Psychometric adaptation of the driving anger expression inventory in a Chinese sample

Transportation Research Part F 33 (2015) 75–86

Contents lists available at ScienceDirect

Transportation Research Part F journal homepage: www.elsevier.com/locate/trf

Psychometric adaptation of the driving anger expression inventory in a Chinese sample Yan Ge, Weina Qu ⇑, Qian Zhang, Wenguo Zhao, Kan Zhang Key Laboratory of Behavioral Science, Institute of Psychology, Chinese Academy of Sciences, Beijing, China

a r t i c l e

i n f o

Article history: Received 3 September 2014 Received in revised form 11 May 2015 Accepted 9 July 2015

Keywords: Driving anger expression Dangerous driving Reliability and validity

a b s t r a c t The purpose of this study was to assess the psychometric properties and the factorial structure of the Driving Anger Expression Inventory (DAX) in a Chinese sample. We also explored the relationships among driving anger expression, general anger expression, and driving outcomes. Three hundred and fifty-eight drivers completed the Chinese version of the DAX, the Anger Expression Scale (AX), the Dula Dangerous Driving Index (DDDI) and a questionnaire about several types of traffic violations. A confirmatory factor analysis of the Chinese DAX yielded a four-factor solution with 20 items. This solution showed the best goodness of fit of the data and acceptable reliability. The validity of the revised DAX was also verified. The aggressive expression forms were positively correlated with dangerous driving behaviors. Using the vehicle to express anger was associated with fines. The aggressive forms were also positively correlated with general anger expression-out and negatively correlated with general anger control. The adaptive expression of anger was positively correlated with anger control but negatively correlated with dangerous driving behaviors, penalty points and fines. Furthermore, young drivers (<30 years old) reported more personal and physical aggressive expressions of anger than other drivers. Gender differences were only found in some age groups. Thus, the revised DAX was confirmed to be a reliable and valuable instrument to measure forms of driving anger expression in traffic environments in China. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction As a new automobile consumption market, China has witnessed a dramatic increase in the number of vehicles and drivers on its roads over the past 10 years. According to the 2011 report of China Road Traffic Accidents Statistics (CRTAS, 2011), the number of vehicles in China increased from 18 million in 2001 to over 105 million in 2011. Over the same period, the number of drivers increased from 42 million to 173 million. The rapid increase in the number of drivers has resulted in a large number of novice drivers whose driving style differs from that of US drivers. In particular, Chinese drivers concentrate on driving skills and capabilities, whereas US drivers concentrate on practical driving safety guidelines. For example, Chinese drivers seldom use running lights during rainy or snowy weather, and less than half of drivers use turn signals to indicate their intention to change lanes (Zhang, Huang, Roetting, Wang, & Wei, 2006). Therefore, the traffic environment in China differs from that in other countries. Chinese drivers are involved in a greater number of traffic accidents than drivers in the US and

⇑ Corresponding author at: 16 Lincui Road, Chaoyang District, Beijing 100101, China. Tel.: +86 10 64836956; fax: +86 10 64836047. E-mail address: [email protected] (W. Qu). http://dx.doi.org/10.1016/j.trf.2015.07.008 1369-8478/Ó 2015 Elsevier Ltd. All rights reserved.

76

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

Japan (Atchley, Shi, & Yamamoto, 2014; Zhang, Tsimhoni, Sivak, & Flannagan, 2010; Zhang et al., 2006). Thus, there is an acute need to systematically and deeply study the driving behavior of Chinese drivers. Anger is an emotion that drivers often experience while driving (Underwood, Chapman, Wright, & Crundall, 1999). Deffenbacher, Oetting, and Lynch (1994) provided a clear definition of driving anger as a situation-specific form of trait anger that refers to the propensity to become angry behind the wheel. They also found that high-anger drivers were more easily provoked by traffic situations and engaged in more aggressive and risky driving behaviors than low-anger drivers (Deffenbacher, Filetti, Richards, Lynch, & Oetting, 2003; Deffenbacher, Huff, Lynch, Oetting, & Salvatore, 2000). Similar results have been reported in many countries (Iversen & Rundmo, 2002; Lajunen & Parker, 2001; Li, Yao, Jiang, & Li, 2014; Sullman, 2006). According to the Cognitive Neoassociation Theory (Anderson & Bushman, 2002; Berkowitz, 1989, 1990), aversive events produce negative affect, which automatically stimulates various thoughts, memories, expressive motor reactions, and physiological responses related to fight and flight tendencies. In the driving environment, different situations may provoke anger, and drivers’ irritated thoughts and expressed reactions may differ. For example, two drivers may experience the same level of anger when they encounter a trigger situation, but they may address the same situation in different ways. Specifically, one angry driver may yell and attempt to force the trigger driver to the side of the road, whereas the other angry driver may tell himself that it is not worth getting mad at the trigger driver. These various methods of managing anger may result in different driving behaviors and violation outcomes (Deffenbacher, Lynch, Deffenbacher, & Getting, 2001; Deffenbacher, Lynch, Oetting, & Swaim, 2002; Deffenbacher, Oetting, Lynch, & Morris, 1996). Moreover, the two major dimensions of dangerous driving behavior, aggressive driving and risky driving, have different features (Dula & Ballard, 2003; Richer & Bergeron, 2012). Aggressive driving refers to any behavior in which a driver intends to physically or psychologically harm others (Dula & Ballard, 2003), such as using verbal expressions (e.g., yelling or cursing at another driver), physical expressions (e.g., leaving the vehicle and confronting or physically fighting with another driver) or their vehicle (e.g., flashing lights, honking horns, following close behind and cutting off another driver) to express anger (Deffenbacher et al., 2002). Risky driving refers to behaviors that do not intend to cause harm to others but potentially have negative outcomes because precautions are not taken. Such behaviors may be socially unacceptable or socially acceptable but dangerous (Turner, McClure, & Pirozzo, 2004; Willemsen, Dula, Declercq, & Verhaeghe, 2008). Risky driving behaviors include running red lights, weaving through traffic and speeding (Aarts & van Schagen, 2006; Dula & Ballard, 2003; Elvik, 2012; Rosen & Sander, 2009). Driving anger has been associated with both aggressive driving and risky driving (Bachoo et al., 2013; Jovanovic, Lipovac, Stanojevic, & Stanojevic, 2011), but different forms of anger expression may have different influences on these two dimensions of driving behavior. Considering the rapid motorization of China and the associated problems of traffic congestion and the resulting stress and frustration, an exploration of the effect of anger expression and control on Chinese drivers is warranted. To explore how people express their anger while driving, Deffenbacher et al. (2002) developed a self-report questionnaire, the Driving Anger Expression Inventory (DAX), to measure various forms of anger expression in traffic environments. The original version of the DAX contains 53 items, but the authors recommended a 49-item version with 4 factors because the fifth factor (4 items) demonstrated low reliability. These four factors are summarized as follows. First, the Verbal Aggressive Expression factor refers to expressing anger at the offending driver by yelling, name-calling or using non-verbal behaviors with verbal aggression (e.g., shaking one’s head). Second, the Personal Physical Aggressive Expression factor refers to drivers’ expression of anger with their body or through gestures, including shaking their fist and making hostile gestures, to scare other drivers. Third, the Use of Vehicle to Express Anger factor refers to using some part of the vehicle to express anger, such as flashing one’s lights or purposely blocking the other driver from performing the action that he/she wants to perform. The first three factors summarize the general aggressive expression index, which has been shown to be positively correlated with aggressive and risky driving behavior and some crash-related conditions (Deffenbacher, White, & Lynch, 2004). Fourth, the Adaptive/Constructive Expression factor refers to using a positive method to cope with driving anger. In this form of expression, the driver attempts to accept the frustrating situation and think of relieving ways to cope with it. This factor has been correlated negatively or has been uncorrelated with the first three factors and is unrelated to accident involvement (Deffenbacher et al., 2002; Sarbescu, 2012; Sullman, Stephens, & Kuzu, 2013). The DAX has been translated into several languages, including Turkish (Esßiyok, Yasak, & Korkusuz, 2007; Sullman et al., 2013), French (Villieux & Delhomme, 2010), Spanish (Herrero-Fernández, 2011) and Romanian (Sarbescu, 2012), and for several cultures, but the factors in each version differ. The original 49-item English version with four factors has been used widely without the removal of any items (Dahlen & Ragan, 2004; Deffenbacher et al., 2001; Jovanovic et al., 2011; Moore & Dahlen, 2008). By contrast, the Turkish version maintains the four factors but excludes two items from the 49-item version (Esßiyok et al., 2007; Sullman et al., 2013). In addition, Villieux and Delhomme (2010) removed all of the items in the Personal Physical Aggressive Expression factor of the French version because French drivers were unlikely to report the behavior described in this factor. The authors also modified items in the remaining factors according to some modification indices. Thus, the final French version of the DAX contains three factors with 11 items. Herrero-Fernández (2011) tested a 53-item Spanish version of the DAX and verified a five-factor model with good fit in the Spanish culture. Sarbescu (2012) removed some items and combined the verbal and physical factors into one factor to produce a 30-item Romanian version of the DAX that includes three factors and reached an acceptable level of model fit. Sullman (2015) verified a three-factor version without the Physical Aggressive Expression factor in a sample of New Zealand drivers. More importantly, Stephens and Sullman (2014) developed two short versions of the DAX, which can more easily be combined with other questionnaires and requires a smaller sample size than the original version.

