculties to obtain representative samples). It is likely that behaviour of drivers originating from different countries might still be different even though they are sharing the same trafc environment (see Ozkan, 2006). Therefore, the interaction between local drivers and expatriates in trafc should be taken into account while interpreting the results of the present study. Finally, the DBQ is a measure for aberrant driver behaviour, therefore, by its nature, the DBQ item scores have not been normally distributed in any DBQ article in the literature. This feature of the scale should be taken into account in statistical analyses and specic statistical techniques (i.e., logistic regression) should be used. Acknowledgements This project was nancially funded by the Qatar Red Crescent Society, and supported by Hamad Medical Corporation, Doha, State

Note: Clasication rate (CR) of the cases were lower than chance accuracy rate (CAR) both in Qatar (CR: 84.2% and CAR: 86.7%) and UAE (CR: 56.6% and CAR: 63.2%).

(e.g., cognitive mechanisms, attention etc.) for shaping the factor structure of the DBQ. For instance, Dubai (very congested network in UAE) might have more congested trafc compared to Al Ain in UAE (see Lajunen et al., 1999). In addition, trafc laws and their application (i.e., trafc nes) in Arab Gulf countries might not be as standardized as in the European countries. Besides, trafc laws might vary within in one Arab Gulf country (i.e., UAE) because of its federal state structure but not in the another Arab Gulf country (i.e., Qatar). It was also the rst time that the DBQ factors were found to be related (although weakly) to accident involvement among Emirati drivers. It seems that Emirati drivers just drive without being aware of the differences between the types of driver behaviours and their possible consequences in trafc. Thus, the differentiation of errors, lapses, and violations did not emerge among Emirati drivers well. Consistent with the tendency found in Ozkan et al. (2006b) study, the DBQ factor structure might get more blurred while the frequency of the behaviours gets higher when going far from West/North to East/South. For example, it was found that Arab Gulf drivers, especially drivers in UAE, scored much higher on the DBQ items than Western/Northern European drivers (e.g., Blockey and Hartley, 1995; Lajunen et al., 2004; Reason et al., 1990) and even higher than Southern European/Middle Eastern drivers (e.g., Ozkan et al., 2006b). The results of the factor analysis showed that lapses items loaded mostly on errors factor, which was a mixture of lapses and errors and some violation items. Similar to Spanish factor structure (Gras et al., 2006), the factor structure did not support the original division to aggressive and ordinary violations either. The content of the factor was rather based on mainly hurrying up and tailgating or pushing among Arab Gulf drivers. The reliabilities scores were also, in general, lower than in the original British data and, for example, aggression-speeding factor was highly unreliable in the UAE sample. It might be possible to nd culture-specic items for higher reliability scores in Arab Gulf countries. On the other hand, it also likely that scaling or ways of response to items might vary from Western societies to Arab Gulf countries. Therefore, theoretical, methodological and practical reasons for the low reliability scores should be investigated in a further study. In contrast to general ndings in literature that women tend to commit violations less frequently than men and commit more

A. Bener et al. / Accident Analysis and Prevention 40 (2008) 14111417

1417

of Qatar. We greatly acknowledge their help for execution of this work and manuscript. References
Aberg, L., Rimmo, P.-A., 1998. Dimensions of aberrant driver behaviour. Ergonomics 41, 3956. Bener, A., Abu-Zidan, F.M., Bensiali, A.K., Al-Mulla, A.A.K., Jadaan, K.S., 2003. Strategy to improve road safety in developing countries. Saudi Med. J. 24, 603608. Bener, A., Crundall, D., Haigney, D., Bensiali, A.K., Al-Falasi, A.S., 2007. Driving behaviour stress, error and violations on the road: a cross-cultural comparisons study. Adv. Transportation Stud. 12, 514. Blockey, P.N., Hartley, L.R., 1995. Aberrant driving behaviour: errors and violations. Ergonomics 38, 17591771. Elander, J., West, R., French, D., 1993. Behavioral correlates of individual differences in road trafc crash risk: an examination of methods and ndings. Psychol. Bull. 113, 279294. Gras, M.E., Sullman, M.J.M., Cunill, M., Planes, M., Aymerich, M., Font-Mayolas, S., 2006. Spanish drivers and their aberrant driving behaviours. Transport. Res. Part F: Trafc Psychol. Behav. 9, 129137. Ingham, R., 1991. The young driver. In: Paper presented at the new insights into driver behaviour conference organized by PACTS (Parliamentary Advisory Council for Transport Safety), London, 21 October 1991. International Road Federation IRF, 2003. Transportation Stat., Brussels, Belgium. Lajunen, T., Summala, H., 2003. Can we trust self-reports of driving? Effects of impression management on driver behaviour questionnaire responses. Transportation Res. Part F 6, 97107. Lajunen, T., Parker, D., Summala, H., 1999. Does trafc congestion increase aggression? Transportation Res. Part F: Trafc Psychol. Behav. 2, 225236. Lajunen, T., Parker, D., Summala, H., 2004. The Manchester Driver Behaviour Questionnaire: a cross-cultural study. Accid. Anal. Prev. 36, 231238.

Lawton, R., Parker, D., Manstead, A.S.R., Stradling, S., 1997. The role of affect in predicting social behaviours: the case of road trafc violations. J. Appl. Soc. Psychol. 27, 12581276. Lewin, I., 1982. Driver training: A perceptual-motor skill approach. Ergonomics 25, 917924. Maycock, J., Lockwood, C.R., Lester, J.F., 1991. The accident liability of car drivers. (Rep. No. 315). Transport and Road Research Laboratory, Crowthorne, England. Mesken, J., Lajunen, T., Summala, H., 2002. Interpersonal violations, speeding violations and their relation to accident involvement in Finland. Ergonomics 45, 469483. Norusis, M.J., 1998. SPSS/PC+ for Windows, Base System and Advanced Statistics Users Guide, Window Version 12.0, SPSS, Inc., Chicago, IL. Ozkan, T., 2006. The Regional Differences Between Countries in Trafc Safety: A Cross-Cultural Study and Turkish Case. University of Helsinki, Helsinki, Finland, http://ethesis.helsinki./julkaisut/kay/psyko/vk/ozkan/theregio.pdf Ozkan, T., Lajunen, T., Summala, H., 2006a. Driver Behaviour Questionnaire: A followup study. Accid. Anal. Prev. 38, 386395. Ozkan, T., Lajunen, T., Chliaoutakis, J., Parker, D., Summala, H., 2006b. Cross-cultural differences in driving behaviours: a comparison of six countries. Transportation Res. Part F 9, 227242. Parker, D., McDonald, L., Rabbitt, P., Sutcliffe, P., 2000. Elderly drivers and their accidents: the aging driver questionnaire. Accid. Anal. Prev. 32, 751759. Parker, D., Reason, J.T., Manstead, A.S.R., Stradling, S.G., 1995a. Driving errors, driving violations and accident involvement. Ergonomics 38, 10361048. Parker, D., West, R., Stradling, S.G., Manstead, A.S.R., 1995b. Behavioural characteristics and involvement in different types of trafc accident. Accid. Anal. Prev. 27, 571581. Reason, J.T., 1990. Human Error. Cambridge University Press, New York. Reason, J.T., Manstead, A.S.R., Stradling, S., Baxter, J., Campbell, K., 1990. Errors and violations on the roads. Ergonomics 33, 13151332. Sullman, M.J.M., Meadows, M.L., Pajo, K.B., 2002. Aberrant driving behaviours amongst New Zealand truck drivers. Transportation Res. Part F 5, 217232.