Accident Analysis and Prevention 38 (2006) 1011–1018
Cross-cultural differences in driving skills: A comparison of six countries a , Timo Lajunen b,∗ , Joannes El. Chliaoutakis c , ¨ T¨urker Ozkan Dianne Parker d , Heikki Summala a b
a Traffic Research Unit, Department of Psychology, P.O. Box 9, 00014 University of Helsinki, Finland ¨ Ankara, Turkey Safety Research Unit, Department of Psychology, Middle East Technical University (ODTU), c Department of Social Work, Technological Educational Institute (TEI) of Crete, Heraklion, Greece d Department of Psychology, University of Manchester, UK
Received 14 July 2005; received in revised form 2 January 2006; accepted 5 April 2006
Abstract The first aim of the present study was to investigate the applicability of the two-factor structure (perceptual-motor skills by 11 items, e.g., “fluent driving”; safety skills by 9 items, e.g., “conforming to the speed limits”) of the Driver Skill Inventory (DSI) among British, Dutch, Finnish, Greek, Iranian, and Turkish drivers. It was also hypothesized that the combination of self reported high ratings of perceptual-motor skills and low ratings of safety skills creates a serious risk for dangerous driving and road accident involvement. The second aim of this study was, therefore, to investigate this asymmetric relationship between perceptual-motor and safety skills in traffic penalties and accident involvement. Two hundred and forty two drivers were chosen from each of the six countries, matched for age and sex. The results of exploratory factor analyses together with target rotation showed that the two-factor structure of DSI found in “safe” Northern and Western European countries were highly congruent. However, the safety skills factor of DSI in Greece, Iran, and Turkey was relatively incongruent in spite of high factor similarity found in perceptual-motor skills. The asymmetric relationship between perceptual-motor and safety skills on traffic penalties was found in Finland and Turkey. A negative relationship between safety skills and the number of accidents was found both in Greece and Iran while a positive relationship between perceptual-motor skills and the number of accidents was found only in Iran. © 2006 Elsevier Ltd. All rights reserved. Keywords: Driver Skill Inventory; Cross-cultural differences; Perceptual-motor skills; Safety skills; Traffic accidents; Traffic penalties
1. Introduction It has been estimated that annually one million people die in road traffic accidents in the world (Elvik and Vaa, 2004). Although a road traffic accident is one of the most important public health problems everywhere, there are considerable regional differences between countries. According to accident statistics, 9.1 Finns, 8.9 Dutch, and 7.5 Britons per one billion vehicle kilometres were killed in traffic accidents in 2002 whereas the corresponding figure for Greeks and Turks were 26.7 and 73, respectively (IRTAD, 2003). However, only a few cross-cultural studies about traffic safety and behaviour have been conducted so far. Human error is a sole or a main contributory reason for most road traffic accidents (Lewin, 1982). Human factors in driving
∗
Corresponding author. Tel.: +90 312 2105113; fax: +90 312 2103154. E-mail address:
[email protected] (T. Lajunen).
0001-4575/$ – see front matter © 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.aap.2006.04.006
can be seen as being composed of two separate components, driving style and driving skills (Elander et al., 1993). Driving style refers to the ways drivers choose to drive or habitually drive (e.g., the choice of driving speed). Driving skills include information processing and motor skills, which improve with practice and training (Elander et al., 1993). Spolander (1983) made a distinction between technical and defensive driving skills and developed a self-report instrument for measuring these dimensions. According to this distinction, technical driving skills include quick and fluent car control and traffic situation management, while defensive driving skills consist of anticipatory accident avoidance skills. In Spolander’s (1983) instrument, drivers were asked to compare themselves with “an average driver” in 13 aspects of driving. Later, Hatakka et al. (1992) replaced this external reference with an internal one in which drivers were asked to assess their own abilities in different aspects of driving skills. Lajunen and Summala (1995) developed the Driver Skill Inventory (DSI) further by extending the contents of the instrument and verifying the two-factor
1012
¨ T. Ozkan et al. / Accident Analysis and Prevention 38 (2006) 1011–1018
structure of DSI as perceptual-motor and safety skills by using factor analysis. A consistent factor structure and high reliability of the DSI was obtained in very different populations (e.g., among male traffic criminals, male policemen, and male traffic instructor candidates by Summala and Hyv´en, 1990). Later, the English version of the DSI was used in Australia (Lajunen et al., 1998a) and in the UK (Lajunen et al., 1998b). However, the DSI has not been validated in countries such as Iran, Greece, and Turkey, which have worse safety records than Scandinavian and AngloAmerican countries. The first aim of the present study was, therefore, to investigate the suitability of the two-factor structure and cross-cultural applicability of the DSI in different countries representing different levels of traffic safety and motorization. A well-known fact is that there are considerable differences ¨ between countries in driving style (see Ozkan et al., 2006). Earlier studies have shown that there are differences among nationalities in driver self-assessment (Sivak et