Risky riding behavior on two wheels: The role of cognitive, social, and personality variables among young adolescents

Journal of Safety Research 46 (2013) 47–57

Contents lists available at SciVerse ScienceDirect

Journal of Safety Research journal homepage: www.elsevier.com/locate/jsr

Risky riding behavior on two wheels: The role of cognitive, social, and personality variables among young adolescents Alessandra Falco ⁎, Alessandra Piccirelli, Damiano Girardi, Laura Dal Corso, Nicola A. De Carlo Department FISPPA - Section of Applied Psychology, University of Padova, Via Venezia 8, 35131 Padova, Italy

a r t i c l e

i n f o

Article history: Received 17 January 2012 Received in revised form 21 January 2013 Accepted 7 March 2013 Available online 6 April 2013 Keywords: Risky riding behavior Adolescents people on two wheels Risk perception Gender Personality

a b s t r a c t Objective: The main objective of this study was to analyze and estimate the relations between risky riding behaviors and some personality and sociocognitive variables through structural equation modeling. We focused on two-wheel riding behavior among a sample of 1,028 Italian adolescents at their first driving experience. Conclusions: The main findings confirmed the role of personality in influencing riding behavior directly as well as indirectly through risk perception. In particular, risk perception was a significant mediator between personality, social norm, and riding behavior. The significant relations that emerged in the general sample were further confirmed in the two specific sub-samples of males and females. In terms of social marketing and educational communication, it may consequently be advisable to proceed in an integrated and coordinated manner at both the cognitive and social level, taking into account some "dispositions to risk" related to personality. Impact on industry: The integrated and coordinated action on different levels - cognitive, social, and personality - may therefore allow more effective and significant results in reducing those risky riding behaviors that often underlie young two-wheel riders' higher involvement in traffic accidents. © 2013 National Safety Council and Elsevier Ltd. All rights reserved.

1. Introduction 1.1. Adolescents and road accidents on two wheels Road traffic injuries represent a severe threat to health and are therefore a priority in most countries. An alarming issue is the ever-increasing involvement of young road users. Every year, 32,000 people younger than 25 years lose their life to road traffic injuries, either as drivers/riders or passengers (World Health Organization, 2007). The same trend is observed in Italy, where 360 youths aged 15–19 years lost their life in two-wheeled accidents in 2008 and 31,584 were injured. Important gender differences are evident: the risk of injury is twofold and the risk of mortality fourfold greater for men compared to women in the age bracket 15–19 years (National Institute of Statistics, 2009). Because of this increased accident rate, it is important to identify and understand possible risk factors for young two-wheeled riders. Many studies have tried to define and identify high-risk riders by combining several variables and assessing their interactions. In general, those who persist with dangerous behaviors while driving/

⁎ Corresponding author. Tel./fax: +39 0498276590. E-mail addresses: [email protected] (A. Falco), [email protected] (A. Piccirelli), [email protected] (D. Girardi), [email protected] (L. Dal Corso), [email protected] (N.A. De Carlo).

riding are considered high-risk drivers/riders and represent a higher accident risk. As observed by Vézina (2001) this population is very heterogeneous and comprises different subgroups with different sociodemographic profiles, attitudes, and behaviors. Many studies have investigated social, cognitive, and personality variables for prediction of high-risk behavior, measured in terms of the number of accidents and/or traffic violations committed in a given period, usually 2 or 3 years, or the frequency of behaviors that are dangerous and do not comply with the highway code while driving/riding. The first few studies revealed an association between high-risk drivers and individual and cognitive characteristics such as social and personal maladjustment, impulsiveness, and deficit in information processing (Mayer & Treat, 1977), hostility and alienation from the educational system (Pelz & Shuman, 1973), and use of alcohol and/or drugs (Farrow, 1985). Peck (1993) claimed that there is no single variable or combination of variables that can accurately predict risky driving behavior. Some of a driver’s characteristics may increase his/her chance of being involved in an accident. These are related to social maladjustment and to personality and attitude traits, together with being young, male, belonging to a low socioeconomic class, and having little driving experience and no previous road accidents or fines. At-risk drivers are therefore more deviant in terms of social maladjustment, alcohol and drug use, and individual characteristics (gender, age, personality traits, etc.) that predispose them to risktaking behavior. Among these, sensation-seeking has attracted a great deal of attention, in particular because of its effects on favoring

0022-4375/$ – see front matter © 2013 National Safety Council and Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jsr.2013.03.002

48

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

risky driving behavior among the young, especially if associated with alcohol use (Arnett, 1990; Yu & Williford, 1993). The social environment also plays an important role, in particular family members and parents who may influence the young person in terms of future risky driving behaviors. In their role as behavior models, parents can indeed pass on their driving style (Bianchi & Summala, 2004) and influence their children’s driving behavior through educational styles and expression of attitudes. Shope, Waller, Raghunathan, and Patil (2001) noted that the use of substances such as cigarettes, marijuana, and alcohol, together with negative parental influences, observed at 15 years of age may have a bearing on the increase in road accidents for young people, especially females, aged 23–24 years. This confirms the predictive value of substance use for risky driving behaviors by adolescents. Parental influence is very important as well. Specifically, a 15-year-old who perceives low levels of monitoring, nurturance, and family connectedness and high levels of parental indulgence towards alcohol use by young people will have a higher risk of violations and road accidents when driving, independent of gender. Many studies have aimed to identify distinctive variables able to predict risky driving behaviors among young drivers. Young twowheeled riders are an emergent at-risk category that requires further efforts to understand possible risk factors. Our study focuses on young moped riders, a category largely composed of adolescents with little or no driving/riding experience who are influenced by specific sets of values, beliefs, and social norms, as well as their sociocultural context, and are characterized by a greater propensity for risky behavior. Issues related to the physical development of adolescents must also be considered. Many adolescents are in a stage of growth and physical maturation in which hormones are “raging” and energy levels are high. Even their brain functions are not fully developed, especially with regard to the prefrontal cortex, where impulse inhibition, decision-making, and feedback are processed (Paus, 2005). Adolescents have a peculiar propensity to adopt risky behaviors that play several roles related to identity development, social participation (Arnett, 1992), and overall personality. Through risky behaviors, an adolescent can prove him/herself, assert his/her “adultness,” be accepted by peer groups, and satisfy a need for transgression or challenge while claiming his/her autonomy and independence (Silbereisen, Eyferth, & Rudinger, 1986; Silbereisen & Todt, 1994). More than driving/riding inexperience, the psychosocial aspects and personality traits connected to adolescence have the most significant bearing on traffic accidents and on the adoption of risky behaviors in general (Boyce & Geller, 2002; Deery, 1999; Dworkin, 2005; Fergusson, Swain-Campbell, & Horwood, 2003; Harré, 2000; Iversen & Rundmo, 2004; Rodham, Brewer, Mistral, & Stallard, 2006; Vavrik, 1997). These specific characteristics can help to differentiate the behavior of adolescent motorcyclists from that of motorcyclists and car drivers over the age of 18 years. In fact, older individuals have a more stable personality and are less prone to emotions and excitement-seeking, and have more control and awareness of their ability to drive and of potential traffic hazards, and are thus more able to understand and direct their own risk-taking behavior (Irwin & Millstein, 1986; Jessor, 1987; McKnight, 1999). According to cognitive psychology, some behaviors, such as driving, become habits and automatisms over time (Ranney, 1994). It is also known that individuals with more entrenched habits take into account less information than those with less entrenched habits, who assess situational contingencies more carefully and choose the driving behavior to adopt more actively (Fuji & Kitamura, 2003). Considering the often random nature of road accidents and the lack of awareness of what characterizes risky behavior, it is possible that a person unintentionally develops a stable risky driving habit. This willingness to take risks may arise especially during adolescence

