The traffic climate in China: The mediating effect of traffic safety climate between personality and dangerous driving behavior

The traffic climate in China: The mediating effect of traffic safety climate between personality and dangerous driving behavior

Accident Analysis and Prevention 113 (2018) 213–223 Contents lists available at ScienceDirect Accident Analysis and Prevention journal homepage: www...

759KB Sizes 0 Downloads 0 Views

Accident Analysis and Prevention 113 (2018) 213–223

Contents lists available at ScienceDirect

Accident Analysis and Prevention journal homepage:

The traffic climate in China: The mediating effect of traffic safety climate between personality and dangerous driving behavior ⁎


Qian Zhanga,b, Yan Gea,b, , Weina Qua,b, , Kan Zhanga,b, Xianghong Suna,b a b

Key Laboratory of Behavioral Science, Institute of Psychology, Chinese Academy of Sciences, Beijing, China Department of Psychology, University of Chinese Academy of Sciences, Beijing, China



Keywords: Traffic climate Safety culture Dangerous driving behavior Personality

Traffic safety climate is defined as road users’ attitudes and perceptions of traffic in a specific context at a given point in time. The current study aimed to validate the Chinese version of the Traffic Climate Scale (TCS) and to explore its relation to drivers’ personality and dangerous driving behavior. A sample of 413 drivers completed the Big Five Inventory (BFI), the Chinese version of the TCS, the Dula Dangerous Driving Index (DDDI) and a demographic questionnaire. Exploratory factor analysis and confirmatory factor analysis were performed to confirm a three-factor (external affective demands, internal requirements and functionality) solution of the TCS. The reliability and validity of the Chinese version of TCS were verified. More importantly, the results showed that the effect of personality on dangerous driving behavior was mediated by traffic climate. Specifically, the functionality of the TCS mediated the effect of neuroticism on negative cognitive/emotional driving and drunk driving, while openness had an indirect impact on aggressive driving, risky driving and drunk driving based on the internal requirements of the TCS. Additionally, agreeableness had a negative direct impact on four factors of the DDDI, while neuroticism had a positive direct impact on negative cognitive/emotional driving, drunk driving and risky driving. In conclusion, the Chinese version of the TCS will be useful to evaluate drivers’ attitudes towards and perceptions of the requirements of traffic environment in which they participate and will also be valuable for comparing traffic cultures and environments in different countries.

1. Introduction Safety climate has concerned road safety researchers in recent years (Gehlert et al., 2014). Researchers have noted that the development of scientific safety studies in the context of road traffic has occurred in four stages (Özkan and Lajunen, 2011). At the beginning, technical safety measures were the focus; then, behavioral and individual factors gradually became the primary research targets. In the third stage, ergonomics and sociotechnical systems drew public attention, and more recently, in the fourth stage, road safety researchers have focused on the impact of traffic culture and climate (Gehlert et al., 2014; Guggenheim and Taubman-Ben-Ari, 2015; Krishen et al., 2015; Leviäkangas, 1998; Özkan and Lajunen, 2011; Schlembach et al., 2016). Safety culture and safety climate are interrelated concepts. Although they have only recently been introduced into the field of traffic safety, they have been valued in other work organizations for a long time (Beus et al., 2010). Safety culture was defined as “the product of individual and group values, attitudes, competencies, and patterns of behavior that determine the commitment to, and the style and

proficiency of, an organization’s health and safety programs” by the Advisory Committee on the Safety of Nuclear Installations (ACSNI) Study Group (1993). Safety climate, which can be viewed as part of safety culture, represents the individual’s perception of the value and importance of safety in relation to his or her organization’s policies, processes and patterns among its members at any given time, as manifested by recent or current events (Griffin and Neal, 2000; Zohar, 1980, 2000, 2011). Safety culture exists at a higher level of abstraction as the underlying belief in creating a climate. Climate reflects a perception of organizational structures and how it feels to be a member of the organization; culture refers to core values and beliefs regarding how to behave within an organizational unit (Mearns et al., 1998; Neal et al., 2000). From another perspective, safety climate was considered a manifestation of safety culture (Cheyne et al., 1998; Schein, 1985). Safety climate is closely aligned with a temporal “state of safety,” a relatively unstable “snapshot” of safety culture (Bhattacharya, 2015). Therefore, safety climate is more easily measured by the participants’ perception than safety culture. Safety climate has been applied in the traffic safety research field in

Corresponding author at: 16 Lincui Road, Chaoyang District, Beijing, 100101, China. E-mail addresses: [email protected] (Y. Ge), [email protected] (W. Qu). Received 20 October 2017; Received in revised form 22 December 2017; Accepted 23 January 2018 0001-4575/ © 2018 Elsevier Ltd. All rights reserved.

Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

behavior from different specific perspectives (Guggenheim and Taubman-Ben-Ari, 2015; Lee et al., 2016; Naevestad et al., 2015; Schlembach et al., 2016). Research exploring the relationship between safety climate and work-related driving behavior confirmed that traffic violations, errors and distraction were strongly related to safety climate factors (Wills et al., 2006). Consistently, Amponsah-Tawiah and Mensah (2016) found that safety climate predicts safe work-related driving behaviors. Other researchers even demonstrated that family and community climate regarding traffic influences driving behavior in young drivers (Guggenheim and Taubman-Ben-Ari, 2015; TaubmanBen-Ari and Katz-Ben-Ami, 2012). A study by Zohar et al. not only proved that the relationship between safety climate and traffic nearmiss events is represented by the actual recorded hard-braking frequency of truck drivers but also found that the relationship between safety climate and traffic near-miss events was fully mediated by the self-reported frequency of driving-safety shortcuts (Zohar et al., 2014). Another study found that self-reported speeding and aggressive driving mediated the path from attitude towards traffic safety to accident involvement (Mohamed and Bromfield, 2017).

recent years (Gehlert et al., 2014; Mader and Zick, 2014). Traffic safety climate is defined as “road users’ (e.g. drivers, bicyclists, and pedestrians) attitudes and perceptions of the traffic in a context (e.g. country) at a given point in time” (Özkan and Lajunen, 2011). As a potent and integrated concept, traffic safety climate includes all factors related to drivers, vehicles and infrastructure (Leviäkangas, 1998). It refers to an attitude encompassing cognitive, affective and behavioral components (Gehlert et al., 2014) and can be described as road users’ thoughts and feelings towards the traffic around them as well as their possible behavioral intentions (Gehlert et al., 2014). 1.1. Measurement of traffic safety climate To explore traffic safety climate quantitatively, a self-reported questionnaire named the Traffic Climate Scale (TCS) was developed by Özkan and Lajunen (unpublished) and was then validated in Germany (Gehlert et al., 2014). The initial structure of traffic safety climate is based on organizational climate, as there are many similarities between the two. Organizational climate, defined as “shared perceptions of organizational policies, practices, and procedures” (Hoy, 1990; Reichers and Schneider, 1990), refers to individuals’ perceptions of the practices, relationships and processes of their workplace (Nencini et al., 2016). Therefore, organizational climate is a result of interactions among individual, environmental, and instrumental variables, which refer to the three higher-order facets, including affective, cognitive and instrumental components (Carr et al., 2003). The affective facet refers to member involvement and social relations, such as participation and cooperation. The cognitive facet includes the degree of psychological demand and work-related skills or knowledge, such as innovation and autonomy. The instrumental facet relates to tasks or work processes, such as structure and hierarchy. The development of TCS takes into account the structure of organizational safety climate through external affective demands, internal requirements, and functionality, which correspond to the three components of organizational safety climate. External affective demands reflect the emotions of road users when participating in and interacting with the traffic environment or rules (e.g., “stressful,” “time-consuming,” and “dangerous”). The internal requirements factor focuses on road users’ cognition about traffic in which they participate and refers to the skills, workload and abilities that make them capable in traffic (e.g., “demands vigilance,” “demands fast reactions” or “demands knowledge of traffic rules”). Functionality measures road users’ requirements for the properties of a functional traffic system as represented by adjectives that describe the state of the traffic facilities and environment (e.g., “safe,” “free-flowing,” or “forgives mistakes”) (Gehlert et al., 2014). Previously, the TCS was validated in Germany and Lithuania and was written in German and Lithuanian (Gehlert et al., 2014; Marksaityte et al., 2014). In Germany and Lithuania, gender and age are important factors in evaluating traffic safety climate: men reported a more positive attitude than women; young male and older female drivers perceived traffic climate as safer and less challenging than older males and younger females (Marksaityte et al., 2014); and female drivers who had been involved in at least one traffic accident perceived the traffic system as more functional. However, no relationships between traffic safety climate and self-reported penalties were found.