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

77

According to the guidelines of the International Test Commission (International Test Commission (ITC), 2010), language and cultural differences should be fully considered when developing a questionnaire. Previous studies of driving scales have highlighted the importance of driving style and habits, traffic regulations and traffic environments when adapting a version of a questionnaire to a target population (Lajunen, Parker, & Summala, 2004; Özkan, Lajunen, Chliaoutakis, Parker, & Summala, 2006). The DAX has been shown to be a useful tool for evaluating the expression of driving anger in many countries, and its structure and items have been adjusted based on target populations (Esßiyok et al., 2007; Herrero-Fernández, 2011; Sarbescu, 2012; Sullman et al., 2013; Villieux & Delhomme, 2010). However, in consideration of the specificity of the traffic environment in China, the psychometric properties of the DAX must be adapted to Chinese drivers. Additionally, individual differences have an important effect on driving anger expression. For example, age differences have been found in many countries. Specifically, younger drivers score higher than older drivers on driving aggression expression (Esßiyok et al., 2007; Herrero-Fernández, 2011; Sarbescu, 2012). The results concerning gender difference have not been consistent across studies. In most studies, male drivers reported higher scores for using their body or vehicle to express their anger than female drivers, whereas female drivers reported more adaptive/constructive expression than male drivers (Dahlen & Ragan, 2004; Deffenbacher et al., 2002, 2003, 2004; Esßiyok et al., 2007; Sullman, 2015). Studies have found opposite results concerning the Verbal Aggressive Expression factor. Dahlen and Ragan (2004) reported that female drivers scored higher than male drivers on Verbal Aggressive Expression, whereas other studies reported that male drivers scored higher than female drivers on this dimension (Sullman, 2015). Furthermore, no significant gender difference was found in a Spanish sample (Herrero-Fernández, 2011). The mechanism that underlies these results must be explored. Additionally, some studies have explored the relationship between driving experience and anger expression, but the results have not been consistent. Sullman (2015) reported that driving experience revealed small to medium negative correlations with Verbal Aggressive Expression and Use of Vehicle to Express Anger and a small positive correlation with Adaptive/Constructive Expression in a New Zealand sample. A separate study by this author and his colleague showed that driving experience was negatively correlated with Verbal Aggressive Expression, Use of Vehicle to Express Anger and Personal Physical Aggressive Expression among Turkish taxi drivers (Sullman et al., 2013). Stephens and Sullman (2014) found that driving experience was negatively correlated with Use of Vehicle to Express Anger and positively correlated with Adaptive/Constructive Expression. By contrast, no significant correlation with any dimension of the DAX was found in a French sample (Villieux & Delhomme, 2010). In summary, the main purpose of the current study was to assess the reliability and factorial structure of the DAX questionnaire in a Chinese sample. Moreover, this study verified the questionnaire’s convergent validity by assessing its relationship with the Dula Dangerous Driving Index (DDDI, Dula & Ballard, 2003), which is used to evaluate several dangerous driving behaviors, including aggressive driving, risky driving, negative cognitive/emotional driving and drunk driving. We also assessed the criterion validity of the DAX using self-reports of accidents and violations. Moreover, we tested the relationship between driving anger expression and general anger expression. For this evaluation, general anger expression was measured with the Anger Expression Scale (AX) from the State-Trait Anger Expression Inventory 2 (STAXI-2, Spielberger, 1999). Finally, we investigated differences in driving anger expression according to demographic characteristics. 2. Methods 2.1. Participants The questionnaires were distributed to 473 licensed drivers who were randomly selected from parking lots, gas stations, restaurants, shopping centers and office buildings in Beijing, China. All participants completed the questionnaire voluntarily and anonymously. Overall, 382 (80.7% response rate) questionnaires were returned, and 24 were excluded because of incompleteness or inconsistencies. The final sample included 358 drivers, aged from 19 to 56 years old (M = 32.79 years, SD = 8.54), including 215 males and 143 females. The demographics of the final sample are shown in Table 1. 2.2. Instruments 2.2.1. The Driving Anger Expression Inventory (DAX) The DAX is a self-report scale that was developed to differentiate the ways in which drivers express anger when they are on the road (Deffenbacher et al., 2002). The original version of the four-factor DAX included 49 items. The reliability of each factor was good, i.e., Verbal Aggressive Expression (Ver; 12 items, a = 0.88); Personal Physical Aggressive Expression (Phy; 11 items, a = 0.81); Use of the Vehicle to Express Anger (Veh; 11 items, a = 0.86); and Adaptive/Constructive Expression (Adp; 15 items, a = 0.90). The participants were asked to rate the frequency with which they express each item using a 4-point Likert scale that ranged from 1 (‘‘almost never’’) to 4 (‘‘almost always’’). The Chinese version of the DAX used in this study was translated following the translation/back-translation procedure suggested by Brislin (1980) and Regmi, Naidoo, and Pilkington (2010). First, two researchers independently translated the English version of the DAX into Chinese. Then, all of the authors compared and discussed their translations with consideration of accuracy, fluency and adaptation to the Chinese driving culture to create a single version. Second, this draft was back-translated by a professional translator who was proficient at English–Chinese translation. Differences between the

78

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86 Table 1 Participant demographics (n = 358). N Age groups by gender 19–30 years old Males Females

Percent (%)

99 70

27.65 19.56

31–40 years old Males Females

73 45

20.39 12.57

41–56 years old Males Females

43 28

12.01 7.82

26 64 235 30 3

7.26 17.88 65.64 8.38 0.84

43 93 56 98 68

12.01 25.98 15.64 27.38 18.99

68 112 108 70

18.99 31.29 30.17 19.55

Education Less than high school Secondary school College Postgraduate or above Unknown Driving years 61 year 2–3 years 4–5 years 6–10 years >10 years Annual mileage (km) 65000 5001–10,000 10,001–20,000 >20,000