al., 1989a), in risk perception (Sivak, 1987), and the target risk-level of performance (Sivak et al., 1989b). According to Brown and Groeger (1988), a combination of two main inputs, (1) information on potential hazards in the traffic environment and (2) information on the joint capacity of driver and vehicle to prevent potential hazard, influence a driver’s risk perception. It can be supposed that a driver’s view of his/her perceptual-motor and safety skills would influence his/her risk perception and the target risk-level, which, in turn, affect his/her driving style including risk-taking while driving. Studies using the DSI have revealed that both perceptual-motor and safety skills predict the outcome variables directly related to accident risks. However, as compared to perceptual-motor skills, safety skills seemed to be more strongly associated with the correlates of accident risks, e.g., maximum speed (e.g., Lajunen and Summala, 1995). Lajunen et al. (1998a) demonstrated that some minor crosscultural differences were observed in driving skills. Finnish drivers, for instance, are more safety-oriented than Australian drivers. Age correlated significantly with safety skills, but not with perceptual-motor skills in the Australian sample whereas age correlated significantly with perceptual-motor skills, but not with safety skills, in the Finnish sample. Although there are some variations across cultures and countries in self-assessments of perceptual-motor and safety skills, earlier results indicated that there are some similarities within countries in driving skills too. Male drivers, for instance, consistently overestimated their driving skills as compared to female drivers (e.g., McKenna et al., 1991). In addition, it was found that driving experience was associated with confidence in one’s own driving skills, but negatively related to concern for safety (Lajunen and Summala, 1995). Besides, drivers tend to overestimate their perceptual-motor and safety skills compared to the “average driver” (McKenna et al., 1991). This over estimation of driving skills has been found in many countries including Finland (N¨aa¨ t¨anen and Summala, 1976). Overconfidence in perceptual-motor and safety skills, in turn, results in a biased risk assessment and a driver’s sense of control in traffic leading to high levels of risk acceptance (N¨aa¨ t¨anen and Summala, 1976). Since a driver’s sense of control and overcon-
fidence in his/her capabilities increases with driving experience, these factors constitute a real risk factor in traffic where safety skills, e.g., proper alertness and anticipation of possible risks, are essential for safety (see N¨aa¨ t¨anen and Summala, 1976). It is, therefore, possible that not only the perceptual-motor skill evaluation in relation to the average driver, but also the driver’s internal balance between perceptual-motor and safety skills is important for safe driving (Lajunen et al., 1998a). Earlier studies have indicated that driving skills were positively associated with the number of accidents, penalties and the level of speed, while safety skills were negative associated with these variables (Lajunen et al., 1998b). S¨umer et al. (in press) tested the asymmetric link between perceptual-motor and safety skills among Turkish drivers. The results revealed that those who reported a low level of safety skills but a high level of perceptual-motor skills reported the highest levels of accidents and penalties. However, the asymmetric relationship between perceptual-motor and safety skills on the outcome variables has not been investigated in other traffic cultures. Since social behaviour, cognitive processes, and attitudes are influenced by cultural background (Berry et al., 1992), the drivers’ view of their driving skills might vary from country to country. This in turn might change the association between perceptual-motor and safety skills and accident involvement and penalties in different traffic cultures. This study was therefore conducted to investigate the asymmetric relationship between perceptual-motor and safety skills among British, Dutch, Finnish, Greek, Iranian, and Turkish drivers.
2. Method 2.1. Participants and procedure The participants from Finland (1120), Great Britain (840), and the Netherlands (700) were obtained from data colleted in a previous study (see Lajunen et al., 2004, for detailed information). The Turkish data reported in this study initially consisted of four previous data sets including more than two thousand drivers across different age groups (see, for instance, S¨umer et al., in press, for information about the data collection procedure). In Greece (Crete), stratified random sampling using information from the National Statistics Department was performed to collect data from 342 participants in the targeted households (see Chliaoutakis et al., 2005, for detailed information). In Iran, two research assistants trained in data collection approached drivers and only those who agreed to fill out the questionnaire were included in the study. Research assistants also used their own social network to find eligible drivers. They collected data from 311 participants, of which, 242 drivers were then eligible to be included in the present study. Since the Iranian sample was smaller compared to the other datasets, a total number of 242 drivers (matched by age and sex) were sampled from each country. In all studies, participation was voluntary but having a driving licence was obligatory. Participants were assured of anonymity and confidentiality.
¨ T. Ozkan et al. / Accident Analysis and Prevention 38 (2006) 1011–1018
1013
Table 1 Sample characteristics Variables Total sample
N Age Mean S.D.