1.3. Relationship between personality and driving behavior The personality traits of drivers are important factors in traffic safety studies, and the big five personality traits have received much attention in this research field. Researchers found that each of the big five dimensions has a different influence on driving behavior and accident involvement (af Wåhlberg et al., 2017). Conscientiousness and openness were mainly negatively correlated with dangerous driving behavior. Conscientiousness was found to have a negative relation with AD (Benfield et al., 2007; Burtăverde et al., 2016; Guo et al., 2016; Harris et al., 2014; Schwebel et al., 2006; Zhang et al., 2017), especially when it negatively correlated with angry thoughts behind the wheel, such as pejorative labeling and verbally aggressive thinking (Benfield et al., 2007). Similarly, openness was negatively related to self-reported RD behaviors as well as AD behaviors (Benfield et al., 2007; Burtăverde et al., 2016; Dahlen and White, 2006; Harris et al., 2014). In contrast, neuroticism was mainly positively correlated with dangerous driving behavior. AD and RD were predicted by a higher neuroticism score or a lower emotional stability score (Burtăverde et al., 2016; Dahlen and White, 2006; Jovanović et al., 2011; Richer and Bergeron, 2012; Zhang et al., 2017). The path from neuroticism to driving anger was positive (Richer and Bergeron, 2012). Additionally, the effect of agreeableness and extraversion on driving behaviors were complicated. The results for the relationship between agreeableness and dangerous driving behaviors are inconsistent. A number of studies confirmed that agreeableness negatively correlated with or predicted aggressive driving (AD), risky driving (RD) and negative cognitive/emotional driving (NCED) (Benfield et al., 2007; Burtăverde et al., 2016; Dahlen et al., 2012; Harris et al., 2014; Jovanović et al., 2011; Zhang et al., 2017). However, contrary results indicated that lower agreeableness was associated with a lower number of times that a driver violated safety rules (Guo et al., 2016). Referring to extraversion, some researchers found it to be associated with more frequent physically aggressive driving or reckless driving behavior (Benfield et al., 2007; Harris et al., 2014), while certain researchers also found it to be negatively correlated with verbally aggressive driving (Burtăverde et al., 2016).

1.2. Relationship between traffic safety climate and driving behavior A few studies have explored the relationship between traffic climate and driving behavior. For example, Gehlert et al. (2014) found that the internal requirements of road users are related to individuals’ driving or riding style and that external affective demands are associated with individual perception: road users who experienced higher internal requirements and functionality perceived more behavioral control and a less descriptive norm (Gehlert et al., 2014). However, many studies have confirmed the important role of safety climate in traffic-related

1.4. Relationship between personality and traffic climate Few studies have explored the relationship between personality and traffic climate, but studies of personality and safety attitude can provide some inspiration. Some personalities have a direct effect on the attitude towards safety. For example, altruism and anxiety have a positive correlation with positive attitudes towards traffic safety and rules 214

Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

Therefore, valid samples were obtained from 413 (131 females) drivers with an average of 8.12 years (SD = 6.388) of driving experience and an average annual mileage of 10,435.84 km (SD = 7679.298 km). The drivers were 20–60 years old (M = 40.366, SD = 9.936). Table 1 summarizes the demographic details of the final sample.

(Chen, 2009; Lucidi et al., 2014; Mallia et al., 2015; Ulleberg and Rundmo, 2003). In contrast, normlessness, sensation-seeking/excitement-seeking and hostility/aggressiveness are negative predictors of positive attitudes towards traffic safety and rules (Chen, 2009; Lucidi et al., 2014; Mallia et al., 2015; Ulleberg and Rundmo, 2003). Meanwhile, normlessness, sensation-seeking and anger were three positive predictors of agreement of risky driving (Chen, 2009). Researchers also found an indirect effect of personality on dangerous driving behaviors that was mediated by safety climate. Some researchers noted that attitudes towards safe driving mediated the effects of personality on selfreported risky or aberrant driving behaviors, including errors, lapses and violations (Lucidi et al., 2014; Mallia et al., 2015). A conceptual model proposed by Murphy and his colleagues also stressed that the work system through safety climate determines individuals’ behaviors (Murphy et al., 2014). As mentioned earlier, traffic safety climate was defined as individuals’ attitudes and perceptions, so we hypothesized that traffic safety climate could mediate the effect of personality on driving behaviors.

2.2. Instruments 2.2.1. The big five inventory The Big Five Inventory (BFI) is a short-phrase instrument used to measure five factors of personality in an individual (John and Srivastava, 1999). It uses 44 items to construct the five dimensions: extraversion (8 items; α = 0.86; e.g., “I see myself as someone who is talkative”), agreeableness (9 items; α = 0.79; e.g., “I see myself as someone who likes to cooperate with others”), conscientiousness (9 items; α = 0.82; e.g., “I see myself as someone who does a thorough job”), neuroticism (8 items; α = 0.84; e.g., “I see myself as someone who can be moody”), and openness (10 items; α = 0.80; e.g., “I see myself as someone who is inventive”). Each item is evaluated on a 5point Likert scale from 1 (“strongly disagree”) to 5 (“strongly agree”). The Chinese version of the BFI has been used in the Chinese population and demonstrated good reliability and validity, with the internal consistency coefficient higher than 0.71 (Wang et al., 2014). In the present sample, the Cronbach’s alpha coefficient of the 44-item BFI is 0.71.