original items and the back-translated items were discussed to ensure the precision of the Chinese version. Finally, four drivers were recruited to evaluate the translated draft and note unclear items. We finalized the scale by taking the drivers’ suggestions into account. 2.2.2. The Dula Dangerous Driving Index (DDDI) The DDDI was developed by Dula and Ballard (2003) to assess individual propensities for dangerous driving. Qu, Ge, Jiang, Du, and Zhang (2014) validated the Chinese version of the DDDI with a four-factor model, which includes Aggressive Driving (AD, 7 items, a = 0.78), Negative Cognitive/Emotional Driving (NCED, 9 items; a = 0.80), Risky Driving (RD, 10 items; a = 0.78) and Drunk Driving (DD, 2 items; a = 0.63). The internal consistency of the total scale was shown to be excellent (Cronbach’s a = 0.90). Participants rate the items in this index using a 5-point scale (1 = ‘‘never’’ to 5 = ‘‘always’’) according to the frequency with which they express each dangerous driving behavior. In the current study, the Chinese version of the DDDI (Qu et al., 2014) was used to assess dangerous driving behavior. 2.2.3. The Anger Expression (AX) scale The Anger Expression (AX) scale was constructed to assess individual differences in retaining anger or expressing anger as aggressive behavior. The AX is a subscale of the State-Trait Anger Expression Inventory 2 (STAXI-2, Spielberger, 1999). The AX scale comprises 32 items that measure the following four factors: Anger Expression-In (AI), Anger Expression-Out (AO), Anger Control-In (ACI) and Anger Control-Out (ACO). AI is defined as how often an individual experiences but suppresses angry feelings. AO is defined as how often an individual expresses angry feelings with verbally or physically aggressive behavior. ACI is defined as how often an individual attempts to calm down by reducing the intensity of the suppressed anger. Finally, ACO is defined as how often an individual attempts to regulate and prevent the external expression of anger. Each factor contains 8 items, and participants rate each item on a four-point scale (1 = ‘‘almost never’’; 4 = ‘‘almost always’’). The Chinese version used in the current study was verified by Liu and Gao (2012), and each factor demonstrated acceptable reliability, with Cronbach’s a coefficients between 0.62 and 0.87. 2.2.4. Demographic questionnaire In addition to the basic sociodemographic variables that are required in typical traffic psychology research (i.e., gender, age, level of education, driving years and estimated average annual mileage), the current participants were asked to report the number of accidents that they had been involved in over the previous three years and the penalty points and fines that they had received in the last year. More specifically, they were asked to state the frequency with which they had been

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

79

involved in the following traffic violations in the last month: speeding, violating traffic signs or lines, driving while intoxicated and running a red light. 2.3. Procedure Research interns distributed the questionnaires in parking lots, gas stations, restaurants, shopping centers and office buildings in Beijing. The drivers participated in this study voluntarily and were assured that their information would be kept confidential and used only for scientific research. Drivers who completed the questionnaires were given 20 yuan RMB. All of the forms were completed in approximately 20 min. The study was approved by the Institutional Review Board of the Institute of Psychology, Chinese Academy of Sciences. 3. Results 3.1. Psychometric properties of the original DAX scale In terms of the original DAX scale, the drivers in the current study reported a large amount of Adp (evidenced by the average score per item of 2.77 on a 4-point scale), a moderate amount of Phy and Veh, and some Ver Expression. Table 2 shows the means and standard deviations (SD) for each item of the original DAX scale. The item-scale correlation, skewness, and kurtosis of each item are also reported. 3.2. Confirmatory factor analysis (CFA) A CFA was conducted to analyze the internal structure of the DAX using Mplus 6.0. All items can be assumed to be normally distributed when skewness is less than 3 and kurtosis is less than 7 (Kline, 2011; West, Finch, & Curran, 1995). However, the data does not obey the multivariate normal distribution for multivariate kurtosis (Finney & DiStefano, 2006). The consequences of using ML under conditions of severe non-normality lead to inflating the risk of Type I error (West et al., 1995). The most commonly used estimator of robust Maximum Likelihood Estimation (MLE) is Mean-Adjusted Maximum Likelihood (MLM) in the Mplus program (Brown, 2014). For CFA, a Steiger–Lind root mean square error of approximation (RMSEA) value under 0.08 indicates good model fit (Browne, Cudeck, Bollen, & Long, 1993), the acceptable value of v2/df index is lower than 5 (Wheaton, 1977). Moreover, the Joreskog–Sorbom goodness-of-fit index (GFI P 0.90 indicates a good model fit), Bentler’s comparative fit index (CFI P 0.90) and the Tucker–Lewis Index (TLI P 0.90) were examined to compare the models and improve the model fit during the iterative CFA process (Bentler & Bonett, 1980). Akaike’s information criterion (AIC) is the main index that is used to indicate model parsimony; a lower AIC value suggests a more parsimonious model. The original 49-item four-factor model was tested first. The internal consistency (Cronbach’s a) of the four subscales of the DAX was satisfactory, i.e., 0.90 for Adp, 0.91 for Phy, 0.87 for Veh and 0.83 for Ver. However, the indices of model fit were not acceptable, except the RMSEA was under 0.08. The model fit details are shown in Table 3. Considering that the short form was easy to use, we tested the two short forms of the DAX developed by Stephens and Sullman (2014). The 25-item model showed an improved fit, but none of the indices reached a satisfactory level. The 15-item model showed a better fit, and all of the indices met the satisfactory level, but the reliabilities of two subscales were not acceptable (Ver, a = 0.57; Veh, a = 0.52; Phy, a = 0.79; Adp, a = 0.80) in the current sample. To develop a short version of the DAX that has a satisfactory reliability and stable internal structure, we revised this scale based on the item-total correlation coefficients (ITCs) for each subscale (Inoue et al., 2014; Stanton, Sinar, Balzer, & Smith, 2002; Yasak & Esiyok, 2009). The 5 items with the highest ITCs in each subscale were selected to compose the revised version of the DAX with 20 items. This version revealed acceptable fit levels for all indices. Furthermore, the variances of the error terms were analyzed through the modification indices (MIs), which demonstrated that some variables were redundant (Kaplan, 1989). Lagrange Multiplier Tests revealed that three error-pairs should be covaried (47–48: MI = 34.63, r = 0.65; 29–30: MI = 19.84, r = 0.62; 20–21: MI = 15.13, r = 0.73), following the criterion of MIs equal to or higher than 15.00. After allowing the errors of these conceptually similar items to covary, the final model reached a good level of model fit. Table 3 shows the corresponding goodness-of-fit indices. 3.3. Psychometric properties of the revised DAX scale The revised DAX scale reached acceptable internal consistency: Phy (M = 1.32, SD = 0.53; a = 0.87); Ver (M = 1.82, SD = 0.57; a = 0.76); Veh (M = 1.62, SD = 0.55; a = 0.79); Adp (M = 2.77, SD = 0.69; a = 0.84). The correlations between each subscale are shown in Table 4. Phy, Ver and Veh were positively correlated with each other and negatively correlated with Adp. The factors that were derived from the original DAX (49-items) and revised DAX (20-items) were strongly correlated (Phy: r = 0.96, p < 0.001; Ver: r = 0.92, p < 0.001; Veh: r = 0.94, p < 0.001; Adp: r = 0.92, p < 0.001). The following analysis used the revised DAX scale with 20 items.

80

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

Table 2 The descriptive statistics of the DAX items and subscales. DAX items

M(SD)

Adaptive/Constructive Expression 49 I pay even closer attention to the other’s driving to avoid accidents. 23 I pay even closer attention to being a safe driver. 25 I think things through before I respond. 42 I just try to accept that there are bad drivers on the road. 26 I try to think of positive solutions to deal with the situation. 36 I tell myself it’s not worth getting involved in. 30 I decide not to stoop to their level. 29 I tell myself it’s not worth getting all mad about. 45 I just try to accept that there are frustrating situations while driving. 48 I tell myself to ignore it. 47 I think about things that distract me from the frustration on the road. 35 I try to think of positive things to do. 24 I think about things that distract me from thinking about the other driver. 44 I do things like take deep breaths to calm down. 32 I turn on the radio or music to calm down.