Male sample
Female sample
FIN
GB
GR
IRN
NL
TR
FIN
GB
GR
IRN
NL
TR
FIN
GB
GR
IRN
NL
TR
242
242
242
242
242
242
146
146
146
146
146
146
96
96
96
96
96
96
29.69 10.84
29.48 11.10
30.47 9.71
29.50 10.63
32.25 8.26
30.03 10.36
31.24 11.81
29.73 11.93
31.06 10.66
31.00 11.47
33.06 8.81
31.08 27.34 29.09 29.57 27.20 31.02 28.42 11.39 8.70 9.76 8.01 8.79 7.22 8.37
Driving experience Mean 10.96 10.52 S.D. 9.74 9.69
8.84 8.08
9.00 9.36
12.00 7.85
8.77 8.37
12.70 10.60
11.28 10.78
10.42 9.04
10.61 9.99
13.11 8.47
10.18 9.39
Annual mileage (× 1000 km) Mean 22.21 10.99 87.18 S.D. 17.76 9.87 97.29
45.77 78.78
25.99 32.92
13.90 19.13
26.97 19.68
12.56 103.51 10.75 106.18
63.17 90.00
30.79 38.52
17.01 14.96 21.52 11.03
8.33 7.58
9.35 7.64
6.44 5.61
6.52 10.31 7.70 6.49
6.63 5.97
8.59 62.35 19.31 18.69 9.17 7.81 75.97 46.98 18.97 13.58
Total penalties Mean 1.46 S.D. 2.64
0.60 1.21
1.73 3.04
1.38 2.26
1.67 2.60
1.39 1.99
1.83 3.03
0.69 1.32
2.06 3.51
1.74 2.76
1.92 2.72
1.60 2.14
0.90 1.77
0.46 1.02
1.21 2.06
0.84 0.88
1.30 2.38
1.07 1.71
Total accidents Mean 0.38 S.D. 0.80
0.51 0.79
1.55 2.07
2.68 3.28
0.55 0.97
1.01 1.34
0.42 0.86
0.50 0.82
1.79 2.31
3.16 3.63
0.64 1.05
1.05 1.40
0.34 0.71
0.53 0.74
1.20 1.59
1.95 2.53
0.43 0.83
0.96 1.24
2.2. Measures 2.2.1. Driver Skill Inventory (DSI) The Driver Skill Inventory (DSI) is a 20 item self-reported measure of perceptual motor (11 items; e.g., fluent driving) and safety skills (9 items; e.g., conforming to the speed limits) (Lajunen and Summala, 1995). DSI was previously translated into English and had been shown to have good reliability and predictive validity in Finland and Australia (Lajunen et al., 1998a; Lajunen and Summala, 1995). The DSI was recently used in Turkey with high reliability coefficients (e.g., see S¨umer et al., in press, for DSI items). In the present study, British, Dutch, Finnish, and Turkish drivers were asked to rate how weak or strong they feel they were in given skills by using 5point scales (0 = very weak, 4 = very strong) whereas Greek and Iranian drivers evaluated each item on 4-point scales (1 = very weak, 4 = very strong). 2.2.2. Demographic measures Respondents answered questions about their age, sex, the number of accidents, and penalties during last 3 years, the number of years a full driving license had been held, and annual mileage. Characteristics for the whole sample as well as for male and female drivers in each country are presented in Table 1. Dutch drivers were older than drivers in other countries. Greek sample had higher annual mileage and traffic penalties than drivers in other countries. Iranian drivers had higher number of traffic accidents than drivers in other countries. 2.3. Statistical analyses The equivalence of the two-factor solution of DSI in six countries were assessed by comparing the rotated factor matrices by using Procrustes target rotation techniques and factorial agreement coefficients. In the present study, the Finnish drivers were
used as a target group and separate principal axis factoring with varimax rotation was run for each sample before calculating factorial agreement coefficients (see the Lajunen et al., 2004 for detailed information). In general, factorial agreement coefficients higher than 0.95 indicate factorial similarity, whereas values lower than 0.90 (van de Vijver and Leung, 1997) or 0.85 (ten Berge, 1986) are taken as a sign of non-negligible incongruities. The data were analysed by using reliability analyses, Pearson product–moment correlations and moderated regression analyses. In each of the moderated regression analyses, age, sex, and annual mileage were entered in the first step. Perceptualmotor and safety skills were entered in the second step, and their interactions were entered in the third step. Moderator and independent variable scores were first centred (i.e., the mean was subtracted from each individual score) and then an interaction term (i.e., scores multiplied by perceptual-motor and safety skill score) was calculated before the analyses (see Aiken and West, 1991).