due to the increased intensity of traits such as sensation-seeking, normlessness, anger, and unrealistic optimism, each of which interacts with cognitive and biopsychosocial development processes of and social influences. Bearing in mind that healthy or unhealthy habits that can stabilize and recur over time are developed during adolescence, it is important to understand which variables have the most influence on driving/ riding behaviors and subsequent involvement in road accidents to implement more effective preventive and educational actions for this age group. The study of personality, cognitive and social characteristics of young riders is therefore an important and productive focus for research. Numerous approaches have been taken to understand the variables and processes underlying the driving/riding behavior of young people. A few have centered on the predictive value of some personality traits, others have highlighted the role of cognitive variables in elaborating external information, while others have tried to explain behavior within social cognition models. Despite the extensive literature on traffic psychology, only a few studies have integrated the different approaches. Furthermore, these studies (Chen, 2009; Machin & Sankey, 2008; Ulleberg & Rundmo, 2003; Wong, Chung, & Huang, 2010) focused exclusively on young adults, but it is unclear whether the findings can be generalized to adolescents. The aim of our study is to contribute to the understanding of the mechanisms and processes underlying risk-taking behaviors on the road by young riders of two-wheeled vehicles. Two-wheeled vehicles are intended to carry up to two people, including the rider. In particular, our study focuses on mopeds with an engine capacity of 50 cc or less and a maximum speed of 45 km/h on a level road (National Institute of Statistics, 2009). We investigate the influence of the main sociocognitive and personality variables used for samples of young drivers on the moped riding behavior of young adolescents aged 14–15 years. This is the age at which young people in Italy can drive a moped after obtaining the driving license. To the best of our knowledge, no studies have investigated the integrated influence of personality variables and cognitive and social skills on 14–15-year-old adolescents when riding a moped. 1.2. The role of social, cognitive, and personality variables Most traffic accidents are due to human factors, namely to risktaking and irresponsible behavior that violates the highway code. A literature review showed that many possible variables can influence risky driving behaviors. The influence of these variables has been confirmed in several studies for samples of young drivers of four-wheeled vehicles and, in recent years, samples of motorcyclists. For the reasons set out in the Introduction and in light of these studies, we chose to focus on some personality and psychosocial variables analyzed in an integrated model. 1.2.1. Risk perception Among the factors influencing risky driving/riding behaviors, risk perception should be considered. It can be defined as subjective assessment of the probability of a specific event happening and of the magnitude of the possible negative consequences (Sjöberg, Moen, & Rundmo, 2004). This construct was investigated in relation to young drivers of four-wheeled vehicles who are more likely to underestimate dangerous situations (Brown & Groeger, 1988; Deery, 1999) and to consider themselves less at risk compared to other road users (Glick, Kronenfeld, Jackson, & Zhang, 1999), and consequently to engage in risky traffic behaviors (Arnett, 1992; Dworkin, 2005; Harré, 2000; Moller, 2004). In these studies, risk perception can be considered an antecedent of behavior: the higher the risk perception associated with a behavior, the lower is the chance that a person will engage in this behavior. In many studies, risk perception appeared to be an important mediator between personality and riding behavior variables (Machin &

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

Sankey, 2008; Rundmo & Iversen, 2004; Wong et al., 2010). Vanlaar and Yannis (2006) observed that the risk perception and prevalence of dangerous driving behaviors in terms of the estimated percentage of drivers engaging in them may influence the level of driver concern regarding road security issues and thus driving behavior. Further examination showed that the risk perception of some behaviors represents the result not only of self-assessment but also of the level of concern of others, namely knowing or thinking that others generally deem some behaviors to be dangerous (e.g., driving when drunk or after using drugs) increases an individual’s risk perception and affects subsequent behavior (Vanlaar, Simpson, & Robertson, 2008). However, the direction of the effect between risk perception and behavior is not completely clear. In a study by Horvath and Zuckerman (1992), a model considering risk perception as a consequence of risky behavior had better fit indices. In this case, reckless driving behavior without negative consequences may strengthen the sense of driving competence and lower the estimate of the connected risk. Regarding the two components of risk perception, some studies considered the influence of the cognitive component (how information is perceived and processed) on driving behavior (Brown & Cotton, 2003; Deery & Fildes, 1999; Horvath & Zuckerman, 1992; Machin & Sankey, 2008). Others pointed out the influence of the affective component, in terms of worry and emotional reactions associated with possible negative consequences of involvement in traffic accidents (Rundmo & Iversen, 2004; Ulleberg & Rundmo, 2003; Wong et al., 2010). In the present study, risk perception was detected in both cognitive and affective components and was considered as an antecedent of risky riding behavior. We assumed that risk perception influences the moped riding behavior of adolescents both directly and by mediating the effects of other variables such as personality, optimism, and social influence. 1.2.2. Personality variables The personality variables most investigated are those that lead a person to ignore or underestimate road dangers and to engage in irresponsible and unsafe actions. These are therefore a set of variables that define a personality at risk in terms of greater likelihood of adopting unsafe driving behaviors and of being involved in a car accident. The influence of traits such as sensation-seeking, anger, and normlessness (Dahlen & White, 2006; Iversen & Rundmo, 2002; Oltedal & Rundmo, 2006; Ulleberg, 2001) on risky driving behavior by young people is known. Sensation-seeking, defined as the seeking of varied, complex, and intense sensations and experiences and the desire to take physical, social, legal, and financial risks for the sake of such sensations and experiences (Zuckerman, 1994), strongly influences risky behaviors in young people, including risky driving (Arnett, Offer, & Fine, 1997; Iversen & Rundmo, 2002; Jonah, 1997; Jonah, Thiessen, & Au-Yeng, 2001; Ulleberg, 2001). In the past few years, this relationship has been analyzed and confirmed for two-wheeled riding behavior among youths older than 18 years (Chen, 2009; Wong et al., 2010). Driving anger is a particular form of context-specific anger defined as a person’s proneness to get angry while driving, usually assessed on the Driving Anger Scale (DAS; Deffenbacher, Oetting, & Lynch, 1994). Several studies have confirmed the relationship between high driving anger scores and road accidents, aggressive driving, infringement of road traffic laws, and high-speed driving (above the speed limit) (Dahlen & White, 2006; Deffenbacher, Deffenbacher, Lynch, & Richards, 2003; Lajunen & Parker, 2001; Oltedal & Rundmo, 2006). Normlessness, a construct within social deviation, is the belief that some behaviors, although socially unapproved and inappropriate, are required and necessary to reach certain goals and results (Kohn & Schooler, 1983). Iversen and Rundmo (2002) found that participants with high normlessness scores were more frequently involved in