2.77(0.58) 3.03(0.94) 3.01(1.00) 2.92(0.86) 2.90(0.91) 2.89(0.85) 2.84(0.90) 2.81(0.91) 2.79(0.88) 2.77(0.92) 2.74(0.87) 2.69(0.87) 2.58(0.90) 2.57(0.92) 2.55(0.94) 2.44(0.99)

Verbal 37 6 43 39 31 11 14 40 28 5 9 38

r

Skew

Kurtosis

Loading

0.60** 0.55** 0.67** 0.68** 0.65** 0.73⁄⁄ 0.71⁄⁄ 0.71⁄⁄ 0.68** 0.72⁄⁄ 0.72⁄⁄ 0.63** 0.59** 0.56** 0.49**

0.29 0.77 0.76 0.52 0.41 0.45 0.42 0.41 0.36 0.34 0.36 0.30 0.19 0.13 0.13 0.05

0.19 0.23 0.48 0.32 0.68 0.33 0.56 0.59 0.52 0.68 0.49 0.53 0.71 0.81 0.85 1.05

0.54 0.49 0.64 0.66 0.62 0.71 0.71 0.70 0.66 0.71 0.70 0.58 0.55 0.49 0.40

1.86(0.48) 2.10(0.95) 2.09(0.83) 2.03(0.90) 2.02(0.80) 1.98(0.81) 1.95(0.77) 1.88(0.75) 1.86(0.83) 1.71(0.81) 1.59(0.76) 1.57(0.80) 1.54(0.74)

0.50** 0.54** 0.56** 0.65⁄⁄ 0.56** 0.68⁄⁄ 0.57** 0.69⁄⁄ 0.65⁄⁄ 0.57** 0.63⁄⁄ 0.60**

0.35 0.36 0.47 0.57 0.36 0.48 0.53 0.60 0.67 0.93 1.19 1.38 1.31

0.49 0.91 0.27 0.45 0.48 0.33 0.01 0.14 0.22 0.19 0.88 1.33 1.26

0.41 0.44 0.45 0.54 0.45 0.65 0.48 0.64 0.65 0.57 0.63 0.62

Use of the Vehicle to Express Anger 27 I drive a lot faster than I was. 3 I drive a little faster than I was. 33 I flash my lights at the other driver. 4 I try to cut in front of the other driver. 46 I slow down to frustrate the other driver. 19 I leave my brights on in the other driver’s rear view mirror. 15 I speed up to frustrate the other driver. 16 I purposely block the other driver from doing what he/she wants to do. 22 I do to other drivers what they did to me. 7 I follow right behind the other driver for a long time. 2 I drive right up on the other driver’s bumper.

1.63(0.48) 1.98(0.79) 1.87(0.73) 1.79(0.78) 1.75(0.76) 1.63(0.81) 1.59(0.76) 1.57(0.74) 1.55(0.76) 1.52(0.74) 1.44(0.68) 1.27(0.54)

0.59** 0.57** 0.67⁄⁄ 0.71⁄⁄ 0.62** 0.67** 0.69⁄⁄ 0.72⁄⁄ 0.66** 0.69⁄⁄ 0.60**

0.86 0.52 0.68 0.68 0.79 1.22 1.23 1.18 1.31 1.43 1.65 2.06

0.26 0.10 0.51 0.21 0.23 0.91 1.10 0.89 1.18 1.60 2.68 3.82

0.49 0.46 0.58 0.64 0.58 0.61 0.65 0.74 0.64 0.68 0.64

Personal Physical Aggressive Expression 10 I roll down the window to help communicate my anger. 18 I go crazy behind the wheel. 34 I make hostile gestures other than giving the finger. 12 I shake my fist at the other driver. 41 I try to get out of the car and have a physical fight with the other driver. 13 I stick my tongue out at the other driver. 8 I try to get out of the car and tell the other driver off. 20 I try to force the other driver to the side of the road. 21 I try to scare the other driver. 1 I give the other driver the finger. 17 I bump the other driver’s bumper with mine.

1.36(0.48) 1.60(0.77) 1.46(0.77) 1.37(0.71) 1.35(0.62) 1.35(0.71) 1.34(0.67) 1.33(0.65) 1.33(0.65) 1.28(0.61) 1.27(0.50) 1.27(0.65)

0.70** 0.68** 0.76⁄⁄ 0.74⁄⁄ 0.71** 0.65** 0.71** 0.80⁄⁄ 0.80⁄⁄ 0.59** 0.80⁄⁄

1.71 1.19 1.75 2.04 1.72 2.13 2.05 2.12 1.96 2.30 1.81 2.58

2.34 0.91 2.48 3.69 2.35 4.01 3.77 4.38 3.03 5.13 3.33 6.19

0.65 0.63 0.72 0.71 0.68 0.58 0.69 0.80 0.79 0.55 0.78

Aggressive Expression I shake my head at the other driver. I make negative comments about the other driver aloud. I think things like ‘Where did you get your license?’ I make negative comments about the other driver under my breath. I swear at the other driver under my breath. I glare at the other driver. I call the other driver names under my breath. I give the other driver a dirty look. I swear at the other driver aloud. I call the other driver names aloud. I yell questions like ‘Where did you get your license?’ I yell at the other driver.

Note: Items selected in the revised model are in bold. ** p < 0.01.

3.4. Validity analysis of the revised DAX scale The correlations between the subscales of the DAX and the other questionnaires were analyzed and are shown in Table 4. The DAX was correlated with general anger expression, as measured by the AX scale. Two forms of control-anger expression were negatively related to Phy, Ver and Veh but positively related to the Adp dimension. Anger-out expression was positively correlated with the three aggressive anger expressions but negatively correlated with adaptive expression. However, the anger-in expression only showed correlations with the Adp and Ver subscales of the DAX. The validity of the revised DAX was analyzed based on its correlation with the DDDI and traffic accidents and violations. The first three subscales (Phy, Ver and Veh) were positively correlated with the DDDI, whereas Adp was negatively correlated

81

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86 Table 3 Goodness-of-fit indices for five models of the DAX scale using MLM method with Mplus. Bollen–Stine

Model Model Model Model Model

1 2 3 4 5

(49 (25 (15 (20 (20

items) items) items) items) items)

v2

v2/df

p

2186.96 471.48 112.73 274.07 228.68

1.95 1.75 1.34 1.67 1.40

0.000 0.000 0.020 0.000 0.001

CFI

TLI

RMSEA(90% C. I.)

AIC

0.82 0.92 0.98 0.95 0.97

0.82 0.91 0.97 0.94 0.96

0.05(0.05–0.06) 0.05(0.04–0.05) 0.03(0.01–0.05) 0.04(0.03–0.05) 0.03(0.02–0.04)

35171.95 18263.77 10934.90 13793.60 13732.69

Note: Model 1 was the original model; models 2 and 3 were the two short forms of the DAX developed by Stephens and Sullman (2014); model 4 included the five items with the highest item-subscale correlation for each subscale; and model 5 allowed three error-pairs for conceptually similar items to covary based on model 4.

with the DDDI. Adp was also negatively correlated with the points and fines received in the last year, whereas Veh was positively correlated with the fines received in the last year. No significant correlation was found between accidents and the DAX. The correlations between demographic variables and the DAX were also calculated. Age was only negatively correlated with Phy. Mileage had significantly positive correlations with Phy and Veh and a negative correlation with Adp. However, there were no significant correlations between gender and the DAX. To explore the relationship between the DAX and traffic violations, the participants were divided into two groups (offenders and non-offenders) based on their self-reported traffic regulation violations in the last month, including speeding, violating traffic signs or lines, driving while intoxicated and running a red light. For example, the speeding offenders were drivers who reported one or more instances of speeding in the last month, whereas drivers who did not speed in the last month were defined as non-offenders. Differences between the speeding and non-speeding groups were found for Phy, Ver and Veh. The offenders who violated traffic signs or lines reported using vehicles to express their anger more frequently than did non-offenders. In addition, the offenders who had driven intoxicated reported lower scores on Adp than non-offenders. Table 5 shows the details of this analysis. 3.5. Differences by age and gender To explore differences by gender and age in the expression of anger, a 2 (gender)  3 (age) multivariate analysis of variance (MANOVA) and Bonferroni post hoc tests (Field, 2009) were conducted with annual mileage as a covariate (Table 6). The drivers were divided into the following three age groups: 19–30, 31–40 and 41–56 years. The results showed that the main effect of age on Phy was significant. Further analysis indicated that drivers under 30 years of age exhibited significantly more Phy than the other two groups, whereas no significant differences were found between the groups aged 31–40 and 41– 60 years. Annual mileage had a significant effect on Adp (F = 9.02, p < 0.001). There was no significant main effect of gender on the four subscales. However, a significant interaction between age and gender was found on the Ver subscale, but the effect size was small (F = 3.80, p < 0.05, g2 = 0.02). The simple effect analysis suggested that females were more likely to express their anger verbally than males in the 31- to 40-year-old group (F = 4.96, p < 0.05), whereas males reported more aggressive expression using their vehicle than females in the 41- to 56- year-old group (F = 4.89, p < 0.05). 4. Discussion The main objective of this study was to assess the psychometric properties and the factorial structure of the DAX in a Chinese sample. Overall, the revised version of the DAX showed good reliability and validity, and our results highlighted correlations between general anger expression and driving anger expression. According to our results, the Chinese DAX showed adequate reliability and a stable structure. The CFA confirmed a good fit of the four-factor model with 20 items, as was found for the American and Turkish versions of the DAX (Deffenbacher et al., 2002; Sullman et al., 2013), although some items were removed to allow for an adequate fit. All of the factors showed acceptable internal consistency and could be compared with other versions of the DAX (Deffenbacher et al., 2002; Sullman et al., 2013). According to the current study, Chinese drivers expressed their anger in four ways, verbally, physically, through the vehicle and adaptively. The Adaptive/Constructive Expression of anger was the most commonly used way to address anger, which is similar to what has been found for drivers in other countries (Esßiyok et al., 2007; Sarbescu, 2012; Sullman et al., 2013; Villieux & Delhomme, 2010). Although the number of items in the Adaptive/Constructive Expression factor is not the same for all countries’ versions of the DAX, the mean score of this factor was the highest of the four factors. These results indicate that drivers are more likely to cognitively reframe the anger trigger and attempt to be safe drivers. The scores for the three aggressive driving anger expression forms were lower than the Adaptive/Constructive Expression factor score. Specifically, the Verbal Aggressive Expression factor showed the lowest score, which was also observed for the French and Turkish versions of the scale (Esßiyok et al., 2007; Sarbescu, 2012; Sullman et al., 2013; Villieux & Delhomme, 2010).