1.5. Purposes and hypotheses The first purpose of the current study is to verify the TCS in the Chinese language, culture and environment. There are many reasons to develop a Chinese version of the TCS. First, China is undergoing a dramatic increase in the number of vehicles and drivers. The number of automobiles in China exceeded 2.83 billion by the end of March 2014, according to recent statistics from the Roads and Traffic Authority (RTA) of the Ministry of Public Security (MPS) (The Ministry of Public Security of the People’s Republic of China, 2016). The traffic environment and facilities in China are unique; therefore, China has implemented traffic rules that are different from those in Germany. Second, Chinese drivers have a unique traffic cultural background that could promote different driving styles and behaviors. For example, in driving anger studies, Chinese drivers reported a lower level of anger in response to discourtesy and illegal driving than their American counterparts as well as drivers in New Zealand and Spain (Li et al., 2014). In terms of driving skills, Chinese drivers exhibit a different driving style and produce more crashes than those in the United States and Japan (Atchley et al., 2014; Zhang et al., 2006, 2010). Finally, although multiple studies have explored the cultural predictors of car accident involvement across countries (Nordfjaern et al., 2012; Poó et al., 2013; Rundmo et al., 2012), no research has directly investigated traffic safety climate in China. Therefore, it is meaningful to assess traffic safety climate in China. The second purpose is to assess the relations between personality, traffic climate and dangerous driving behavior in a Chinese sample. Given previous studies of relations between the big five personality traits, traffic safety climate, dangerous driving behavior and traffic violations, we proposed a path diagram of these variables, as shown in Fig. 1. The effectiveness of personality in predicting dangerous driving behavior and traffic violations has been proved in previous studies, but no study has explored the role of traffic safety climate. Therefore, the main hypotheses we aim to prove in this path diagram are H1 : Traffic safety climate predicts dangerous driving behavior and traffic violations. H2 : Traffic safety climate has a mediating effect between big five personality traits and dangerous driving behavior.

2.2.2. The TCS The TCS is a self-reported tool for investigating people’s attitudes towards traffic safety climate (Gehlert et al., 2014). Participants were asked to indicate to what degree the items describe traffic on a 6-point Likert scale from 1 = “does not describe it at all” to 6 = “describes it fully.” The original scale includes 41 items (Özkan and Lajunen, unpublished), but 11 items were deleted because they appeared in more than one dimension through exploratory factor analysis (EFA) in the German version of the TCS (Gehlert et al., 2014). The two models both contained three factors: “internal requirements,” “functionality” and “external affective demands.” The 41-item version was used in this study. To generate the Chinese version of the TCS, a translation/backtranslation procedure (Brislin, 1980; Regmi et al., 2010) was used to translate the scale. First, the English version of the TCS that was used in Germany (Gehlert et al., 2014) was translated into Chinese independently by two researchers. Next, the authors compared and discussed the two translated scales to create a single document that is more accurate, fluent and adapted to the Chinese driving culture. Then, the integrated draft was back-translated into English by a professional English-Chinese translator and compared to the original items to identify differences. An agreement among all researchers was reached by discussing and modifying the noted differences. This version of the draft was evaluated by four experienced drivers to ensure that each item was clearly expressed. The formal scale was obtained after this procedure. In the final version, one new option, 7 = “don’t understand,” was added to identify items that may be difficult to understand in the Chinese culture. 2.2.3. The dula dangerous driving index In previous studies, the Dula Dangerous Driving Index (DDDI) was frequently used to measure participants’ self-reported likelihood of driving dangerously (Dula and Ballard, 2003). The Chinese DDDI (Qu et al., 2014) was used in this study to help examine the relations among personalities, the Chinese version of the TCS and dangerous driving behaviors. The participants indicated how frequently they had committed each behavior on a five-point Likert scale (1 = “never”; 5 = “always”). The higher the total score of each subscale, the more unsafe the driving behavior. Twenty-eight self-reported items composed four factors referring to unsafe driving behavior: AD (7 items, α = 0.77; e.g., “I flash my headlights when I am annoyed by another

2. Methods 2.1. Participants Of the 449 drivers who completed and returned the questionnaire, 18 were excluded for selecting the same option(s) on one or more scales. In addition, 18 additional drivers were excluded because they considered more than 5 items difficult to understand (choosing “7”). 215

Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

Fig. 1. Theoretical model of big five personality traits, traffic safety climate, dangerous driving behavior and traffic violations.

the survey aimed to investigate the traffic environment and driving behavior and that they could participate voluntarily and anonymously. The TCS, DDDI, BFI, and demographic measurements were combined into one questionnaire as part of a larger survey. It took participants 20 min to complete the entire questionnaire. After completing the questionnaire, participants were rewarded with gifts worth approximately 20 RMB. The study was approved by the Institutional Review Board of the Institute of Psychology, Chinese Academy of Sciences.

Table 1 Participant demographics (N = 413).

Age groups by gender 20–30 years Males Females 31–40 years Males Females 41–50 years Males Females 51–60 years Males Females Driving years ≤5 years 6–10 years 11–15 years 16–20 years > 20 years Annual mileage ≤5000 km 5001–10000 km 10001–20000 km > 20,001 km


Percent (%)

53 21

12.8 5.1

84 40

20.4 9.7

89 44

21.6 10.7

56 26

13.6 6.3

165 159 37 32 20

40.0 38.5 9.0 7.8 4.8

76 217 101 19

18.4 52.5 24.5 4.6

2.4. Statistical analysis All statistical analyses were performed using SPSS 19.0 and AMOS 21.0. First, the date were divided into two parts to perform exploratory factor analysis (EFA) with SPSS and confirmatory factor analysis (CFA) with AMOS by exploring the proper structure of the TCS in the Chinese sample. Second, a reliability analysis was conducted with SPSS to indicate the internal consistency of the Chinese version of the TCS. Third, Spearman correlation coefficients were performed to determine the association of the big five personality traits, the TCS, the DDDI and traffic violations. Finally, these relations were included in a path analysis model to determine the nominated variables’ effects on traffic violations. AMOS 21.0 was used to perform CFA and path analysis. Acceptable model fit was indicated by the Steiger-Lind root mean square error of approximation (RMSEA) of 0.08 or less (Browne et al., 1993) and the Joreskog-Sorbom goodness-of-fit index (GFI) of 0.90 or more. In addition, the Bentler comparative fit index (CFI ≥ 0.90) and the TuckerLewis index (TLI ≥ 0.90) were used to indicate the degree of model fit improvement during the iterative CFA process (Bentler and Bonett, 1980). The Akaike’s information criterion (AIC) was the main index indicating model parsimony; a lower AIC value suggested a more parsimonious model.

driver”), RD (10 items, α = 0.76; e.g., “I will cross double yellow lines to see if I can pass a slow-moving car/truck”), drunk driving (DD; 2 items; α = 0.65; e.g., “I will drive when I am drunk”) and NCED (9 items; α = 0.79; e.g., “I drive when I am angry or upset”). 2.2.4. Demographic questionnaire The demographic information of all respondents consisted of three parts. First, general demographic details, such as age and gender, were requested. Then, participants provided detailed information about their driving behaviors, including years of driving experience, total mileage, average annual mileage, and fines and penalty points received over the past year.

3. Results 3.1. Exploratory factor analysis A total of 205 participants were randomly selected using the Random sample of Select Cases function in SPSS (the sample size was set as approximately 50% of all cases) for EFA. Principal component analysis (PCA) with promax rotation was used to assess the factorial structure of the 41-item TCS, since rather strong intercorrelations existed between items. The initial EFA indicated that some items should

2.3. Procedure Data collection was carried out by a private research company using a convenient sampling method in Beijing, China. Drivers were told that 216

Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

Table 2 Structure matrix of principal component analysis with promax rotation (N = 205). Factor External affective demands Chaotic Fate dictates matters Unpredictable Risky Requires experience Demands compliance Demands patience Requires skillfulness Harmonious Mobile Fast Free-flowing Safe % of variance explained

Table 3 Goodness-of-fit indexes for four models of the brief TCS (N = 208).