49

accidents and quasi-accidents and inclined to engage in dangerous driving behaviors. Oltedal and Rundmo (2006) confirmed the predictive capacity of normlessness associated with masculine gender; according to the authors, normlessness implies not only violation of road regulations but also a tendency towards general irresponsibility in driving behavior. Similarly, a study by Chen (2009) on the driving behavior of young university student motorcyclists showed that normlessness exerts a greater influence on male drivers. We also considered locus of control as a personality trait that can influence the intention to commit traffic violations. Since the focus of our study is on personality characteristics that encourage individuals to engage in more risky behaviors, only the external component of locus of control was considered. The tendency to attribute causes of events to external, uncontrollable factors reduces the threshold of perceived risk and favours reckless and less careful driving behaviors (Katwal & Kamalanabhan, 2001; Taris, 1997; Yagil, 2001). Analysis of the literature revealed that several personality traits can influence risky driving behaviors directly (Dahlen, Martin, Ragan, & Kuhlman, 2005; Machin & Sankey, 2008; Schwebel, Severson, Ball, & Rizzo, 2006; Vassallo et al., 2007), as well as indirectly through risk perception. Rundmo and Iversen (2004) found that personality influences the driving behavior of young drivers, acting through the emotional component of risk perception. In our investigation we assumed that personality has effects, both direct and indirect through the cognitive and affective components of risk perception, on the risk-taking riding behavior of young adolescents. 1.2.3. Social influence In analyzing young adolescents, we considered the effects on risk-taking riding behavior of social influence in terms of social norms (Ajzen, 1985, 1991). In some research studies, the social norm (the social pressure to pursue socially acceptable behaviors) was a weak predictor of behavior (Armitage & Conner, 2001; Benevene & Scopelliti, 2012; Hausenblas, Carron, & Mack, 1997). By contrast, in other studies it was the most important variable in determining the intention to engage in risky riding behaviors (Parker & Manstead, 1996; Wallén Warner & Åberg, 2008). In studies by Gardner and Steinberg (2005) and Michael and Ben-Zur (2007), teenagers were more influenced by their peers than by adults in decisions to engage in risky behaviors. Few studies have examined the relationship between social norms and risk perception. It is interesting to assess whether and to what extent the expectations of others, specifically significant adults and peers, influence the road risk assessment and consequent decisions of young adolescents on whether or not to engage in certain riding behaviors. We hypothesise that stronger perceived social pressure (e.g., parents/ friends expect responsible driving/riding behavior) corresponds to a higher perception of road risk and a lower frequency of irresponsible and illegal riding behavior. 1.2.4. Driving optimism The specific focus on adolescents justifies our consideration of optimism as a variable, defined as a generalized expectancy for positive events (Weinstein, 1980). An optimistic tendency to expect positive outcomes increases the probability of adopting reckless and unsafe behaviors, especially among young males (Gosselin, Gagnon, Stinchcombe, & Joanisse, 2010; Ulleberg, 2001). This tendency is related to a psychological process known as comparative optimism (Harré & Sibley, 2007; Harris & Middleton, 1994). Such an evaluation bias leads a person to believe he/she is more competent and skilful than others in controlling the outcome of events (Causse, Kouabenan, & Delhomme, 2004). In this sense, optimism is different from the risk perception construct: while the former requires an individual to compare him/herself and his/her abilities to