82

Table 4 Correlations among the revised DAX, the AX, the DDDI, accidents, penalty points and fines (n = 358). 1 Age Gender Mileage Phy Ver Veh Adp AI AO ACI ACO DDDI -NCED -AD -RD -DD Acc. Points Fines

2 – 0.03 0.06 0.11* 0.09 0.06 0.03 0.02 0.15** 0.04 0.02 0.10 0.05 0.10 0.10* 0.12* 0.03 0.05 0.01

3 – 0.16** 0.09 0.04 0.06 0.08 0.01 0.01 0.03 0.01 0.09 0.00 0.11* 0.12* 0.09 0.03 0.06 0.09

4

– 0.10 0.01 0.05 0.17** 0.01 0.10 0.01 0.01 0.03 0.02 0.08 0.01 0.01 0.02 0.05 0.03

5

– 0.55** 0.68** 0.33** 0.08 0.55** 0.31** 0.35** 0.56** 0.35** 0.62** 0.51** 0.59** 0.06 0.03 0.04

6

– 0.57** 0.16** 0.11* 0.46** 0.15** 0.20** 0.57** 0.53** 0.56** 0.48** 0.32** 0.09 0.02 0.08

7

– 0.36** 0.01 0.58** 0.33** 0.36** 0.69** 0.53** 0.71** 0.62** 0.49** 0.10 0.07 0.21**

8

– 0.16** 0.32** 0.56** 0.53** 0.33** 0.21** 0.34** 0.32** 0.33** 0.07 0.11* 0.15**

9

– 0.05 0.42** 0.43** 0.08 0.05 0.04 0.11* 0.05 0.16** 0.14** 0.17**

10

– 0.35** 0.41** 0.69** 0.58** 0.70** 0.61** 0.44** 0.05 0.02 0.11*

– 0.85** 0.31** 0.20** 0.33** 0.29** 0.27** 0.11* 0.11* 0.21**

11

– 0.34** 0.26** 0.35** 0.31** 0.26** 0.11* 0.08 0.15**

12

13

14

15

16

17

18

– 0.88** 0.89** 0.93** 0.65** 0.08 0.12* 0.24**

– 0.70** 0.72** 0.39** 0.09 0.10 0.23**

– 0.75** 0.59** 0.08 0.08 0.18**

– 0.62** 0.05 0.14** 0.25**

– 0.01 0.05 0.12*

0.30** 0.34**

0.61**

Notes: Gender: 1 = male, 2 = female; Mileage = annual mileage; Phy = Personal Physical Aggressive Expression; Ver = Verbal Aggressive Expression; Veh = Use of Vehicle to Express Anger; Adp = Adaptive/ Constructive Expression; AI = Anger Expression-In; AO = Anger Expression-Out; ACI = Anger Control-In; ACO = Anger Control-Out; DDDI = Total score of Dula Dangerous Driving Index; NCED = Negative Cognitive/ Emotional Driving; AD = Aggressive Driving; RD = Risky Driving; DD = Drunk Driving; Acc. = Accidents in the last three years; Points = Penalty points received in the last year; Fines = Fines received in the last year. * p < 0.05. ** p < 0.01.

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

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

83

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86 Table 5 Comparison of the DAX factor subscales in terms of traffic offenses. Phy

Ver

Veh

Adp

M(SD)

F

g2

M(SD)

F

g2

M(SD)

F

g2

M(SD)

F

g2

1.28(0.51) 1.48(0.59)

7.63**

0.02

2.04(0.61) 1.78(0.56)

11.11**

0.03

1.87(0.61) 1.56(0.52)

17.10**

0.05

2.67(0.64) 2.81(0.70)

2.13

0.01

Violating traffic signs or lines Offenders (n = 79) 1.37(0.51) Non-offenders (n = 271) 1.31(0.54)

0.90

0.00

1.93(0.57) 1.80(0.57)

3.07

0.01

1.75(0.56) 1.58(0.54)

5.40*

0.02

2.67(0.70) 2.81(0.68)

2.60

0.01

Running a red light Offenders (n = 30) Non-offenders (n = 321)

1.16

0.00

1.89(0.63) 1.82(0.57)

0.32

0.00

1.75(0.55) 1.61(0.55)

1.92

0.01

2.45(0.74) 2.80(0.68)

7.59**

0.02

Speeding Offenders (n = 65) Non-offenders (n = 279)

1.42(0.56) 0.31(0.53)

Notes: Phy = Personal Physical Aggressive Expression; Ver = Verbal Aggressive Expression; Veh = Use of Vehicle to Express Anger; Adp = Adaptive/ Constructive Expression. * p < 0.05. ** p < 0.01.

Table 6 Multivariate differences by age and gender for each type of driving anger expression (n = 358). 18–30 Male (n = 99) Phy Ver Veh Adp

1.39(0.55) 1.95(0.62) 1.68(0.51) 2.65(0.70)

31–40 Female (n = 90) 1.41(0.62) 1.82(0.55) 1.65(0.62) 2.76(0.66)

Male (n = 73) 1.31(0.51) 1.67(0.54) 1.58(0.54) 2.77(0.74)

41–60 Female (n = 45) 1.21(0.42) 1.88(0.55) 1.54(0.51) 2.94(0.71)

Male (n = 43) 1.29(0.51) 1.84(0.53) 1.73(0.58) 2.75(0.65)

Age Female (n = 28) 1.11(0.19) 1.66(0.51) 1.42(0.46) 3.04(0.57)

Age  Gender

F **

4.51 2.16 1.55 2.84

g2

F

g2

0.03 0.01 0.01 0.02

0.73 3.80* 1.55 0.15

0.00 0.02 0.01 0.00

Notes: Phy = Personal Physical Aggressive Expression; Ver = Verbal Aggressive Expression; Veh = Use of Vehicle to Express Anger; Adp = Adaptive/ Constructive Expression. * p < 0.05. ** p < 0.01.