Communalities Internal requirements



0.779 0.777

0.607 0.604

0.769 0.707 0.783

0.596 0.554 0.632







Original model (41 items) German model (30 items) EFA model (13items) Final model (13items)







χ 2 / df















207.3 127.7

3.3 2.2

0.871 0.919

0.827 0.918

0.782 0.892

0.106 0.075

265.3 191.7

Notes: Final model, 3-factor, 13-item TCS with 3 pairs of errors covaried.



0.704 0.696 0.654 0.648 0.577 10.630

When the model was specified, an acceptable fit was obtained, with all indicators meeting the standards. Thus, the final 13-item TCS was used for further analyses. The results are shown in Table 3.

0.542 0.520 0.448 0.432 0.364

3.3. Internal consistency Cronbach’s α coefficients of external affective demands, internal requirements and functionality are 0.810, 0.760 and 0.723, respectively. Internal consistency and reliability should be considered adequate for the Cronbach’s α coefficients of each factor greater than 0.7. Means, standard deviations and correlation coefficients were estimated for the three factors; the results are summarized in Table 4. Internal requirements showed a significantly positive correlation with external affective demands as well as functionality.

Notes: Loadings less than 0.3 and communalities less than 0.2 were suppressed. Items with cross-loading were removed from the scale step by step to optimize the psychometric properties of the brief TCS.

be moved or removed for bad loading value or cross-loading. Then, the minimum average partial (MAP) test (O’connor, 2000; Velicer, 1976; Velicer et al., 2000) was used to clarify the number of factors, and the data best fitted a three-factor solution with 41 items, which accounted for 53.9% of the total variance. Next, PCA was carried out, setting the number of factors to three many times to delete improper items. A total of 14 items was finally included based on three criteria: (1) factor loading greater than 0.300, (2) communality value not less than 0.200, and (3) no cross-loading item. In addition, item 36 (“planned”) was deleted for being assigned to the “internal requirements” factor while belonging to the “functionality” factor in the German version of the TCS; this item also had a different connotation than other items of the “internal requirements” factor. The final EFA results of 13 TCS items are shown in Table 2. The 13 items were divided into three factors, explaining 54.0% of the evaluation scale variance. Factor 1, labeled “external affective demands,” explained 24.5% of the variance. It included 4 items that are all related to emotional engagement in traffic. Factor 2 was labeled “internal requirements” and explained 18.9% of the variance; it was also defined by 4 items that refer to individual skills and abilities to successfully participate in traffic. Factor 3 had 5 items related to requirements for a functional traffic system; it was labeled “functionality” and contributed 10.6% of the variance.

3.4. Correlations among personalities, TCS factors and DDDI factors Spearman correlations among variables are shown in Table 4. The remaining four factors of the BFI, except neuroticism, have significantly positive relations with each other and all show significantly negative relations with neuroticism. Even more remarkable is the relationship between the BFI and the TCS and the BFI and the DDDI as well as driving violations. The neuroticism factor is positively related to the functionality of the TCS and is negatively related to the other two subscales of the TCS. Extraversion, conscientiousness and openness are positively related to the external and internal requirements of the TCS. Agreeableness is significantly correlated only with the internal requirements, and the relationship is positive. These four BFI factors are also positively related to the four subscales and total scores of the DDDI, which shows a significantly negative relationship with neuroticism. The BFI has no significant relation to fines and penalty points over the previous year. There are also some significant correlations between the TCS and demographic variables and the TCS and the DDDI as well as driving violations. External affective demands negatively correlated with gender, indicating that males experience more affective demands from the traffic environment. Drivers of younger ages have more internal requirements and less functionality assessment. DD has been found to negatively correlate with the external affective demands but positively correlate with the functionality of the TCS. The internal requirements of the TCS negatively correlate with both total score and the four subscales of the DDDI. The functionality factor shows a significantly positive relation to fines. The DDDI’s total score and four subscale scores all have significantly positive relations to age, years of driving experience and total mileage. The total score and NCED score of the DDDI show significantly positive relations to penalty points. In addition, years of driving experience, total mileage and annual mileage are positively related to each other and show no gender differences. Age is found to positively relate to driving years and total mileage but to negatively relate to annual mileage.

3.2. Confirmatory factor analysis CFA was conducted on the remaining 208 cases that were not selected for PCA. To explore the optimal structure of the Chinese version of the TCS, the 41-item original model, 30-item German model and 13item EFA model mentioned above were used to fit the Chinese sample data. These three models showed less than adequate fit to the data with RMSEA higher than 0.08 or GFI, CFI, and TLI lower than 0.90. Next, the Lagrange multiplier tests revealed that three error pairs with similar concepts might be covaried. Every item pair was from the same factor, of which “chaotic” and “risky” items and “chaotic” and “unpredictable” items belonged to external affective demands, and the other error pair, “harmonious” and “free-flowing” items, belonged to functionality. 217


Age Gender Driving Years Annual mileage Total mileage Extraversion Agreeableness Conscientiousness Neuroticism Openness External Internal Functionality DDDI -NCED -AD -RD -DD Penalty points Fines

40.37 – 8.12 10,435 91,900 3.25 3.51 3.39 2.74 3.31 4.28 4.59 3.85 2.01 2.18 1.88 1.97 1.74 1.39 140.56

– – – – – 8 9 9 8 10 4 4 5 28 9 7 10 2 – –

9.94 – 6.39 7679 112,570 0.48 0.56 0.51 0.52 0.45 0.94 0.82 0.81 0.70 0.70 0.78 0.77 0.87 2.15 220.84

SD – 0.02 0.31** −0.12* 0.19** −0.20** −0.23** −0.18** 0.23** −0.17** 0.01 −0.12* 0.10* 0.17** 0.11* 0.22** 0.17** 0.18** 0.04 −0.03


– −0.01 −0.07 −0.04 0.17** 0.09 0.15** −0.11* 0.10 −0.10* 0.04 0.01 −0.09 −0.06 −0.09 −0.09 −0.04 0.02 0.01


– 0.29** 0.87** −0.08 −0.12* −0.10* 0.10* −0.10* 0.04 −0.09 −0.06 0.12* 0.09 0.12* 0.12* 0.11* −0.02 −0.02


– 0.66** 0.07 −0.06 0.02 −0.11* −0.01 −0.03 −0.03 −0.07 0.07 0.08 0.07 0.03 0.04 0.07 0.06


– −0.03 −0.13** −0.09 0.03 −0.08 0.02 −0.07 −0.06 0.12* 0.10* 0.13** 0.10* 0.10* 0.03 0.02


– 0.41** 0.51** −0.44** 0.53** 0.14** 0.26** 0.07 −0.26** −0.17** −0.29** −0.26** −0.28** 0.03 0.04


– 0.64** −0.58** 0.47** 0.09 0.32** −0.09 −0.49** −0.34** −0.55** −0.50** −0.55** −0.04 −0.02


– −0.58** 0.48** 0.12* 0.28** −0.09 −0.38** −0.24** −0.42** −0.42** −0.44** −0.03 −0.04


– −0.34** −0.06 −0.23** 0.17** 0.40** 0.30** 0.43** 0.42** 0.40** −0.01 −0.01


– 0.13** 0.28** 0.03 −0.20** −0.09 −0.24** −0.21** −0.23** 0.06 0.08


– 0.42** 0.02 −0.09 −0.07 −0.08 −0.09 −0.15** −0.07 0.01


– 0.22** −0.24** −0.17** −0.30** −0.24** −0.27** −0.08 −0.07


– 0.03 −0.07 0.09 0.08 0.17** −0.1 −0.13**


Notes: Gender: male = 1, female = 2. DDDI = Dula Dangerous Driving Index; NCED = negative cognitive/emotional driving; AD = aggressive driving; RD = risky driving; DD = drunk driving. * p < .05. ** p < .01.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20



– 0.93** 0.91** 0.94** 0.73** 0.11* 0.09


– 0.77** 0.82** 0.56** 0.15** 0.12**


– 0.86** 0.74** 0.07 0.04


Table 4 Descriptive statistics of demographic variables and scales and correlations between age, years of driving experience, gender, annual mileage, total mileage, big five personality traits, the TCS and the DDDI (N = 413).