50

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

average people, risk perception reflects an assessment of the perceived level of risk and the emotions associated with it. In a road context, Moen (2007) proposed three dimensions of driver optimism: driver confidence (confidence in one’s abilities), driver self-assessment (evaluation of oneself in comparison to others), and driver stress (perception of not being in control of situations while driving and a consequent feeling of anxiety). The first two, if positive, can push a person to take risks (Dejoy, 1989; Van der Pligt, 1996), while stress has the opposite effect (Iversen, 2004). Driver confidence is the perception of safety a driver has about him/ herself and his/her abilities while driving; risk perception, instead, involves an assessment and evaluation of risk and therefore of insecurity with regard to certain driving/riding situations and emotions associated with the risk of accidents. We can therefore reasonably assume that a high degree of driving/riding optimism is associated with a general feeling of invulnerability that manifests itself as inaccurate risk perception and risk-taking behavior in traffic, especially among young adolescents with an illusion of control, due in part to their young age. Furthermore, a stronger effect of optimistic bias on males compared to females is expected (McKenna, Stanier, & Lewis, 1991). In this sense, driving optimism is considered to be an antecedent of risk perception and risky riding behavior. 1.3. Study objectives In the present study we analyse and assess the relations between personality traits, optimism, social norms, risk perception, and risky riding behaviors in young adolescents. As previously noted, no studies have examined all of the effects of the variables we identified on driving/riding behavior. Moreover, existing studies did not examine the specific behavior of adolescents aged 14–15 years with no or little riding experience, when at-risk riding behaviors are not yet rooted. In particular, we propose the following hypotheses: (a) personality, social norms, optimism, and risk perception directly influence riding behavior; and (b) risk perception mediates the relationship between personality, social norms, optimism, and riding behavior. The literature has emphasised a connection between gender and the adoption of risky behaviors: young males show a greater tendency and inclination to both seek and engage in dangerous and risky behaviors, in general and when riding. Several studies have pointed out that adolescent boys exhibit more dangerous behaviors than their female counterparts, and have a personality more inclined towards risk-taking and emotion-seeking (Gullone & Moore, 2000; Oltedal & Rundmo, 2006; Ulleberg, 2001; Yagil, 1998). A few studies have shown the opposite, with girls engaging in more risky driving behaviors and exhibiting higher involvement in road accidents. According to Romano, Kelley-Baker, and Voas (2008), the higher involvement in accidents of women can be explained in part by their greater driving exposure, yet the finding also indicates that some groups of women drivers, especially the very young, are more inclined to engage in risky driving behaviors. Specifically, the higher accident rate for young female adolescents is mainly associated with alcohol use/ abuse (Tsai, Anderson, & Vaca, 2008, 2010), use of other substances, and negative environmental influences (family and friends; Elliott, Shope, Raghunathan, & Waller, 2006). Although assessment of the use and/or abuse of alcohol and other substances was not one of our study aims, we analyzed and assessed possible differences due to gender among young adolescents for two-wheeled risky riding behaviors and their predictive factors. 2. Method 2.1. Participants The survey was part of an educational project on road safety sponsored by an Italian region. Various educational institutions, including

high schools, technical institutes, industrial institutes, hotel schools, art schools, and commerce institutes in the region, were involved in the research. The variety of schools ensured participation by young people with different interests and from different social backgrounds. The questionnaire was administered to 1,126 young people in first or second year of high school in the presence of psychologists who explained the instrument and the research objectives and supervised its completion. The sample used for the research is limited to a specific Italian region and it cannot be considered representative of the entire population of young moped riders. Of the participants 56% were male and 44% were female. The mean age was 14.58 years. Some 44% of the participants obtained a motorcycle license after attending courses organized within their school, of whom 77% believed that the course contributed to making them good road users. The remaining participants (56%) obtained a license after attending a course at a driving school. All the participants had a motorcycle license. Adolescents without a license did not fill in the questionnaire. According to the Italian highway code, mopeds can be ridden from the age of 14 years after obtaining a license by attending a course at a driving school or high school. At the end of the course, a license can be obtained by passing the final examination. Therefore, the participants in this research had limited driving experience of a maximum of 1 year. 2.2. Instrument The scale items were formulated on the basis of studies and tools already available in the literature, translated into Italian if necessary and adapted to specific riding behavior on two wheels. Before administration, the questionnaire was subjected to a preliminary test by independent judges who were experts in the phenomenon under investigation, who evaluated the formulation and the clarity of questions, the adequacy of response categories and the overall understanding of the instrument. The questionnaire was then administered to a group of 40 adolescents, who showed a good understanding of the questions and the answer modes. The instrument, called questionnaire of risk perception on two wheels (Q-Per2), comprises the following scales. For personality, four scales were used for traits highlighted in the literature as important predictors of traffic behavior. For sensationseeking, we adapted the Brief Sensation-Seeking Scale (BSSS) of Hoyle, Stephenson, Palmgreen, Lorch, and Donohew (2002). This comprises the following subscales: Thrill- and Adventure-Seeking (e.g., I like to do frightening things), Experience-Seeking (e.g., I would like to explore strange places), Boredom Susceptibility (e.g., I prefer friends who are excitingly unpredictable), and Disinhibition (e.g., I like wild parties). Four specific items pertaining to traffic behavior were added that we categorized as a thrill-seeking while riding subscale (e.g., I like the thrill of speed on a motorcycle). External locus of control was assessed using a scale specific to riding behavior, standardized for the Italian context, made up of four items (e.g., road accidents are governed by chance/fate; Falco, 2007). Normlessness was measured via three items drawn and adapted from the model of Kohn and Schooler (1983; e.g., it is right to commit certain actions even if they are forbidden). One original item of the scale was not included because it was considered difficult to understand by participants in the preliminary test. For sensation-seeking, external locus of control, and normlessness, respondents had to express agreement or disagreement on a fourpoint Likert scale (from strongly disagree to strongly agree). Finally, driving anger was assessed using eight items drawn and adapted from the DAS (Deffenbacher et al., 1994). A new item specific to driving on two wheels was added. Participants had to imagine themselves in potentially frustrating and irritating riding situations and to indicate how much anger they would experience on a scale

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

ranging from “not at all” to “a lot” (e.g., a driver opens the car door without noticing your arrival). The social norm was evaluated in relation to significant people, such as parents, family, and teachers (subjective norm, parents scale), and to peers, who have a considerable impact in adolescence (subjective norm, friends scale). Both scales are composed of four items taken from Ajzen (1991). For both, participants had to express their degree of agreement or disagreement on a four-point Likert scale (e.g., the friends I go out with expect me to drive the scooter in the safest way possible [not going in the wrong direction, doing a U-turn, not changing lanes frequently]). For driving optimism, we used three scales developed by Moen (2007) and suitably adapted to the two-wheeled context. The driver self-assessment scale (seven items; e.g., I am a good driver compared to someone my age and with similar experience) evaluated the respondent’s ability as a moped rider compared to the average; the driver confidence scale (eight items; e.g., I never regret the decisions I make in traffic) measured driver safety; and the driver stress scale (five items; e.g. situations where I am not in control happen often) assessed the level of anxiety and tension experienced while driving. The respondents expressed their degree of agreement or disagreement on a four-point Likert scale. For risk perception, two scales were used. The cognitive component of risk was assessed in terms of aversion to risk-taking, namely the perception of danger when performing some actions (e.g., riding through a red light at an intersection, riding a moped without wearing a helmet) when riding. The item formulation was drawn from the scale proposed by Dorn and Machin (2004) and later used by Machin and Sankey (2008). Among these, one item was formulated specifically with regard to driving on two wheels. For the affective component of risk perception, the emotional reactions scale devised by Rundmo and Iversen (2004) was used, whereby participants indicate which emotion they feel on a bipolar seven-point scale (e.g., worry and concern versus tranquillity, safe versus unsafe) when thinking about the possibility of being involved in a traffic accident. This scale was composed of five couples of adjectives: four pairs taken from the scale of Rundmo et al. (2004) and one pair (bold versus conservative) formulated specifically with regard to driving on two wheels. Finally, a 22-item scale was used to evaluate the moped riding behavior of young people. Some general items regarding ordinary and aggressive violations (10 items) were drawn from the Driver Behavior Questionnaire (DBQ; Parker, Lajunen, & Stradling, 1998; Parker, Reason, Manstead, & Stradling, 1995) and others (four items) were specific to two-wheeled riding and were created ad hoc. To detect transgressive violations, eight items were included, four of which were taken from the DBQ and four were especially created for driving on two wheels. The respondents were asked to indicate how frequently they engaged in certain behaviors on a response scale ranging from “never” to “always.” The riding behaviors investigated included: ordinary violations, that is, violations of the road traffic law (e.g., failure to stop at a red light, driving down a no entry street); aggressive violations, or interpersonal hostile violations (e.g., showing, by means of insults or rude gestures, hostility towards other road users and sounding the horn/bell to make one’s irritation plain); and transgressive violations (e.g., vying with other scooter riders and performing dangerous acrobatics to show off one’s abilities). 2.3. Statistical analyses Before data analysis was performed, missing data were taken into account. Some 98 participants had extensive missing data (over 25% of the variables considered) and were therefore excluded from subsequent analyses (Lin, 2010). The sample thus obtained included 1028 respondents. Missing values were then estimated using multiple