The validity of the Chinese DAX was confirmed in two ways in our study. First, the convergent validity of the DAX was tested via the relationship between the DAX and the DDDI. The DDDI is a widely used questionnaire that mainly measures aggressive, risky and negative emotional or cognitive driving behaviors (Dula & Ballard, 2003; Iliescu & Sârbescu, 2013; Qu et al., 2014; Richer & Bergeron, 2012; Willemsen et al., 2008). The three aggressive anger expression forms of the DAX were positively correlated with all of the subscales of the DDDI, whereas the Adaptive/Constructive Expression factor was negatively correlated with all of the subscales of the DDDI. These results support the acceptable convergent validity of the Chinese DAX. Moreover, an interesting result should be mentioned. From a theoretical perspective, there is a clear difference between aggressive driving and risky driving (Dula & Ballard, 2003; Willemsen et al., 2008). However, the DAX subscales seem to correlate at approximately the same magnitude with these two subscales of the DDDI. Previous studies have shown that trait driving anger is associated with both aggressive driving and risky driving (Bachoo et al., 2013; Jovanovic et al., 2011). The current research implies that the way that drivers express their anger is correlated with both aggressive driving and risky driving. Based on the comprehensive model of aggressive driving (Lennon & Watson, 2011; Soole, Lennon, Watson, & Bingham, 2011), a driver perceives a trigger event on the road and makes cognitive and affective appraisals, which are the basis for his/her behavior choice. When encountering an anger-provoking situation, some drivers consider the other driver’s behavior as an insult and express their anger aggressively. Meanwhile, anger may increase the probability of risk taking (Abdu, Shinar, & Meiran, 2012), which may cause drivers to engage in risky driving behavior. Thus, the driver’s thoughts behind the wheel may be the main reason for his/her choice of behavior. This issue should be explored in future studies. Second, self-reported traffic accidents and violations were used to test the criterion validity of the Chinese DAX. Unfortunately, we did not find a significant correlation between aggressive expressions of driving anger and self-reported accidents. This result is in contrast to the results reported for the Romanian version of the scale, which showed a significant correlation between aggressive expression and accidents (Gonzalez-Iglesias, Antonio Gomez-Fraguela, & Angeles Luengo-Martin, 2012; Sarbescu, 2012). Similar to the results of the Turkish version, no significant correlation between aggressive expression and major or minor accidents was found in taxi drivers (Sullman et al., 2013). This result is not surprising because some studies have reported a connection between aggressive driving and accidents and crashes (Dahlen, Edwards, Tubré, Zyphur, & Warren, 2012), whereas others have not (Sullman et al., 2013; Van Rooy, Rotton, & Burns, 2006). However, we found a positive association between fines received in the last year and using the vehicle to express

84

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

anger. We also found a negative correlation between penalty points/fines received and the adaptive expression of anger. These results indicate that the DAX may be an effective instrument to identify potential offenders of traffic laws. The difference between traffic offenders and non-offenders also supported this conclusion. Drivers who engaged in speeding in the last month reported more physical and verbal expression of anger and more use of their cars to express anger than drivers who did not speed in the last month. Previous studies have shown that speeding is an important factor in drivers’ involvement in collisions (Aarts & van Schagen, 2006; Cooper, 1997), indicating that speeding may be used as a valid index to predict the dangerous outcomes of drivers. Furthermore, drivers who violated traffic signs or lines in the last month reported using their car to express anger more frequently than those who did not, and drivers who ran a red light reported less use of adaptive strategies to adjust their anger compared to non-offenders. These results imply that different violations may be correlated with different forms of anger expression and control. The underlying mechanisms of these phenomena should be explored in future studies. Additionally, we explored the relationship between driving anger expression and general anger expression. The results showed that all three of the aggressive forms of anger expression were positively correlated with anger expression-out and negatively correlated with anger control-in and anger control-out. The adaptive expression of driving anger showed opposite results. This expression type was positively correlated with anger control and anger expression-in and negatively correlated with anger expression-out. These results indicate that the general features of anger expression and control are strongly associated with driving anger expression. To our knowledge, this is the first study to assess the relationship between driving anger expression and general anger expression, which was measured using the AX subscale of the STAXI-2 (Spielberger, 1999). Previous studies have tested only the relationship between driving anger expression and trait anger (Dahlen & Ragan, 2004; Dula & Ballard, 2003), as measured by the trait anger scale, which is another subscale of the STAXI-2 (Spielberger, 1999). Regarding individual differences in demographic variables related to driving anger expression, the results obtained from various countries are not consistent (Deffenbacher et al., 2002; Esßiyok et al., 2007; Herrero-Fernández, 2011; Sarbescu, 2012; Stephens & Sullman, 2014; Sullman, 2015). We found an age difference in personal physical aggressive expression and adaptive expression. Younger drivers reported more physical aggressive expression of anger. This finding is similar to the results of the Spanish and Romanian versions of the scale (Herrero-Fernández, 2011; Sarbescu, 2012). However, young drivers reported more adaptive expression in Romania (Sarbescu, 2012). We did not find any age difference in using the car to express anger, although such a difference was found in the Turkish and Spanish versions (Esßiyok et al., 2007; Herrero-Fernández, 2011). The different driving climates or safety cultures of the countries may be one reason for these different results (Atchley et al., 2014). Concerning gender differences, the results are also inconsistent. In the current study, the main effect of gender was not significant, but we found that female drivers reported more verbal aggressive expression than male drivers in the 31- to 40-year-old group. Whereas some authors identified that males reported more verbal aggressive expression than females (Sullman, 2015), others found that females reported more verbal aggressive expression than males (Dahlen & Ragan, 2004). We found a gender difference of using the vehicle to express anger in the older group (41–56 years old). This result is similar to that of a previous study (Deffenbacher et al., 2004), whereas other studies failed to find such a difference (Herrero-Fernández, 2011; Sarbescu, 2012). Age may play an important role in these results. The drivers in Dahlen and Ragan’s research were undergraduate students (Mean age = 19), but the participants in Sullman’s research (2015) covered a wide age range. We only found gender differences in specific age groups. Future research should deeply explore the interaction between age and gender on anger expression. There are some limitations to this study. One limitation is that the participants did not constitute a representative sample of Chinese drivers because they were sampled conveniently, thus limiting the applicability of our conclusions to the entire population of drivers in China. A second limitation is that the data reported in this study were acquired through self-reporting. Researchers have demonstrated that self-reported data are as useful as archival driving data (Arthur et al., 2005) and data from field studies (Barling, Kelloway, & Iverson, 2003; Zacharatos, Barling, & Iverson, 2005). However, social desirability remains an interference factor. Exact figures concerning accidents and violations recorded by the traffic management department could provide more precise data. Thus, future studies would be enhanced by integrating self-report measures with other methods such as field observations and simulated driving. Finally, unfortunately, we did not measure the fifth factor, Displaced Aggression, in the Chinese version of the DAX because of its low reliability in the original version (Deffenbacher et al., 2002). In fact, Displaced Aggression is an important factor that differs from direct aggression and has its own theoretical support (Denson, Pedersen, & Miller, 2006; Dollard, Miller, Doob, Mowrer, & Sears, 1939; Martinez, Zeichner, Reidy, & Miller, 2008). It is worth measuring Displaced Aggression in driving in future studies. In summary, we verified a short version of the DAX in a Chinese sample. This tool can easily be administered with other instruments. It is useful for classifying drivers according to different types of anger expression. Because anger reactions are learned in the course of life, these harmful behaviors can be replaced by good behaviors (Deffenbacher & Stark, 1992). Indeed, some researchers have explored the effect of cognitive-behavioral treatment on high-anger drivers (Balogun, Shenge, & Oladipo, 2012; Del Vecchio & O’Leary, 2004). Thus, the Chinese version of the DAX may be a useful tool for developing personalized anger management education and training classes based on various expressions of driving anger. For example, drivers who are more likely to express their anger with their vehicles must be trained to express their anger in alternative ways or to practice emotional regulation.