– 0.75** 0.08 0.06


– 0.07 0.09


– 0.81**


Q. Zhang et al.

Accident Analysis and Prevention 113 (2018) 213–223

Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

Fig. 2. Revised path model of variables with significantly standardized regression estimates. Standardized estimates: Chi-square = 51.3, Chi-square/df = 2.4, GFI = 0.975, CFI = 0.986, TLI = 0.970, RMSEA = 0.059, AIC = 119.3.

in China. Most importantly, traffic safety climate played a key role in driving safety. The drivers’ attitude towards traffic climate mediated the effect of personality on driving behavior. First, the Chinese version of the TCS showed preferable reliability and a stable structure. The internal consistency reliabilities of the TCS were high enough, with coefficient values above 0.70 for each factor. The internal structure of the TCS was similar to that of the German version (Gehlert et al., 2014). The construct validity of the TCS structure was proved by both EFA and CFA. Factor analysis revealed a threefactor structure of the Chinese version of traffic safety climate, including internal requirements, external affective demands and functionality. As the model fit in CFA indicated, the 30-item TCS was not a good fit in the Chinese sample; therefore, inappropriate items were removed based on data analysis, including double loading. In addition, considering that traffic rules and driving habits vary across countries (ITC, 2010), some modifications were adopted to make the TCS more suitable for the Chinese sample. Therefore, there are some differences in items between the original version of the TCS and the Chinese version. First, in the original TCS, two items (“mobile” and “fast”) were in the functionality factor, but in the Chinese version, they belonged to the internal requirements factor. On one hand, these two items could be used to describe traffic conditions, which correspond to the functionality factor; for example, the item “free-flowing” in functionality has a meaning similar to that of “fast.” On the other hand, the internal requirements factor emphasizes the needs and demands of skills or abilities, e.g., “requires experience” and “demands patience,” which are different from “fast” and “mobile.” Therefore, the final three-factor Chinese version of the TCS with 13 items showed sufficient construct validity. Next, the relationship between the Chinese version of the TCS and self-reported driving behaviors was explored. Driving behavior was measured by the DDDI, a validity tool used to measure the frequency of AD, RD, DD and NCED (Dula and Ballard, 2003). The results showed that the TCS factors were significantly related to the DDDI. Internal requirements have negative correlations with all DDDI subscales. These results indicated that drivers who reported a higher need of driving skills and abilities reported less dangerous driving behavior. These findings supported previous results that less cognitively demanding of driving related to more traffic violations (Gehlert et al., 2014). This result implied a higher need for skills and abilities related to cognition to make drivers more cautious in driving. In addition, external affective demands and functionality showed opposite relationships to DD. External affective demands were negatively correlated with DD, while functionality was positively correlated with DD. These results indicated that the more emotional demands and less functionality, the less unsafe

3.5. Path analysis We used maximum likelihood procedures to test the fit of the path model shown in Fig. 1. According to model fit indexes, as mentioned in Section 2.3, after deleting all non-significant pathways between big five personality traits, the TCS, the DDDI and violations, the revised model showed a good model fit, with all indexes meeting the standards. Furthermore, all pathways in the revised model were significant and modified indexes were lower than 6. The final model included ten factors (see Fig. 2): the three personality factors were agreeableness, neuroticism and openness; the two TCS factors were functionality and internal requirements; the four DDDI factors were AD, RD, DD and NCED; and the one traffic violation factor was penalty points over the previous year. The model indicated, first, that functionality mediated the effect of neuroticism on NCED and DD and openness had an indirect impact on RD, DD and AD by the internal requirements of the TCS; second, that agreeableness had a direct impact on the four factors of the DDDI and neuroticism had a direct impact on NCED and RD; and finally, that only NCED directly predicted penalty points in the previous year. The values of the standardized effects are shown in Table 5. 4. Discussion The main purpose of this study was to verify the psychometric properties of the Chinese version of the TCS and to explore its relationship with personality, self-reported driving behavior and traffic violations. The results confirmed the reliability and validity of the TCS Table 5 Standardized effects on the DDDI and penalty points from path analysis.

Agreeableness → Points Agreeableness → NCED Agreeableness → RD Agreeableness → DD Agreeableness → AD Neuroticism → Points Neuroticism → NCED Neuroticism → RD Neuroticism → DD Openness → RD Openness → DD Openness → AD

Total effect

Direct effect

Indirect effect

−0.03 −0.30 −0.40 −0.46 −0.50 0.01 0.11 0.14 0.01 −0.02 −0.04 −0.03

– −0.30 −0.40 −0.46 −0.50 – 0.14 0.14 – – – –

−0.03 – – – – 0.01 −0.03 – 0.01 −0.02 −0.04 −0.03

Notes Points = penalty points over the previous year; DDDI = Dula Dangerous Driving Index; NCED = negative cognitive/emotional driving; AD = aggressive driving; RD = risky driving; DD = drunk driving.


Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

mobile phone conversation (Haigney et al., 2000). The lower speed effect was considered an effort to compensate for the increased mental workload. In this study, it is possible that drivers realized they should apply more cognitive resources to handling vehicles so that unsafe driving behaviors decreased to enable enough resources to maintain the level of perceived safety standards. However, functionality positively predicted dangerous driving behavior but negatively predicted negative cognitive/emotional thinking. When drivers assessed the traffic system as well organized, they might experience less arousal of negative emotion or cognition. However, they may display more dangerous driving behavior if they believe the traffic system is poorly organized. More importantly, the results showed that the effect of neuroticism on negative emotion or cognition was partially mediated by functionality. Meanwhile, the effect of openness on dangerous driving behavior was completely mediated by internal requirements. Although no previous studies have directly explored the role of traffic climate in the relationship between personalities and driving behaviors, the findings of this study were supported by some relevant studies. In studies by Ulleberg and Rundmo (2003) and Lucidi et al. (2014), the effects of anxiety, normlessness, sensation-seeking, and hostility/aggression on RD were mediated by attitudes towards traffic safety. Traffic safety climate was defined as individuals’ attitudes and perceptions (Özkan and Lajunen, 2011). A similar pattern of results was obtained in the present study, thus confirming the mediating role of traffic safety climate in personality-driving behavior relationships. The study is not free of limitations. The data were self-reported by drivers, so it is difficult to eliminate contamination by social desirability. Indeed, a study reported that employees avoid disadvantages by less reporting of accidents (Gadd and Collins, 2002). It is preferable to use actual traffic violation and accident data from the traffic management system, which can allow actual measurements of driving performance indicators. In addition, the participants were recruited using a convenient sampling method in Beijing, China. But China is a large country with diverse road and traffic conditions, vehicle use, cultures and climates, etc. So the representativeness of this study is insufficient. This may limit the generalizability of our conclusions to the entire population of drivers in China. Future research could enlarge the sample distribution and apply a well-designed sampling method to validate the questionnaire. Furthermore, we just validated the original version of TCS in China, taking the specificity of the Chinese traffic environment into account, more unique items according with Chinese Culture should be added into the Chinese version of the TCS. In conclusion, the Chinese version of the TCS is a valid and reliable tool for assessing drivers’ attitudes towards the driving environment in which they participate. The results of EFA and CFA confirmed the threefactor (internal requirements, external affective demands and functionality) structure of TCS, which is consistent with the three components of attitude, supporting the assumption of the traffic safety climate structure proposed by Gehlert and collaborators (Gehlert et al., 2014). Meanwhile, the sufficient internal consistency for each factor proved its reliability. Its direct relationship with self-reported driving behaviors and violations implied the validity of the Chinese version of the TCS. More importantly, traffic safety climate mediated the effect between personalities and dangerous driving behaviors. If the driving environment is perceived as dysfunctional and demanding more cognition, drivers are more careful to ensure driving safety. The TCS can be used to classify drivers according to attitude, thus predicting their reaction types during driving. Moreover, high scores of external affective demands and internal requirements as well as low score of functionality in Chinese drivers indicate that traffic management and equipment should be strengthen to reduce mental load of drivers and increase transportation efficiency. These evidence and information could provide some support for policy proposal and improving the traffic climates in China.

driving behavior. A high score in the external affective demands dimension indicates that drivers believe the traffic in their city is chaotic, risky and uncontrollable and may therefore drive carefully. However, a high score in functionality indicates that the traffic is perceived as orderly and well-behaved; thus, drivers appear to be less careful because of their confidence in the traffic system. Although we did not find direct support from driving behavior research, similar results were found in the field of organizational safety climate. A model analysis study found that a positive attitude towards behavior may lead to less perceived behavioral control, in turn resulting in more self-reported speeding and even logged speeding (Warner and Aberg, 2006). In this study, the drivers’ attitude towards the surrounding traffic situation was related to their behaviors. When drivers believed the traffic environment was stressful (high external affective demands), required skillfulness (high internal cognitive demands) and was dysfunctional (low functionality), they might drive safely and engage in fewer dangerous driving behaviors. To explore the relationship of traffic climate with personality, selfreported driving behavior and traffic violations, a personality-traffic climate-driving behavior model was tested. In agreement with previous research, we found that dangerous driving behaviors were predicted by personality traits. Agreeableness negatively predicted all types of dangerous driving behaviors, while neuroticism had a positive direct path to NCED and RD. Multiple previous studies have supported the statistically significant relations between agreeableness and neuroticism and dangerous driving behavior (Benfield et al., 2007; Burtăverde et al., 2016; Dahlen et al., 2012; Guo et al., 2016; Harris et al., 2014; Jovanović et al., 2011; Richer and Bergeron, 2012; Zhang et al., 2017). Although no statistically significant direct path between dangerous driving behavior and the other three personality traits, openness, conscientiousness and extraversion, was found in the model analysis, the results of the correlation analysis indicated that these three personality traits negatively correlated with dangerous driving behavior, corresponding with the findings of previous studies (Benfield et al., 2007; Burtăverde et al., 2016; Dahlen and White, 2006; Guo et al., 2016; Harris et al., 2014; Jovanović et al., 2011; Schwebel et al., 2006; Zhang et al., 2017). Notably, traffic safety climate is also directly predicted by personality traits. The findings of this study indicated that neuroticism positively predicted functionality, which means drivers with high neuroticism tend to optimistically evaluate the traffic environment. Moreover, openness directly predicted internal requirements. Openness is likely to be associated with tolerance, leading to situational attributions of other drivers’ behavior (Dahlen et al., 2012). Therefore, drivers with high openness experience more of a cognitive load from their surroundings and thus reduce the likelihood of invoking hostile reactions. Regarding the direct effect of traffic climate on dangerous driving behavior, the results of this study showed that functionality and internal requirements significantly predicted dangerous driving behavior and negative cognitive/emotional thinking. On the one hand, more skill or cognitive demands significantly predicted less unsafe driving behavior. A possible explanation is that subjects regulate the distribution of cognitive resources when multitasking to ensure the normal performance of multiple tasks. In the context of attentional resource models (Wickens, 1991), with more than one time-sharing task and demand for resources exceeding the available supply, the subject modulates the supply of resources between the tasks to obtain the desired level of differential performance. This hypothesis might refer to a process of risk homeostasis (Wilde, 1982, 1988), which posits that people continuously modify their actions to sustain a balance between perceived and acceptable risks. Some research in the driving behavior field supports this explanation. A study focusing on mobile phone use during driving found that dialing and conversation lead to decreased self-reported and actual driving speeds (Tornros and Bolling, 2005). A similar study conducted in a simulator found that participants engage in a process of risk compensation, with driving speed being lower at times of 220

Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.


Natural Science Foundation of China (31771225, 31400886, and 31100750), the Basic Project of National Science and Technology of China (2009FY110100), the Pioneer Initiative of the Chinese Academy of Sciences, Feature Institutes Program (TSS-2015-06).

This study was partially supported by grants from the National Key Research and Development Plan (2017YFB0802800), the National

Appendix A. 41 items of the Traffic Climate Scale in English and in Chinese



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

Annoying Makes you irritated Pressuring Stressful Provokes tension Chaotic Risky Dangerous Aggressive Exciting Time-consuming Unpredictable Fate dictates matters Makes you feel worthless Depends on one’s luck Complex Monotonous Demands vigilance Demands knowledge of traffic rules Demands alertness Demands compliance Demands caution Demands fast reactions Requires experience Requires skillfulness Demands mutual courtesy Demands patience Mobile Fast Directs your behavior Non-compliance is rewarding Harmonious Egalitarian Safe Free-flowing Planned Includes preventive measures Functional Forgives mistakes Dynamic Under enforcement

Chinese (

Appendix B. Table of all abbreviations

Complete expression


Traffic Climate Scale Big Five Inventory Dula Dangerous Driving Index Aggressive Driving Risky Driving Drunk Driving Negative Cognitive/Emotional Driving Roads and Traffic Authority




Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

Ministry of Public Security Exploratory factor analysis Confirmatory factor analysis Root mean square error of approximation Goodness-of-fit index Comparative fit index Akaike’s information criterion Principal component analysis Minimum average partial