51

imputation (Expected Maximization algorithm, EM). This technique guarantees more accurate assessment of the population parameters than listwise deletion or mean substitution (Schafer & Graham, 2002). To investigate the hypothesised relations, a structural equations model was estimated. The data covariance matrix was analysed using LISREL 8.8 software (Jöreskog & Sörbom, 2006). We assessed the model fit using the χ2 test, the root mean square error of approximation (RMSEA), the comparative fit index (CFI), the non-normed fit index (NNFI), and the standardized root mean square residual (SRMR). RMSEA values close to 0.06, CFI and NNFI values close to or greater than 0.95, and SRMR values close to 0.08 indicate a good model fit (Hu & Bentler, 1999). The first step in data analysis was to test the metric properties of the revised instrument used in this study. These were sensation-seeking, normlessness, and driving anger for personality; aversion to risktaking and emotional reactions for risk perception; an ordinary violations, aggressive violations, and transgressive violations for riding behaviors. For this purpose, we conducted three sets of confirmatory factor analysis (CFA), one for each construct. Since most of the items were not normally distributed, robust maximum likelihood (Satorra & Bentler, 1994) was used as estimation method. To evaluate both the convergent and discriminant validity of the three revised personality scales (sensation-seeking, normlessness, and driving anger), the external locus of control was also included in this first analysis. Thus, the hypothesised model comprised 28 scale items and four latent constructs. The fit indices showed an acceptable fit of the theoretical model to the data: S-Bχ2(344, N =1028) = 1576.881, p b 0.001; RMSEA = 0.059; NNFI = 0.931; CFI =0.937; and SRMR = 0.063. Analysis of the modification indices suggested that we should freely estimate the correlation between two items of the sensation-seeking scale (items 1 and 2 of the revised BSSS). This is justifiable from a theoretical point of view, given the content overlap of the two items. Then a new CFA was conducted. The fit indices showed a good fit of the theoretical model to the data: S-Bχ 2(343, N = 1028) = 1410.917, p b 0.001; RMSEA = 0.055; NNFI = 0.940; CFI = 0.946; and SRMR = 0.064. To test the metric properties of the two revised risk perception scales (aversion to risk-taking and emotional reactions) we conducted a second CFA. The hypothesised model comprised 12 scale items and two latent constructs. The fit indices showed a good fit of the theoretical model to the data: S-Bχ 2(53, N = 1028) = 248.612, p b 0.001; RMSEA = 0.060; NNFI = 0.953; CFI = 0.963; and SRMR = 0.045. Finally, a third CFA was conducted to test the metric properties of the revised riding behaviors scales (ordinary, aggressive, and transgressive violations). The hypothesized model comprised 22 scale items and three latent constructs. The fit indices showed a good fit of the theoretical model to the data: S-Bχ 2(206, N = 1028) = 960.257, p b 0.001; RMSEA = 0.060; NNFI = 0.976; CFI = 0.979; SRMR = 0.048. Overall, all the revised scales showed satisfactory convergent and discriminant validity (Brown, 2006). Before analyzing the structural model, CFA was conducted to evaluate the adequacy of the measuring model (Kline, 2010). The hypothesized model comprised five latent variables (personality, social norm, optimism, risk perception, and dangerous riding behavior) and 14 observed variables (average score for each scale of the questionnaire). These are: external locus of control, sensation-seeking, normlessness, and anger for personality; subjective norm parents and subjective norm friends for the social norm; driver self-assessment, driver self-confidence, and driver stress for optimism; emotional reactions and aversion to risk for risk perception; and ordinary, aggressive, and transgressive violations for dangerous riding behavior. The measuring model does not fit the data well: χ2(67, N = 1028) = 460.195, p b 0.001; RMSEA = 0.075; NNFI = 0.920; CFI = 0.941; SRMR = 0.060. In particular, the observed variables anger and driver stress have a low factorial saturation of the respective latent factors (λanger = 0.19; λdriver_stress = 0.02). For this reason, the two variables were removed and a new CFA was conducted. The measuring model then provided a good fit to the data: χ2(44, N =