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

85

Acknowledgements The authors thank Jing Chai for data collection. This study was partially supported by grants from the National Natural Science Foundation of China (Grants No. 31100750, No. 31400886 and No. 91124003) and the Basic Project of National Science and Technology of China (No. 2009FY110100). References Aarts, L., & van Schagen, I. (2006). Driving speed and the risk of road crashes: A review. Accident Analysis and Prevention, 38(2), 215–224. http://dx.doi.org/ 10.1016/j.aap.2005.07.004. Abdu, R., Shinar, D., & Meiran, N. (2012). Situational (state) anger and driving. Transportation Research Part F – Traffic Psychology and Behaviour, 15(5), 575–580. http://dx.doi.org/10.1016/j.trf.2012.05.007. Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology, 53, 27–51. http://dx.doi.org/10.1146/ annurev.psych.53.100901.135231. Arthur, W., Bell, S. T., Edwards, B. D., Day, E. A., Tubre, T. C., & Tubre, A. H. (2005). Convergence of self-report and archival crash involvement data: A two-year longitudinal follow-up. Human Factors: The Journal of the Human Factors and Ergonomics Society, 47(2), 303–313. http://dx.doi.org/10.1518/ 0018720054679416. 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. Transportation Research Part F – Traffic Psychology and Behaviour, 26, 317–325. http://dx.doi.org/10.1016/j.trf.2014.01.004. Bachoo, S., Bhagwanjee, A., & Govender, K. (2013). The influence of anger, impulsivity, sensation seeking and driver attitudes on risky driving behaviour among post-graduate university students in Durban, South Africa. Accident Analysis and Prevention, 55, 67–76. http://dx.doi.org/10.1016/ j.aap.2013.02.021. Balogun, S. K., Shenge, N. A., & Oladipo, S. E. (2012). Psychosocial factors influencing aggressive driving among commercial and private automobile drivers in Lagos metropolis. Social Science Journal, 49(1), 83–89. http://dx.doi.org/10.1016/j.soscij.2011.07.004. Barling, J., Kelloway, E. K., & Iverson, R. D. (2003). High-quality work, job satisfaction, and occupational injuries. Journal of Applied Psychology, 88(2), 276–283. http://dx.doi.org/10.1037/0021-9010.88.2.276. Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88(3), 588. Berkowitz, L. (1989). Frustration-aggression hypothesis: Examination and reformulation. Psychological Bulletin, 106(1), 59–73. Berkowitz, L. (1990). On the formation and regulation of anger and aggression: A cognitive-neoassociationistic analysis. American Psychologist, 45(4), 494–503. Brislin, R. W. (1980). Translation and content analysis of oral and written materials. In H. C. Triandis & J. W. Berry (Eds.), Handbook of cross-cultural psychology: Methodology (pp. 389–444). Boston: Allyn and Bacon. Brown, T. A. (2014). Confirmatory factor analysis for applied research. Guilford Publications. Browne, M. W., Cudeck, R., Bollen, K. A., & Long, J. S. (1993). Alternative ways of assessing model fit. In K. A. Bollen (Ed.), Testing structural equation models (pp. 136–162). Newbury Park, CA: Sage. Cooper, P. J. (1997). The relationship between speeding behaviour (as measured by violation convictions) and crash involvement. Journal of Safety Research, 28(2), 83–95. http://dx.doi.org/10.1016/S0022-4375(96)00040-0. CRTAS. (2011). China Road Traffic Accidents Statistics. Traffic Administration Bureau of China State security Ministry, Beijing, China. 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. Accident Analysis and Prevention, 45, 1–9. http://dx.doi.org/10.1016/j.aap.2011.11.012. Dahlen, E. R., & Ragan, K. M. (2004). Validation of the propensity for angry driving scale. Journal of Safety Research, 35(5), 557–563. http://dx.doi.org/10.1016/ j.jsr.2004.09.002. Deffenbacher, J. L., Filetti, L. B., Richards, T. L., Lynch, R. S., & Oetting, E. R. (2003). Characteristics of two groups of angry drivers. Journal of Counseling Psychology, 50(2), 123–132. http://dx.doi.org/10.1037/0022-0167.50.2.123. Deffenbacher, J. L., Huff, M. E., Lynch, R. S., Oetting, E. R., & Salvatore, N. F. (2000). Characteristics and treatment of high-anger drivers. Journal of Counseling Psychology, 47(1), 5–17. http://dx.doi.org/10.1037/0022-0167.47.1.5. Deffenbacher, J. L., Lynch, R. S., Deffenbacher, D. M., & Getting, E. R. (2001). Further evidence of reliability and validity for the driving anger expression inventory. Psychological Reports, 89(3), 535–540. http://dx.doi.org/10.2466/pr0.2001.89.3.535. Deffenbacher, J. L., Lynch, R. S., Oetting, E. R., & Swaim, R. C. (2002). The driving anger expression inventory: A measure of how people express their anger on the road. Behaviour Research and Therapy, 40(6), 717–737. http://dx.doi.org/10.1016/s0005-7967(01)00063-8. Deffenbacher, J. L., Oetting, E. R., & Lynch, R. S. (1994). Development of a driving anger scale. Psychological Reports, 74(1), 83–91. http://dx.doi.org/10.2466/ pr0.1994.74.1.83. Deffenbacher, J. L., Oetting, E. R., Lynch, R. S., & Morris, C. D. (1996). The expression of anger and its consequences. Behaviour Research and Therapy, 34(7), 575–590. http://dx.doi.org/10.1016/0005-7967(96)00018-6. Deffenbacher, J. L., & Stark, R. S. (1992). Relaxation and cognitive-relaxation treatments of general anger. Journal of Counseling Psychology, 39(2), 158–167. http://dx.doi.org/10.1037/0022-0167.39.2.158. Deffenbacher, J. L., White, G. S., & Lynch, R. S. (2004). Evaluation of two new scales assessing driving anger: The driving anger expression inventory and the driver’s angry thoughts questionnaire. Journal of Psychopathology and Behavioral Assessment, 26(2), 87–99. http://dx.doi.org/10.1023/ b:joba.0000013656.68429.69. Del Vecchio, T., & O’Leary, K. D. (2004). Effectiveness of anger treatments for specific anger problems: A meta-analytic review. Clinical Psychology Review, 24(1), 15–34. http://dx.doi.org/10.1016/j.cpr.2003.09.006. Denson, T. F., Pedersen, W. C., & Miller, N. (2006). The displaced aggression questionnaire. Journal of Personality and Social Psychology, 90(6), 1032–1051. http://dx.doi.org/10.1037/0022-3514.90.6.1032. Dollard, J., Miller, N. E., Doob, L. W., Mowrer, O. H., & Sears, R. R. (1939). Frustration and aggression. New Haven, CT: Yale University Press. Dula, C. S., & Ballard, M. E. (2003). Development and evaluation of a measure of dangerous, aggressive, negative emotional, and risky driving. Journal of Applied Social Psychology, 33(2), 263–282. http://dx.doi.org/10.1111/j.1559-1816.2003.tb01896.x. Elvik, R. (2012). Speed limits, enforcement, and health consequences. Annual Review of Public Health, 33, 225–238. http://dx.doi.org/10.1146/annurevpublhealth-031811-124634. Esßiyok, B., Yasak, Y., & Korkusuz, I. (2007). Anger expression on the road: Validity and reliability of the driving anger expression inventory. Turkish Journal of Psychiatry, 18(3), 231–243. Field, A. (2009). Discovering statistics using SPSS (3rd ed.). SAGE Publications. Finney, S. J., & DiStefano, C. (2006). Non-normal and categorical data in structural equation modeling. Structural Equation Modeling: A Second Course, 269–314. Gonzalez-Iglesias, B., Antonio Gomez-Fraguela, J., & Angeles Luengo-Martin, M. (2012). Driving anger and traffic violations: Gender differences. Transportation Research Part F – Traffic Psychology and Behaviour, 15(4), 404–412. http://dx.doi.org/10.1016/j.trf.2012.03.002. Herrero-Fernández, D. (2011). Psychometric adaptation of the driving anger expression inventory in a Spanish sample: Differences by age and gender. Transportation Research Part F – Traffic Psychology and Behaviour, 14, 324–329. http://dx.doi.org/10.1016/j.trf.2011.03.001.