Jovanović, D., Lipovac, K., Stanojević, P., Stanojević, D., 2011. The effects of personality traits on driving-related anger and aggressive behaviour in traffic among Serbian drivers. Transp. Res. Part F: Traffic Psychol. Behav. 14 (1), 43–53. Krishen, A.S., Kachroo, P., Agarwal, S., Sastry, S.S., Wilson, M., 2015. Safety culture from an interdisciplinary perspective: conceptualizing a hierarchical feedback-based transportation framework. Transp. J. 54 (4), 516–534. Lee, J., Huang, Y.H., Murphy, L.A., Robertson, M.M., Garabet, A., 2016. Measurement equivalence of a safety climate scale across multiple trucking companies. J. Occup. Organ. Psychol. 89 (2), 352–376. Leviäkangas, P., 1998. Accident risk of foreign drivers—the case of Russian drivers in South-Eastern Finland. Accid. Anal. Prev. 30 (2), 245–254. Li, F., Yao, X., Jiang, L., Li, Y.J., 2014. Driving anger in China: psychometric properties of the driving anger scale (DAS) and its relationship with aggressive driving. Personal. Individ. Differ. 68, 130–135. Lucidi, F., Mallia, L., Lazuras, L., Violani, C., 2014. Personality and attitudes as predictors of risky driving among older drivers. Accid. Anal. Prev. 72, 318–324. http://dx.doi. org/10.1016/j.aap.2014.07.022. Mader, E.M., Zick, C.D., 2014. Active transportation: do current traffic safety policies protect non-motorists? Accid. Anal. Prev. 67, 7–13. 2014.01.022. Mallia, L., Lazuras, L., Violani, C., Lucidi, F., 2015. Crash risk and aberrant driving behaviors among bus drivers: the role of personality and attitudes towards traffic safety. Accid. Anal. Prev. 79, 145–151. Marksaityte, R., Zardeckaite-Matulaitiene, K., Pranckeviciene, A., Seibokaite, L., Endriulaitiene, A., 2014. Perception of the traffic safety climate. In: Comparison of Different Groups of Lithuanian Drivers. Paper Presented at the 18th International Conference on Transport Means. Kaunas Univ Technol, P.. Mearns, K., Flin, R., Gordon, R., Fleming, M., 1998. Measuring safety climate on offshore installations. Work Stress 12 (3), 238–254. 02678379808256864. Mohamed, M., Bromfield, N.F., 2017. Attitudes, driving behavior, and accident involvement among young male drivers in Saudi Arabia. Transp. Res. Part F: Traffic Psychol. Behav. 47, 59–71. Murphy, L.A., Robertson, M.M., Carayon, P., 2014. The next generation of macroergonomics: integrating safety climate. Accid. Anal. Prev. 68 (7), 16–24. Naevestad, T.O., Phillips, R.O., Elvebakk, B., 2015. Traffic accidents triggered by drivers at work - a survey and analysis of contributing factors. Transp. Res. Part F: Traffic Psychol. Behav. 34, 94–107. Neal, A., Griffin, M.A., Hart, P.M., 2000. The impact of organizational climate on safety climate and individual behavior. Saf. Sci. 34 (1–3), 99–109. 1016/s0925-7535(00)00008-4. Nencini, A., Romaioli, D., Meneghini, A.M., 2016. Volunteer motivation and organizational climate: factors that promote satisfaction and sustained volunteerism in NPOs. Voluntas 27 (2), 618–639. Nordfjaern, T., Jorgensen, S., Rundmo, T., 2012. Cultural and socio-demographic predictors of car accident involvement in Norway, Ghana, Tanzania and Uganda. Saf. Sci. 50 (9), 1862–1872. O’connor, B.P., 2000. SPSS and SAS programs for determining the number of components using parallel analysis and velicer’s MAP test. Behav. Res. Methods Instrum. Comput. 32 (3), 396–402. Özkan, T., Lajunen, T., 2011. Person and environment. Chapter 14. Traffic Culture A2 Porter, Bryan E Handbook of Traffic Psychology. Academic Press, San Diego, pp. 179–192. Özkan, T., & Lajunen, T. Traffic culture and traffic climate scale. Unpublished manuscript. Poó, F.M., Taubman-Ben-Ari, O., Ledesma, R.D., Díaz-Lázaro, C.M., 2013. Reliability and validity of a Spanish-language version of the multidimensional driving style inventory. Transp. Res. Part F: Traffic Psychol. Behav. 17 (0), 75–87. 10.1016/j.trf.2012.10.003. Qu, W., Ge, Y., Jiang, C., Du, F., Zhang, K., 2014. The dula dangerous driving index in China: an investigation of reliability and validity. Accid. Anal. Prev. 64, 62–68. Regmi, K., Naidoo, J., Pilkington, P., 2010. Understanding the processes of translation and transliteration in qualitative research. Int. J. Qual. Methods 9 (1), 16–26. Reichers, A.E., Schneider, B., 1990. Climate and culture: an evolution of constructs. In: Schneider, B. (Ed.), Organizational Climate and Culture. Jossey Bass, San Francisco, pp. 5–39. Richer, I., Bergeron, J., 2012. Differentiating risky and aggressive driving: further support of the internal validity of the dula dangerous driving index. Accid. Anal. Prev. 45, 620–627. Rundmo, T., Granskaya, J., Klempe, H., 2012. Traffic culture as symbol exchange - a cross-country comparison of Russia and Norway. Saf. Sci. 50 (5), 1261–1267. http://