52

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

1028) = 157.348, p b 0.001; RMSEA = 0.050; NNFI = 0.973; CFI = 0.982; SRMR = 0.035. Factorial saturations had values greater than 0.35. Correlations between latent factors ranged in intensity between – 0.01 (optimism and dangerous riding behavior) and 0.82 (the social norm and risk perception). This proves that the latent variables are distinct from each other (discriminant validity; Brown, 2006). The structural model was then estimated for the overall sample. The first estimate showed that optimism does not significantly influence either risk perception (γ = 0.05, ns) or risky riding behavior (γ = –0.02, ns). To make the model more parsimonious, the optimism latent variable was removed and the model was estimated again. Finally, the final structural model was estimated for male and female respondents separately. 3. Results Descriptive analyses and Cronbach’s α were calculated for the scales used in the instrument. It should be noted that one item was eliminated from the emotional reactions scale and three items for the driver confidence scale since these lowered the overall α for the corresponding scales. Table 1 lists the number of items, mean, standard deviation, and internal consistency for all measures. The reliability coefficients were satisfactory for all the scales, with α between 0.70 and 0.89 (Nunnally, 1978). Three SEM models were evaluated: one for the whole sample, one for male participants, and one for female participants. Table 2 presents the covariance matrix. The first estimated model, including all participants, is presented in Fig. 1, which shows the significant standardized coefficients are illustrated. The model fit indices showed a good fit to the data: χ 2(29, N = 1028) = 88.371, p b 0.001; RMSEA = 0.044; NNFI = 0.984; CFI = 0.990; and SRMR = 0.024. The results show that personality, social norms, and risk perception have an important influence on risky riding behavior by young people. Personality, identified in the model as an at-risk trait, had a significant and positive influence on risky riding behavior (γ = 0.36, p b 0.001). The more a young person tends to seek strong emotions, to justify illegal and immoral conducts, and to believe that accidents are due to destiny and fate, the more easily and naturally he/she will commit ordinary, aggressive, and transgressive riding violations. Risk perception had a very strong direct negative influence on the frequency of irresponsible behaviors that violate the highway code (β = –0.57, p = 0.01). The hypothesis of a direct effect of the social norm on riding behavior was not confirmed (γ = 0.14, ns). Both personality and the social norm influenced risk perception. Specifically, an at-risk personality reduced the risk perception

Table 1 Number of items, mean scores, standard deviations, and Cronbach’s α.

1.Ordinary violations 2.Aggressive violations 3.Transgressive violations 4.Emotional reactions† 5.Aversion to risk taking 6.Subjective norm parents 7.Subjective norm friends 8.External locus of control 9.Sensation seeking 10.Normlessness 11.Anger 12.Driver self-assessment 13.Driver confidence 14.Driver stress †

Number of items

Mean score

Standard deviations

Cronbach’s α

13 4 5 4 8 4 4 4 12 3 9 7 5 5

1.66 1.86 1.42 4.21 3.22 3.23 2.83 2.02 2.44 2.24 3.13 2.77 2.88 2.36

.55 .67 .62 1.09 .53 .62 .75 .73 .62 .77 .54 .50 .52 .62

.88 .75 .85 .71 .81 .70 .77 .75 .84 .70 .80 .71 .71 .73

Response scale from 1 to 7. The other response scales are from 1 to 4.

associated with involvement in a traffic accident (γ = –0.29, p b 0.001); whereas the expectations of parents and friends of safe and responsible riding increased the level of risk perception in the young people interviewed (γ = 0.71, p b 0.001). The proportion of total variance in risk perception explained by the social norm and personality was satisfactory (R 2 = 0.76). On the whole, the variables in the model explained 53% of the total variance for risky riding behavior. The hypothesis on the mediating effects of risk perception in the relation between personality/social norms and riding behavior was tested using the distribution-of-product method (Tofighi & MacKinnon, 2011). The indirect effect of personality traits on risky riding behavior through risk perception was significant [95% confidence interval (CI) 0.06–0.51], similar to the indirect effect of the social norm on risky riding behavior through risk perception (95% CI –0.85 to –0.11). The mediation was partial for personality traits and total for the influence exerted by parents and peer groups. The second objective of the study was analysis and verification of the mode for males and females separately (Table 3). The results reveal gender differences in personality, social norms, risk perception, and risky riding behavior. Significant differences were found for all constructs. To examine more thoroughly the effects of personality, social norms, and risk perception on riding behavior, the group of participants was split into male and female groups. The fit indices for the male subsample are satisfactory: χ 2(29, n = 568) = 55.112, p b 0.01; RMSEA = 0.039; NNFI = 0.986; CFI = 0.991; and SRMR = 0.024. Riding behavior was directly influenced by personality (γ = 0.45, p b 0.001) and risk perception (β = –0.30, p b 0.01). The level of risk perception was in turn determined by both personality traits (γ = –0.28, p b 0.001) and social norms (γ = 0.60, p b 0.001). The effects of mediation in the male subsample were confirmed by the distribution-of-product method (Tofighi & MacKinnon, 2011). The indirect effect of personality traits on risky riding behavior through risk perception was significant (95% CI 0.03–0.29), as was the indirect effect of social norms on risky riding behavior through risk perception (95% CI –0.41 to –0.05). The mediation was partial for personality traits and total for the social norm in the male subsample. Overall, the variables included explained 47% of the variance of risk-taking riding behavior by adolescent boys. In the female subsample, the first estimate of the model gave one unacceptable parameter – the direct relation of social norms to riding behavior – so we excluded this from the model. The fit indices for the model estimated are satisfactory: χ2(30, n = 448) = 65.631, p b 0.01; RMSEA = 0.050; NNFI = 0.978; CFI = 0.985; and SRMR = 0.038. For adolescent girls, risky riding behavior was influenced directly by personality (γ = 0.38, p b 0.001) and risk perception (β = –0.42, p b 0.001). Risk perception was in turn negatively influenced by at-risk personality traits (γ = –0.23, p b 0.01) and positively by perceived social pressure to adopt correct riding behaviors (γ = 0.76, p b 0.001). Overall, the variables included explained 51% of the variance of risk-taking riding behavior by adolescent girls. The effects of mediation in the female subsample were confirmed by the distribution-of-product method. The indirect effect of personality traits on risky riding behavior through risk perception was significant (95% CI 0.05–0.28), as was the indirect effect of social norms on risky riding behavior through risk perception (95% CI – 0.44 to –0.20). 4. Discussion The main objective of our study was to integrate the approaches based on personality traits and on social cognition to explore the relations between personality, social norms, riding optimism, risk perception, and riding behavior in a sample of adolescent Italian

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

53

Table 2 Model covariance matrix.