3. Results 3.1. Target rotation and agreement coefficients The values of these indexes are listed in Table 2. It shows that the most widely applied index, i.e., coefficient of proportionality (also known as Tucker’s phi), and the most stringent index (identity coefficient) both indicated full identity between the Finnish and all other samples on the perceptual-motor skills factor of DSI. However, the safety skills factor of DSI in Greece, Iran, and Turkey was relatively incongruent. It seems that the content of the safety skills factor in these three countries differ from the content of the safety skills factor in Great Britain, Finland, and the Netherlands.
¨ T. Ozkan et al. / Accident Analysis and Prevention 38 (2006) 1011–1018
1014
Table 2 Four identity indexes for three factor solutions of the DSI in five countries DSI factors
Countries
Identity coefficient
Additivity coefficient
Proportionality coefficient
Correlation coefficient
Perceptual-motor skills
GB GR IRN NL TR
0.97 0.90 0.93 0.96 0.92
0.93 0.73 0.84 0.92 0.83
0.95 0.90 0.95 0.97 0.92
0.93 0.84 0.90 0.93 0.84
Safety skills
GB GR IRN NL TR
0.97 0.73 0.79 0.93 0.76
0.95 0.54 0.64 0.89 0.60
0.97 0.74 0.79 0.94 0.76
0.95 0.56 0.65 0.91 0.60
Table 3 Alpha reliability coefficients of the DSI scales in six countries DSI subscales
FIN
GB
GR
IRN
NL
TR
Perceptual-motor skills (11 items) Safety skills (9 items)
0.88 0.81
0.85 0.80
0.79 0.73
0.79 0.71
0.85 0.73
0.88 0.82
3.2. Reliability analysis Alpha reliability coefficients for the DSI scales for six nationalities are listed in Table 3. In all six samples, the “perceptualmotor” scale seemed to be internally consistent (α = 0.79–0.88). The “safety skills” scale also had fairly acceptable alpha values (α = 0.71–0.82). 3.3. Correlates of DSI scales Table 4 lists correlations between background variables, perceptual-motor and safety skills, and the number of traffic penalties and accidents in the six countries. Age, sex, and annual mileage correlated with the number of traffic penalties and accidents. Age was negatively related to the number of traffic accidents in Finland and Great Britain whereas it was positively related to the number of traffic accidents in Greece and Iran. Being male was related to a higher number of traffic accidents in Greece and Iran. A high number of self-reported penalties was related to accidents in every country except Great Britain. Annual mileage was positively associated with either the number of traffic accidents or penalties in every country except Turkey. Age was positively related to safety skills while annual mileage was positively associated with perceptual-motor skills. Also, being female was related to a low level of self-reported perceptual-motor skills in every country. Therefore, the effects of these demographic variables were later statistically controlled in the regression analyses by forcing them into the model in the first step. Self-reported accident involvement was positively related to perceptual-motor skills only in Iran, but negatively related to safety skills in Finland and the Netherlands. The number of penalties correlated positively with perceptual-motor skills in Finland, Greece, and the Netherlands whereas it correlated negatively with safety skills in every country except Greece.
Fig. 1. The interaction between safety skills and perceptual-motor skills on traffic penalties in Finland.
3.4. Hierarchical regression analyses 3.4.1. Traffic penalties As presented in Table 5, age was negatively related to the number of penalties in Finland and Great Britain but positively related to the number of penalties in Greece. Sex (being female) was negatively related to the number of penalties only in Iran. Annual mileage was also positively related to the number of penalties in Finland and Great Britain. Perceptual-motor skills were positively related to the number of penalties in Finland, Greece, and the Netherlands. However, safety skills were negatively related to the number of penalties in every country. The statistically significant interaction effect