86

Y. Ge et al. / Transportation Research Part F 33 (2015) 75–86

Iliescu, D., & Sârbescu, P. (2013). The relationship of dangerous driving with traffic offenses: A study on an adapted measure of dangerous driving. Accident Analysis and Prevention, 51, 33–41. http://dx.doi.org/10.1016/j.aap.2012.10.014. Inoue, A., Kawakami, N., Shimomitsu, T., Tsutsumi, A., Haratani, T., Yoshikawa, T., et al (2014). Development of a short version of the new brief job stress questionnaire. Industrial Health, 52(6), 535–540. http://dx.doi.org/10.2486/indhealth.2014-0114. International Test Commission (ITC), (2010). International test commission guidelines for translating and adapting tests. . Iversen, H., & Rundmo, T. (2002). Personality, risky driving and accident involvement among Norwegian drivers. Personality and Individual Differences, 33(8), 1251–1263. http://dx.doi.org/10.1016/s0191-8869(02)00010-7. Jovanovic, D., Lipovac, K., Stanojevic, P., & Stanojevic, D. (2011). The effects of personality traits on driving-related anger and aggressive behaviour in traffic among Serbian drivers. Transportation Research Part F – Traffic Psychology and Behaviour, 14(1), 43–53. http://dx.doi.org/10.1016/j.trf.2010.09.005. Kaplan, D. (1989). A study of the sampling variability and z-values of parameter estimates from misspecified structural equation models. Multivariate Behavioral Research, 24(1), 41–57. Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New York: Guilford Press. Lajunen, T., & Parker, D. (2001). Are aggressive people aggressive drivers? A study of the relationship between self-reported general aggressiveness, driver anger and aggressive driving. Accident Analysis and Prevention, 33(2), 243–255. http://dx.doi.org/10.1016/s0001-4575(00)00039-7. Lajunen, T., Parker, D., & Summala, H. (2004). The Manchester driver behaviour questionnaire: A cross-cultural study. Accident Analysis and Prevention, 36(2), 231–238. http://dx.doi.org/10.1016/s0001-4575(02)00152-5. Lennon, A., & Watson, B. (2011). ‘‘Teaching them a lesson?’’ A qualitative exploration of underlying motivations for driver aggression. Accident Analysis and Prevention, 43(6), 2200–2208. http://dx.doi.org/10.1016/j.aap.2011.06.015. Li, F., Yao, Y., Jiang, L., & Li, Y. J. (2014). Driving anger in China: Psychometric properties of the driving anger scale (DAS) and its relationship with aggressive driving. Personality and Individual Differences, 68, 130–135. http://dx.doi.org/10.1016/j.paid.2014.04.018. Liu, H. J., & Gao, H. M. (2012). Reliability and validity of the Chinese version of State-trait anger expression inventory – 2 in college students. Chinese Mental Health Journal, 26(1), 70–76. http://dx.doi.org/10.3969/j.issn.1000-6729.2012.01.016. Martinez, M. A., Zeichner, A., Reidy, D. E., & Miller, J. D. (2008). Narcissism and displaced aggression: Effects of positive, negative, and delayed feedback. Personality and Individual Differences, 44(1), 140–149. http://dx.doi.org/10.1016/j.paid.2007.07.012. Moore, M., & Dahlen, E. R. (2008). Forgiveness and consideration of future consequences in aggressive driving. Accident Analysis and Prevention, 40(5), 1661–1666. http://dx.doi.org/10.1016/j.aap.2008.05.007. Özkan, T., Lajunen, T., Chliaoutakis, J. E., Parker, D., & Summala, H. (2006). Cross-cultural differences in driving behaviours: A comparison of six countries. Transportation Research Part F – Traffic Psychology and Behaviour, 9(3), 227–242. http://dx.doi.org/10.1016/j.trf.2006.01.002. Qu, W., Ge, Y., Jiang, C., Du, F., & Zhang, K. (2014). The Dula dangerous driving index in China: An investigation of reliability and validity. Accident Analysis and Prevention, 64, 62–68. http://dx.doi.org/10.1016/j.aap.2013.11.004. Regmi, K., Naidoo, J., & Pilkington, P. (2010). Understanding the processes of translation and transliteration in qualitative research. International Journal of Qualitative Methods, 9(1), 16–26. Richer, I., & Bergeron, J. (2012). Differentiating risky and aggressive driving: Further support of the internal validity of the Dula dangerous driving index. Accident Analysis and Prevention, 45, 620–627. http://dx.doi.org/10.1016/j.aap.2011.09.014. Rosen, E., & Sander, U. (2009). Pedestrian fatality risk as a function of car impact speed. Accident Analysis and Prevention, 41(3), 536–542. Sarbescu, P. (2012). Aggressive driving in Romania: Psychometric properties of the driving anger expression inventory. Transportation Research Part F – Traffic Psychology and Behaviour, 15(5), 556–564. http://dx.doi.org/10.1016/j.trf.2012.05.009. Soole, D. W., Lennon, A. J., Watson, B. C., & Bingham, C. R. (2011). Towards a comprehensive model of driver aggression: A review of the literature and directions for the future. In C. N. Ferraro (Ed.), Traffic safety. Nova Science Publishers. Spielberger, C. D. (1999). STAXI-2: State-trait anger expression inventory-2: Professional manual. Psychological Assessment Resources. Stanton, J. M., Sinar, E. F., Balzer, W. K., & Smith, P. C. (2002). Issues and strategies for reducing the length of self-report scales. Personnel Psychology, 55(1), 167–194. Stephens, A. N., & Sullman, M. J. (2014). Development of a short form of the driving anger expression inventory. Accident Analysis and Prevention, 72, 169–176. http://dx.doi.org/10.1016/j.aap.2014.06.021. Sullman, M. J. M. (2006). Anger amongst New Zealand drivers. Transportation Research Part F – Traffic Psychology and Behaviour, 9(3), 173–184. http:// dx.doi.org/10.1016/j.trf.2005.10.003. Sullman, M. J. (2015). The expression of anger on the road. Safety Science, 72, 153–159. Sullman, M. J. M., Stephens, A. N., & Kuzu, D. (2013). The expression of anger amongst Turkish taxi drivers. Accident Analysis and Prevention, 56, 42–50. http:// dx.doi.org/10.1016/j.aap.2013.03.013. Turner, C., McClure, R., & Pirozzo, S. (2004). Injury and risk-taking behavior—A systematic review. Accident Analysis and Prevention, 36(1), 93–101. http:// dx.doi.org/10.1016/S0001-4575(02)00131-8. Underwood, G., Chapman, P., Wright, S., & Crundall, D. (1999). Anger while driving. Transportation Research Part F – Traffic Psychology and Behaviour, 2(1), 55–68. http://dx.doi.org/10.1016/S1369-8478(99)00006-6. Van Rooy, D. L., Rotton, J., & Burns, T. M. (2006). Convergent, discriminant, and predictive validity of aggressive driving inventories: They drive as they live. Aggressive Behavior, 32(2), 89–98. http://dx.doi.org/10.1002/ab.20113. Villieux, A., & Delhomme, P. (2010). Driving anger and its expressions: Further evidence of validity and reliability for the driving anger expression inventory French adaptation. Journal of Safety Research, 41(5), 417–422. http://dx.doi.org/10.1016/j.jsr.2010.08.003. West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables: Problems and remedies. In R. H. Hoyle (Ed.), Structural equation modeling: Concepts, issues and applications (pp. 56–75). Thousand Oaks, CA: Sage. Wheaton, D. (1977). Assessing reliability and stability in panel models. Sociological Methodology. Willemsen, J., Dula, C. S., Declercq, F., & Verhaeghe, P. (2008). The Dula dangerous driving index: An investigation of reliability and validity across cultures. Accident Analysis and Prevention, 40(2), 798–806. http://dx.doi.org/10.1016/j.aap.2007.09.019. Yasak, Y., & Esiyok, B. (2009). Anger amongst Turkish drivers: Driving anger scale and its adapted, long and short version. Safety Science, 47(1), 138–144. http://dx.doi.org/10.1016/j.ssci.2008.02.003. Zacharatos, A., Barling, J., & Iverson, R. D. (2005). High-performance work systems and occupational safety. Journal of Applied Psychology, 90(1), 77–93. http://dx.doi.org/10.1037/0021-9010.90.1.77. 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? Accident Analysis and Prevention, 38(1), 22–27. http://dx.doi.org/10.1016/j.aap.2005.06.015. Zhang, W., Tsimhoni, O., Sivak, M., & Flannagan, M. J. (2010). Road safety in China: Analysis of current challenges. Journal of Safety Research, 41(1), 25–30. http://dx.doi.org/10.1016/j.jsr.2009.12.003.