1. Ordinary violations 2. Aggressive violations 3. Transgressive violations 4. Emotional reactions 5. Aversion to risk taking 6. Subjective norm parents 7. Subjective norm friends 8. External locus of control 9. Sensation seeking 10. Normlessness 11. Anger 12. Driver self-assessment 13. Driver confidence 14. Driver stress

1

2

3

4

5

6

7

8

9

10

11

12

13

14

.303 .255 .253 -.120 -.143 -.122 -.132 .110 .166 .186 .010 -.008 -.003 .026

.445 .248 -.134 -.131 -.119 -.139 .102 .175 .182 .055 .017 .016 .039

.390 -.142 -.120 -.102 -.108 .106 .124 .167 -.024 .006 .000 .030

1.188 .162 .155 .177 -.071 -.137 -.183 .015 -.066 -.064 .156

.283 .156 .186 -.060 -.115 -.135 .018 .039 .012 .043

.383 .253 -.071 -.093 -.129 .018 .038 .034 .023

.562 -.020 -.114 -.123 .006 .041 .027 .045

.539 .154 .188 .005 .030 .033 .094

.389 .285 .080 .022 .034 .026

.600 .039 .032 .043 .052

.296 .022 .038 .009

.247 .169 .008

.275 -.018

.381

We based the cognitive component of risk perception on the aversion risk scale (Machin & Sankey, 2008), which requires subjects to assess the level of perceived risk or danger in relation to some specific situations and driving/riding behavior. Ulleberg and Rundmo (2003) measured the cognitive component of risk perception using two items and subjects had to evaluate the likelihood of being involved in a car accident. The authors acknowledged the low reliability of this measure for risk perception. We used the measure proposed by Machin and Sankey (2008), which shows good reliability and a significant effect on behavior. It is likely that a risk assessment of potentially real and previously experienced driving/riding situations and behaviors is a more valid measure of the construct compared to the estimated probability of being involved in an accident. Thus, further applications of both scales are desirable to verify the goodness of the overall scale used to assess risk perception, especially since risk perception, besides being a mediator, was a significant predictor of risky riding behavior by our study participants. Hence, the higher the cognitive evaluation of danger connected to risky riding behaviors and negative emotions at the thought of being involved in a traffic accident, the lower is the frequency with which young adolescents adopt risky behaviors when riding two-wheeled vehicles. These results will be useful in more effective planning of educational campaigns and of the courses required to obtain a license. From a practical perspective, it is opportune to take actions to increase risk perception and favour safe driving/riding behaviors among young people. Increasing perceived susceptibility and eliciting

moped riders. There is interest in explorative assessment of the influence of these variables on the behavior of a specific sample with a high accident rate and characterized by a propensity to engage in risky behavior. The results point to the predictive capacity of personality traits because of their direct influence on riding behavior among our adolescent participants. It is highly plausible that a 14–15-year-old adolescent may be oriented towards strong emotions and new sensation-seeking, with a craving to assert his/her own identity and overcome limits, an inclination towards anti-social behaviors, and a certain presumption of invulnerability and immunity to involvement in negative events. This makes a young person more likely to engage in risk-taking and irresponsible riding behaviors. In particular, the results show that an at-risk personality influences behavior not only directly but also indirectly by lowering the threshold of road risk perceived by young people. Previous studies showed a relation between personality, risk perception, and riding behaviors in terms of the cognitive (Machin & Sankey, 2008) or affective (Rundmo & Iversen, 2004) component of risk perception. In the study by Ulleberg and Rundmo (2003), in which both components of risk perception were examined, no significant mediation was found between personality and behavior. In the current study, the affective and cognitive components of risk perception mediated the effects of personality on risk-taking riding behavior. This result can be explained using a different scale from that applied by Ulleberg and Rundmo (2003) for the cognitive component.

Subj. norm parents Subj. norm friends

.75 Social norm .14 .72

.71 R2 = .76 Emotional reactions Aversion to risk

Ext. locus of control

.43

Sensation seeking

.79

Normlessness

.95

Ordinary violations

.74

Aggressive violations

R2 = .53

.37 Risk perception

-.57

Risky riding behavior

.76 -.29

.78

Transgressive violations

.36

Personality

.75

Fig. 1. Relations between personality, social norm, risk perception, and risky riding behavior.

54

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

Table 3 Gender differences in personality, social norm, risk perception, and risky riding behavior. Measures

Sensation Seeking External Locus of control-fate Normlessness Subjective norm, parents Subjective norm, friends Emotional reactions (risk perception affective component)† Aversion to risk (risk perception cognitive component) Ordinary violations Aggressive violations Transgressive violations

Males

Females

Difference (p value)

Mean

SD

Mean

SD

2.50 2.07 2.32 3.16 2.73 3.99

.61 .75 .77 .65 .74 1.07

2.37 1.95 2.14 3.32 2.98 4.49

.64 .71 .78 .57 .73 1.06

.13** .12** .18*** -.16*** -.25*** -.49***

3.15

.57

3.30

.46

-.16***

1.75 1.98 1.56

.58 .71 .67

1.55 1.70 1.24

.49 .57 .51

.20*** .28*** .32***

† Response scale from 1 to 7. The other response scales are from 1 to 4. ** p b 0.01, *** p b 0.001.

negative emotions, such as worry and fear, is an effective strategy for promoting preventive behaviors in the field of general health (McKenna & Horswill, 2006; Van der Pligt, 1996). However, the efficacy of negative emotions in communication strategies targeted at young people; their persuasive effectiveness has been questioned in several recent studies displaying greater caution in using guilt and shame appeals. It is becoming increasingly apparent that the use of framed messages and tactics can help young people to bear and handle the possible emotional repercussions of a message. Agrawal and Duhachek (2010) examined the effect of the interaction between emotions induced by different framed messages and a subject’s emotional state on the persuasive capacity of a message in an educational campaign aimed at reducing alcohol use among the young. The findings show that for young people in an emotional state characterized by negative feelings such as guilt and shame (emotions placing the self in an unfavourable perspective), a framed message that stimulates the negative emotion itself − compatible framing − has a counterproductive effect to the point of increasing the intention to drink alcohol. According to Agrawal and Duhachek (2010), this effect is explained by defence processes by which a young person protects him/herself from further negative self-evaluation. The subject, already beset by guilt or shame, will strongly oppose messages generating the same emotions and will be more willing to accept messages associated with other emotions. Likewise, messages incompatible with a young person’s emotional state are not effectively persuasive, which emphasizes the need for further research to identify in which conditions incompatibility leads to an effective persuasive process. Our results do not confirm a direct influence of social norms on behavior as reported in the literature (Åberg, 1993; Brown, 1998; Parker, Manstead, Stradling, Reason, & Baxter, 1992; Wallén Warner & Åberg, 2008); this may arise from having considered together the social pressure from parental/institutional figures and from peers and/or friends. It could be hypothesized that the latter are somehow counteracted by the former during adolescence. A further hypothesis is that a young person riding on two wheels with no passengers or witnesses of improper actions may feel free to act and commit violations without fear of negative consequences (Baxter et al., 1990). Lardelli-Claret et al. (2005) pointed out the protective effect of carrying a passenger on a motorcycle or moped. However, other studies have shown an increased accident risk when carrying a passenger on a motorcycle (Quddus, Noland, & Chin, 2002) or moped (Moskal, Martin, & Laumon, 2012), just as happens with young four-wheeled vehicle drivers (Harré, 2000; Simons-Morton, Lerner, & Singer, 2005). Further studies and research are needed to investigate and determine more precisely the nature and direction of this relation. In Italy, this issue involves young people older than 18 years, because younger riders/drivers cannot legally carry a passenger.