between perceptual-motor and safety skills on the number of penalties was found only in Finland and Turkey. SIMPLE syntax program developed by O’Connor (1998) was used to plot statistically significant interactions. Significant interactions were plotted by generating simple regression equations of a given outcome (dependent) variable at low (i.e., 1 standard deviation below the mean), moderate (mean), and high (i.e., 1 standard deviation above the mean) levels of perceptual-motor and safety skills (cf. Aiken and West, 1991). Fig. 1 shows that the relationship between perceptual-motor skills and traffic penalties
Table 4 Correlations among DSI scales, demographic variables, and the number of accidents and penalties 1
2
1. Age 2. Sex (1 = male, 2 = female) 3. Annual mileage 4. Perceptual-motor skills 5. Safety skills 6. Total penalties 7. Total accidents
GB −0.03 0.15a 0.10 0.25c −0.14a −0.14a
GB −0.19b −0.33c 0.20c −0.10 0.02
1. Age 2. Sex (1 = male, 2 = female) 3. Annual mileage 4. Perceptual-motor skills 5. Safety skills 6. Total penalties 7. Total accidents
GR −0.07 0.45c 0.09 0.33c 0.27c 0.26c
GR −0.21c −0.30c 0.01 −0.14a −0.14a
a b c
3
GB 0.25c −0.17b 0.13a 0.05
GR 0.13a 0.13a 0.20b 0.16a
4
GB −0.00 0.09 0.07
GR 0.41c 0.19b 0.10
5
GB −0.25c −0.09
GR −0.05 −0.01
6
1
2
GB 0.07
FIN −0.17b 0.12 0.12 0.23c −0.12 −0.21c
FIN −0.33c −0.61c 0.11 −0.17b −0.04
GR 0.31c
IRN −0.17b 0.53c 0.05 0.25c 0.09 0.40c
IRN −0.27c −0.27c 0.03 −0.20b −0.18b
3
FIN 0.38c −0.14a 0.27c 0.18b
IRN 0.20b 0.10 0.13a 0.44c
4
FIN −0.08 0.28c 0.03
IRN 0.25c 0.01 0.18b
5
FIN −0.31c −0.19b
IRN −0.13a −0.01
6
1
2
3
4
5
6
FIN 0.24c
NL −0.12 0.11 −0.04 0.21c −0.01 −0.09
NL −0.18b −0.39c 0.05 −0.12 −0.11
NL 0.15a −0.05 0.10 0.19b
NL 0.06 0.23c 0.09
NL −0.14a −0.13a
NL 0.22c
IRN 0.27c
TR −0.13a 0.16b 0.15a 0.19b −0.02 −0.02
TR −0.20b −0.36c 0.17b −0.13a −0.03
TR 0.29c −0.01 0.07 0.10
TR 0.23c 0.04 0.00
TR −0.24c −0.12
TR 0.43c
p < 0.05. p < 0.01. p < 0.001.
Table 5 Moderated hierarchical regression analyses on the number of penalties Variables
FIN
GB
GR
NL
IRN
R2
R2
F
R2
F
R2
R2
F
0.052
4.24b
0.093
8.11c
0.045
3.70a
1. Step Age Sex (1 = M, 2 = F) Annual mileage
0.111
2. Step Perceptual-motor Safety
0.086
3. Step Perceptual-motor × safety
0.035
Total
R2
F
Beta
9.89c
−0.17b −0.07 0.15a
−0.17b −0.12 0.26c 11.55c
0.040
4.67c
0.22b 0.00 −0.19c 0.231
0.072
3.88c
0.004
0.022
F
9.30c
0.169
Beta
1.76 −0.03 −0.10 0.09
0.061
4.25c 0.23c
−0.27c
−0.16a
7.97c
0.006
3.81c
R2
F
0.020
1.63
3.41b
0.000
0.051
2.84a
3.61b 0.06 −0.25c
0.017 −0.02
0.068
Beta −0.04 −0.12 0.05
−0.01 −0.15a
−0.08 0.089
Beta 0.03 −0.17b 0.07
0.023
0.26c
−0.07
0.00 0.092
Beta 0.23c −0.10 0.07
0.07 −0.21b
−0.25c 11.79c
Beta
TR
¨ T. Ozkan et al. / Accident Analysis and Prevention 38 (2006) 1011–1018
Variables
3.80c −0.14a
0.088
d.f. for F-tests: 1. step = 3,239; 2. step = 5,237; 3. step = 6,236. a p < 0.05. b p < 0.01. c p < 0.001. 1015
¨ T. Ozkan et al. / Accident Analysis and Prevention 38 (2006) 1011–1018
0.004
0.271 0.02
varied for drivers with high, moderate and low levels of safety skills in Finland. The simple slope analysis indicates that drivers with low levels of safety skills reported the highest number of traffic penalties (t(238) = 5.49, p < 0.001). It was also significant for those with moderate levels of safety skills (t(238) = 4.48, p < 0.001), but not for high levels of safety skills. Fig. 2 shows that the relationship between perceptual-motor and traffic penalties varied for drivers with high, moderate and low levels of safety skills in Turkey. The simple slope analysis indicates that drivers with low levels of safety skills reported the highest number of traffic penalties (t(238) = 2.38, p < 0.05), but not for moderate and high levels of safety skills. It can be concluded, therefore, that (over)confidence in one’s perceptualmotor skills should be buffered by high emphasis on safety skills to avoid dangerous driving (e.g., speeding).
Total R2
0.099
0.01
−0.06 −0.11
d.f. for F-tests: 1. step = 3,239; 2. step = 5,237; 3. step = 6,236. a p < 0.05. b p < 0.01. c p < 0.001.