References ACSNI, 1993. Advisory Committee on the Safety of Nuclear Installations, Study Group on Human Factors. Third Report: Organizing for safety. HMSO, London. af Wåhlberg, A., Barraclough, P., Freeman, J., 2017. Personality versus traffic accidents: meta-analysis of real and method effects. Transp. Res. Part F: Traffic Psychol. Behav. 44, 90–104. Amponsah-Tawiah, K., Mensah, J., 2016. The impact of safety climate on safety related driving behaviors. Transp. Res. Part F: Traffic Psychol. Behav. 40, 48–55. Atchley, P., Shi, J., Yamamoto, T., 2014. Cultural foundations of safety culture: a comparison of traffic safety culture in China, Japan and the United States. Transp. Res. Part F: Traffic Psychol. Behav. 26, 317–325. Benfield, J.A., Szlemko, W.J., Bell, P.A., 2007. Driver personality and anthropomorphic attributions of vehicle personality relate to reported aggressive driving tendencies. Personal. Individ. Differ. 42 (2), 247–258. Bentler, P.M., Bonett, D.G., 1980. Significance tests and goodness of fit in the analysis of covariance structures. Psychol. Bull. 88 (3), 588. Beus, J.M., Payne, S.C., Bergman, M.E., Arthur Jr., W., 2010. Safety climate and injuries: an examination of theoretical and empirical relationships. J. Appl. Psychol. 95 (4), 713. Bhattacharya, Y., 2015. Measuring safety culture on ships using safety climate: a study among Indian officers. Int. J. e-Navig. Marit. Econ. 3, 51–70. Brislin, R.W., 1980. Translation and content analysis of oral and written material. Handb. Cross-Cult. Psychol. 2 (2), 349–444. Browne, M.W., Cudeck, R., Bollen, K.A., Long, J.S., 1993. Alternative ways of assessing model fit. Sage Focus Ed. 154 136-136. Burtăverde, V., Chraif, M., Aniţei, M., Mihăilă, T., 2016. The incremental validity of the dark triad in predicting driving aggression. Accid. Anal. Prev. 96, 1–11. Carr, J.Z., Schmidt, A.M., Ford, J.K., DeShon, R.P., 2003. Climate perceptions matter: a meta-analytic path analysis relating molar climate, cognitive and affective states, and individual level work outcomes. J. Appl. Psychol. 88 (4), 605. Chen, C.-F., 2009. Personality, safety attitudes and risky driving behaviors—evidence from young Taiwanese motorcyclists. Accid. Anal. Prev. 41 (5), 963–968. http://dx. Cheyne, A., Cox, S., Oliver, A., Tomás, J.M., 1998. Modelling safety climate in the prediction of levels of safety activity. Work Stress 12 (3), 255–271. Dahlen, E.R., Edwards, B.D., Tubré, T., Zyphur, M.J., Warren, C.R., 2012. Taking a look behind the wheel: an investigation into the personality predictors of aggressive driving. Accid. Anal. Prev. 45, 1–9. Dahlen, E.R., White, R.P., 2006. The big five factors, sensation seeking, and driving anger in the prediction of unsafe driving. Personal. Individ. Differ. 41 (5), 903–915. Dula, C.S., Ballard, M.E., 2003. Development and evaluation of a measure of dangerous, aggressive, negative emotional, and risky driving1. J. Appl. Soc. Psychol. 33 (2), 263–282. Gadd, S., Collins, A., 2002. Safety culture: A review of the literature. Health and Safety Laboratory. Gehlert, T., Hagemeister, C., Özkan, T., 2014. Traffic safety climate attitudes of road users in Germany. Transp. Res. Part F: Traffic Psychol. Behav. 26, 326–336. Griffin, M.A., Neal, A., 2000. Perceptions of safety at work: a framework for linking safety climate to safety performance, knowledge, and motivation. J. Occup. Health Psychol. 5 (3), 347. Guggenheim, N., Taubman–Ben-Ari, O., 2015. Ultraorthodox young drivers in Israel–driving through cultural lenses. Transp. Res. Part F: Traffic Psychol. Behav. 33, 87–96. Guo, M., Wei, W., Liao, G., Chu, F., 2016. The impact of personality on driving safety among Chinese high-speed railway drivers. Accid. Anal. Prev. 92, 9–14. Haigney, D., Taylor, R., Westerman, S., 2000. Concurrent mobile (cellular) phone use and driving performance: task demand characteristics and compensatory processes. Transp. Res. Part F: Traffic Psychol. Behav. 3 (3), 113–121. Harris, P.B., Houston, J.M., Vazquez, J.A., Smither, J.A., Harms, A., Dahlke, J.A., Sachau, D.A., 2014. The prosocial and aggressive driving inventory (PADI): a self-report measure of safe and unsafe driving behaviors. Accid. Anal. Prev. 72, 1–8. Hoy, W.K., 1990. Organizational climate and culture: a conceptual analysis of the school workplace. J. Educ. Psychol. Consult. 1 (2), 149–168. s1532768xjepc0102_4. International Test Commission, 2010. International Test Commission Guidelines for Translating and Adapting Tests. John, O.P., Srivastava, S., 1999. The big five trait taxonomy: history, measurement, and theoretical perspectives. Handbook of Personality: Theory and Research, vol. 2. pp. 102–138.


Accident Analysis and Prevention 113 (2018) 213–223

Q. Zhang et al.

Warner, H.W., Aberg, L., 2006. Drivers’ decision to speed: a study inspired by the theory of planned behavior. Res. Part F: Traffic Psychol. Behav. 9 (6), 427–433. http://dx. Wickens, C.D., 1991. Processing resources and attention. Mult. -Task Perform. 1991, 3–34. Wilde, G.J.S., 1982. The theory of risk homeostasis: implications for safety and health. Risk Anal. 2 (4), 209–225. Wilde, G.J.S., 1988. Risk homostasis theory and traffic accidents - propositions, deductions and discussion of dissension in recent reactions. Ergonomics 31 (4), 441–468. Wills, A.R., Watson, B., Biggs, H.C., 2006. Comparing safety climate factors as predictors of work-related driving behavior. J. Saf. Res. 37 (4), 375–383. Zhang, H., Qu, W., Ge, Y., Sun, X., Zhang, K., 2017. Effect of personality traits, age and sex on aggressive driving: psychometric adaptation of the driver aggression indicators scale in China. Accid. Anal. Prev. 103, 29–36. Zhang, W., Huang, Y.H., Roetting, M., Wang, Y., Wei, H., 2006. Driver’s views and behaviors about safety in China - what do they NOT know about driving? Accid. Anal. Prev. 38 (1), 22–27. Zhang, W., Tsimhoni, O., Sivak, M., Flannagan, M.J., 2010. Road safety in China: analysis of current challenges. J. Saf. Res. 41 (1), 25–30. 2009.12.003. Zohar, D., 1980. Safety climate in industrial organizations: theoretical and applied implications. J. Appl. Psychol. 65 (1), 96. Zohar, D., 2000. A group-level model of safety climate: testing the effect of group climate on microaccidents in manufacturing jobs. J. Appl. Psychol. 85 (4), 587. Zohar, D., 2011. Safety climate: conceptual and measurement issues. In: Quick, J.C., Tetrick, L.E. (Eds.), Handbook of Occupational Health Psychology, 2nd ed. American Psychological Association, Washington, DC, pp. 141–164. Zohar, D., Huang, Y.-h., Lee, J., Robertson, M., 2014. A mediation model linking dispatcher leadership and work ownership with safety climate as predictors of truck driver safety performance. Accid. Anal. Prev. 62, 17–25.

Schein, E.H., 1985. Organisational Culture and Leadership: A Dynamic View. San Francisco. . Schlembach, C., Furian, G., Brandstatter, C., 2016. Traffic (safety) culture and alcohol use: cultural patterns in the light of results of the SARTRE 4 study. Eur. Transp. Res. Rev. 8 (1). Schwebel, D.C., Severson, J., Ball, K.K., Rizzo, M., 2006. Individual difference factors in risky driving: the roles of anger/hostility, conscientiousness, and sensation-seeking. Accid. Anal. Prev. 38 (4), 801–810. Taubman-Ben-Ari, O., Katz-Ben-Ami, L., 2012. The contribution of family climate for road safety and social environment to the reported driving behavior of young drivers. Accid. Anal. Prev. 47, 1–10. The Ministry of Public Security of the People’s Republic of China. (2016, April 15). Retrieved July 4, 2016, from: c5196029.content.html. Tornros, J.E.B., Bolling, A.K., 2005. Mobile phone use - effects of handheld and handsfree phones on driving performance. Accid. Anal. Prev. 37 (5), 902–909. http://dx.doi. org/10.1016/j.aap.2005.04.007. Ulleberg, P., Rundmo, T., 2003. Personality, attitudes and risk perception as predictors of risky driving behaviour among young drivers. Saf. Sci. 41 (5), 427–443. http://dx. Velicer, W.F., 1976. Determining the number of components from the matrix of partial correlations. Psychometrika 41 (3), 321–327. bf02293557. Velicer, W.F., Eaton, C.A., Fava, J.L., 2000. Construct explication through factor or component analysis: a review and evaluation of alternative procedures for determining the number of factors or components. In: Goffin, R.D., Helmes, E. (Eds.), Problems and Solutions in Human Assessment: Honoring Douglas N. Jackson at Seventy. Springer US, Boston, MA, pp. 41–71. Wang, Y., Yao, L., Liu, L., Yang, X., Wu, H., Wang, J., Wang, L., 2014. The mediating role of self-efficacy in the relationship between big five personality and depressive symptoms among Chinese unemployed population: a cross-sectional study. BMC Psychiatry 14 (1), 61.