4.29c 0.000 3. Step Perceptual-motor × safety
0.037
0.000
1.49
−0.02
0.103
0.003
4.49c
−0.06
0.066
0.000
0.010 0.11 −0.16a 5.25c 0.019 0.09 −0.05 1.77 0.014 2. Step Perceptual-motor Safety
0.009 5.17c
0.085 1. Step Age Sex (1 = M, 2 = F) Annual mileage
Fig. 2. The interaction between safety skills and perceptual-motor skills on traffic penalties in Turkey.
2.77a
3.33b
0.05 −0.10
−0.07
0.24c −0.06 0.29c 2.25 7.34c
Beta
−0.23c −0.02 0.20b
F
0.028
0.030
14.52c
0.032
0.005
1.18
−0.07
0.02 −0.12 1.19 17.13c
0.235 −0.12 −0.09 0.19b 0.081 −0.16a 0.03 0.08
Beta
0.23c −0.11 0.03 7.01c
0.056
4.66c
F Beta R2 R2
F
0.15a −0.16b
24.31c
Beta Beta R2 R2
F NL GR GB FIN Variables
Table 6 Moderated hierarchical regression analyses on the number of accidents
0.013
0.96
R2 F R2
0.012
TR IRN
F
Beta
−0.04 −0.02 0.10
1016
3.4.2. Traffic accidents As presented in Table 6, age was negatively related to the number of accidents in Finland and Great Britain, whereas age was positively related to the number of traffic accidents in Greece and Iran. Annual mileage was positively related to the number of accidents in Finland, Iran, and the Netherlands. Perceptual-motor skills were positively related to the number of traffic accidents only in Iran whereas safety skills were negatively related to the number of traffic accidents both in Greece and Iran. However, there was no interaction effect between perceptual-motor and safety skills in the number of traffic accidents. 4. Discussion The two-factor structure of the DSI was found to be fairly consistent and had acceptable reliability coefficients in six countries included in this study. The present study clearly suggests that DSI is a viable instrument for measuring drivers’ self-assessment of their perceptual-motor and safety skills in different traffic cultures. It should be, however, noted that the factorial agreement was unsatisfactory for safety skills in
¨ T. Ozkan et al. / Accident Analysis and Prevention 38 (2006) 1011–1018
Greece, Iran, and Turkey. The safety skills include aspects of social tolerance and traffic rule compliance. This distinction is parallel with Lawton et al.’s (1997) distinction of violation types. According to their distinction, violations can be divided into two categories. The first, named ordinary violations, involves deliberately breaking the Highway Code and/or the law. The second involves interpersonal aggressive violations. Recently, it was found that there were significant differences between Western/Northern European (British, Finnish, and Dutch) and Southern European/Middle Eastern (Greek, Iranian, ¨ and Turkish) drivers especially in aggressive violations (Ozkan et al., 2006). The Southern Europeans/Middle Easterners reported committing more aggressive violations compared to ¨ Northern/Western Europeans (Ozkan et al., 2006). It seems that the lack of social tolerance and interpersonal aggressive violations are important components of driving in Greece, Iran, and Turkey. It can be, therefore, supposed that the content of the safety skills factor of DSI does not exclusively include the safety skills in these countries. It might vary from country to country because of the high number of potential interpersonal conflicts areas in these counties’ traffic environments. On the other hand, driving has the same requirements of vehicle handling everywhere, consisting of “pan-cultural” perceptual-motor skills. Consistent with the previous studies (Lajunen and Summala, 1995; Lajunen et al., 1998b), age correlated significantly with safety skills across all countries in the present study. It seems that as drivers become older, the more social tolerance and ruleabiding approach they have. Similarly, it was found that being male was significantly associated with self-reported perceptualmotor skills. These results support partly the general finding in literature that greater accident-risk and more deviant driving styles are related to being male and young (Elander et al., 1993). In addition, annual mileage correlated significantly with self-reported perceptual-motor skills. N¨aa¨ t¨anen and Summala (1976) proposed that increasing driving experience and exposure to traffic increases the level of driving skills, which, in turn, support the sense of subjective control and decreases the sense of subjective risk and concern for safety aspects while driving. It can, therefore, be suggested that driver training should also give drivers a realistic view of their perceptual-motor skills. The only significant positive association between annual mileage and safety skills was found among Greek drivers. It seems that safety-oriented responses learnt in driving schools are forgotten quickly with driving experience due to selective learning and differential feedback (Lajunen and Summala, 1995; Lajunen et al., 1998a). The present study showed that safety skills were negatively associated with number of penalties in all countries whereas perceptual-motor skills were positively associated with the number of penalties in Finland, Greece, and the Netherlands. This result suggests that safety skills are highly related to law-abiding driving whereas a strong emphasis on perceptual-motor skills is associated with reckless driving (Lajunen et al., 1998b). The statistically significant interaction between perceptual-motor and safety skills on the number of penalties was found only in Finland and Turkey. This result suggests that high safety skills are essential for buffering the effect of a high level of perceptual-
1017
motor skills on reckless driving. The results of the present study indicated that perceptual-motor skills were positively associated with the number of accidents only in Iran while safety skills were negatively associated with the number of accidents both in Greece and Iran. Although the number of accidents and penalties are widely used as criteria for safe driving, they are rare events and statistically complex (see Elander et al., 1993). The results of the present study showed that the DSI factor structures are almost the same in different countries but there may still be differences in safety skills, especially in Southern Europe and Iran. It might therefore be better to develop “nationspecific safety skill items” to catch the local aspects of safety skills as well as include “core DSI items” for cross-national comparisons. The results of the present study also emphasize the importance of thorough knowledge of the countries and cultures involved, and the need for comprehensive examination of the traffic cultures before conclusions about cultural differences in driving can be drawn. 4.1. Implications of the present study The present study showed that safety and perceptual-motor skills are separate factors and that high levels of safety skills buffered the negative effects of perceptual-motor skills, especially on traffic penalties in some countries. These findings have serious implications for driver education and training as well as for safety campaigns in different traffic cultures. For instance, Gregersen (1996) argued that additional training on a specific driving skill might increase confidence especially among novice drivers. Overconfidence, in turn, results in a biased risk assessment, which can lead to high levels of risk acceptance (e.g., Matthews and Moran, 1986). Recent studies have shown that drivers tend to have a very pervasive “self-enhancement bias,” especially when they compare themselves to other drivers (e.g., Walton, 1999). Therefore, self-awareness of “real” driving skills should be enforced by giving adequate feedback during driving practice. Specifically, safety skills should be intensively incorporated into driving skills through driver education. In this way, development of the false sense of safety, overestimation of control over traffic situations (“unrealistic optimism” and “illusion of control”, McKenna, 1993), and indifference to messages of road safety campaigns (e.g., Ulleberg, 2002) might be prevented. 4.2. Limitations of the study The present study has some methodological limitations that should be taken into account. First, the driver samples were relatively small and might not represent the national driver population, especially in Turkey and in Iran. Second, the data were based solely on drivers’ self-reports. It is possible that some male respondents “embellished” their answers by reporting higher level of perceptual-motor skills and traffic violations (e.g., speeding tickets) whereas some female respondents might have underestimated their perceptual-motor skills. However, the respondents completed the questionnaires anonymously and could not gain anything by giving biased responses. In their quasi-experimental study about driver behaviours and social
1018
¨ T. Ozkan et al. / Accident Analysis and Prevention 38 (2006) 1011–1018
desirability bias, Lajunen and Summala (2003) concluded that the bias caused by socially desirable responding is very small. Third, the measurement of accident involvement was based on self-reports of all past accidents during the previous three years. Despite this time limit, some drivers might still have forgot to report their minor accidents, although the forgetting rate for the time period of three years can be expected to be small and a threeyear retrospective reporting period is widely used in literature. Fourth, two different response scales were used (4-point Likert scale in Iran and Greece versus 5-point Likert in other countries). It should be noted, however, that samples were not directly compared and, therefore, two different response scales could be used. Still, use of different scales might have a minor impact on correlation and regression coefficients within the countries. Fifth, this study emphasized the importance of special care when translating questionnaires, choosing the samples, and collecting data in different countries. Misinterpretation or inaccurate translation of the items, very different samples and data collection methods may lead to increase of measurement error and, in the worst case, biased results. Acknowledgements ¨ ˙ITAK (SBB-3038) and This work has been supported by TUB ¨ in Turkey and the Graduate School of Psychology and ODTU Henry Ford Foundation in Finland, and Marie Curie Transfer of Knowledge programme (“SAFEAST” project No: MTKDCT-2004-509813). We would like to thank Nihan Alptekin and Atoosa Sharafi for collecting the Iranian data. References Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and Interpreting Interactions. Sage Publications, Newbury Park, CA. Berry, J.W., Poortinga, Y.H., Segall, M.H., Dasen, P.R., 1992. Cross-Cultural Psychology: Research and Applications. Cambridge University Press, Cambridge. Brown, I.D., Groeger, J.A., 1988. Risk perception and decision taking during the transition from novice to experienced driver status. Ergonomics 31, 585–598. Chliaoutakis, J., Koukouli, S., Lajunen, T., 2005. Lifestyle patterns as predictors of driving behavior in urban areas of Greece. Transport. Res. F 8, 413–428. Elander, J., West, R., French, D., 1993. Behavioral correlates of individual differences in road traffic crash risk: an examination of methods and findings. Psychol. Bull. 113, 279–294. Elvik, R., Vaa, T., 2004. The Handbook of Road Safety Measures. Elsevier, Oxford. Gregersen, N.P., 1996. Young drivers’ overestimation of their own skill: an experiment on the relation between training strategy and skill. Accident Anal. Prev. 28, 243–250. Hatakka, M., Keskinen, E., Laapotti, S., Katila, A., Kiiski, H., 1992. Driver’s self-confidence—the cause of the effect of mileage. J. Traffic Med. 21, 313–315. IRTAD, 2003. Selected risk values for the year 2001. International Road Traffic and Accident Database: OECD. Available from: http://www.bast.de/ htdocs/fachthemen/irtad/english/we2.html.
Lewin, I., 1982. Driver training: a perceptual-motor skill approach. Ergonomics 25, 917–924. Lajunen, T., Corry, A., Summala, H., Hartley, L., 1998a. Cross-cultural differences in Drivers’ self-assessments of their perceptual-motor and safety skills: Australians and Finns. Pers. Indiv. Diff. 24, 539– 550. Lajunen, T., Parker, D., Stradling, S., 1998b. Dimensions of driver anger, aggressive and highway code violations and their mediation by safety orientation in UK drivers. Transport. Res. F 1, 107–121. Lajunen, T., Parker, D., Summala, H., 2004. The Manchester Driver Behaviour Questionnaire: a cross-cultural study. Accident Anal. Prev. 36, 231–238. Lajunen, T., Summala, H., 2003. Can we trust self-reports of driving? Effects of impression management on driver behaviour questionnaire responses. Transport. Res. F 6, 97–107. Lajunen, T., Summala, H., 1995. Driver experience, personality, and skill and safety motive dimensions in drivers’ self-assessments. Pers. Indiv. Diff. 19, 307–318. Lawton, R., Parker, D., Manstead, A.S.R., Stradling, S.G., 1997. The role of affect in predicting social behaviours: the case of road traffic violations. J. Appl. Soc. Psychol. 27, 1258–1276. Matthews, M.L., Moran, A.R., 1986. Age differences in male drivers’ perception of accident risk: the role of perceived driving ability. Accident Anal. Prev. 18 (4), 299–313. McKenna, F.P., 1993. It won’t happen to me: unrealistic optimism or illusion of control? Br. J. Psychol. 84, 39–50. McKenna, F.P., Stainer, R.A., Lewis, C., 1991. Factors underlying illusory self-assessment of driving skills in males and females. Accident Anal. Prev. 23, 45–52. N¨aa¨ t¨anen, R., Summala, H., 1976. Road User Behavior and Traffic Accidents. North-Holland/American Elsevier, Amsterdam/New York. O’Connor, B.P., 1998. SIMPLE: All-in-one programs for exploring interactions in moderated multiple regression. Educ. Psychol. Meas. 58, 836–840. ¨ Ozkan, T., Lajunen, T., Chliaoutakis, J., Parker, D., Summala, H., 2006. Cross-cultural differences on driving behaviours: a comparison of six countries. Transportation Research Part F 9, 227–242. Sivak, M., 1987. Human factors and road safety. Appl. Ergonom. 18, 289–296. Sivak, M., Soler, J., Tr¨ankle, U., 1989a. Cross-cultural differences in driver self-assessment. Accident Anal. Prev. 4, 371–376. Sivak, M., Soler, J., Tr¨ankle, U., 1989b. Cross-cultural differences in driver risk-taking. Accident Anal. Prev. 4, 363–370. Spolander, K., 1983. Bilf¨orares uppfatting om egen k¨orf¨orm˚aga (Drivers’ assessment of their own driving ability). Report No.: 252. Swedish Road and Traffic Research Institute, Link¨oping. Summala, H., Hyv´en, A., 1990. Measuring (safety-related) driver skills in the laboratory: towards a new test for the driver examination (Report No. 21). Traffic Research Unit, University of Helsinki, Helsinki. ¨ S¨umer, N., Ozkan, T., Lajunen, T. Asymmetric relationship between driving and safety skills: a test on turkish drivers. Accident Anal. Prev., in press. ten Berge, J.M.F., 1986. Rotatie naar perfecte congruentie en de Multipele roep Methode. Netherlands Tijdschrift voor Psychologie 41, 218– 225. Ulleberg, P., 2002. Personality subtypes of young drivers. Relationship to risk-taking preferences, accident involvement, and response to a traffic safety campaign. Transport. Res. F 4, 279–297. van de Vijver, F., Leung, K., 1997. Methods and Data Analysis for Crosscultural Research. Sage, London. Walton, D., 1999. Examining the self-enhancement bias: professional truck drivers’ perceptions of speed, safety, skill and consideration. Transport. Res. F 2, 91–113.