An interesting result concerns the indirect influence of social norms on behavior via risk perception. Being surrounded by parents and friends who expect a responsible and safe riding style increases road risk perception or makes riders more cautious, favoring non-risky riding behavior. This indicates that communication campaigns designed to promote road safety should consider greater involvement of family and school by including group activities and didactic lessons to strengthen social norms and to specify and clarify other people’s expectations in terms of safe and responsible riding. Another educational communication strategy that successfully influences young people’s intention to drink alcohol is media marketing focused on social norms. Social norms refer to behaviors and attitudes that a person believes are expressed by the majority of people in a group or community, that is, one’s peers. People often incorrectly perceive the norms of their peers. For instance, they overestimate their tendency to drink and drive. This risky behavior thus becomes a practice perceived as acceptable and therefore easily carried out. Modifying or correcting such normative misperception by reliably communicating data on the real norm with reference to peers (e.g., most of my peers do not drink alcohol before driving) is the objective of a social norms campaign. The young person will then tend to behave in accordance with his/her community behavioral standards and avoid drinking before driving/riding (Linkenbach & Perkins, 2005; Perkins, Linkenbach, Lewis, & Neighbors, 2010). It would be interesting to consider the efficacy of such social norm interventions for other risky behaviors in both road safety and public health in general. Our study confirms the important role of personality variables in risky driving/riding behavior. The results indicate that sensationseeking and normlessness prevail in the personality of young people at risk. As Jonah (1997) suggested, individuals with high levels of sensation-seeking, despite perceiving the risk associated with some behaviors, perform them nevertheless to satisfy their need for strong emotions. Such thrill- and emotion-seeking is associated with the possibility of infringing some socially shared rules to reach one’s goals. Such personality traits, typical of adolescence, have several specific functions related to identity development and social participation (Arnett, 1992). Our study also highlights the role of locus of control. The personal tendency to attribute the cause of traffic accidents to external and therefore uncontrollable factors probably allows a young person to justify his/her offences and dangerous riding behaviors. An external locus of control, by shifting responsibility to external events, alters risk perception, lowering its threshold and increasing the frequency of risky riding behaviors. The influence of propensity to anger while riding is not incisive. This can be explained by the specific two-wheeled context, which, unlike four-wheeled driving, allows for easier traffic flow and is associated with a lower likelihood of being stopped by the police and suffering from rude and aggressive reactions of other road users. It would therefore be interesting to examine whether the lack of influence of anger arises from the fact that riding does not determine particularly irritating and stressful situations or from the limitations of the scale used in our research for a sample of young riders. In addition, we hypothesize that young people have usually not yet been exposed to the stress and frustrations that adults often experience. In general, knowing the personal disposition to risk and its weight in influencing riding behaviors can guide and support educators in identifying people and/or groups that require focused efforts and attention in terms of social marketing and educational communication. In addition, a preliminary assessment of young people’s personality tendencies may be useful as a screening process on commencement of driving/riding education and training, such as courses taken to obtain a driving license. An unexpected result is that riding optimism (a typical trait of the young) has no influence on riding behavior, contrary to what is typically found in the literature. One of the possible explanations is that the effects of optimism may be partly absorbed by the construct of

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

risk perception: high risk perception may correspond to a low level of driving optimism. Another explanation involves measurement problems for the scale. In previous studies, optimism was generally assessed using a single item to estimate the probability of being involved in a road accident compared to the average person. In our study, we used a recently formulated three-component scale that might require excessive cognitive and evaluation effort by a young adolescent. Furthermore, there is a clear link between optimistic bias and the sense of control/controllability of events. Therefore, the more control a person has over situations, the more optimistic he/she is (Dejoy, 1989; McKenna, 1993). This control derives from experience, and the sample we investigated is characterized by a lack of driving/riding experience. The low sense of control over events and circumstances of adolescent moped riders explains the non-significant effect of optimism in our model. Therefore, before a potential factor is excluded in future studies it would be appropriate to consider a different measure of the construct or its connection to the degree of perceived control. The second objective of our study was to verify the proposed model on the basis of gender. On the whole, significant relations between the variables that emerged in the whole sample are confirmed in the male and female subgroups. Although we cannot make significant comparisons, some qualitative considerations should be reported. The male subgroup exhibits greater influence of personality on reckless and unsafe riding behavior. Conversely, the riding behavior of adolescent girls is influenced more by levels of risk perception and social norms. This gender difference, although not statistically significant, may be due to the sociocultural context and to gender stereotyping, according to which women are supposed to meet others’ expectations and tend to comply with a gender-related role that considers them as responsible and altruistic, as opposed to males, who are “allowed” to manifest greater competitiveness, aggressiveness, and risk-taking (Simon & Corbett, 1996). In the sample considered, young males showed a greater propensity for risk (Gullone & Moore, 2000; Oltedal & Rundmo, 2006; Ulleberg, 2001; Yagil, 1998). At the practical level, the results suggest that stronger actions should be targeted at boys, who are more inclined to seek strong and dangerous sensations, to have inaccurate risk perception, and to engage in risky riding behaviors because of their higher inclination towards aggressive and transgressive violations. A further step in future research should be the introduction of measures to detect the use of alcohol and/or drugs; which will allow more specific and careful actions aimed at the most at-risk cases of either gender. Overall, this paper represents a first step in a comprehensive approach to understanding factors that can influence risky two-wheeled riding behaviors among young adolescents. Subsequent actions should include assessment of the model for a broader set of data that is representative of the population surveyed. This step will facilitate verification of the relations between variables and a generalization of the results. It will also be possible to adapt educational and social communication campaigns using cognitive, social, and personality characteristics of the specific targets to improve the efficiency of prevention. Integrated and coordinated action on different levels – cognitive, social, and personality – together with communicative strategies based on positive rather than frightening and threatening messages and on social norming strategies may therefore lead to more effective and significant results in reducing risky riding behaviors that often underlie the higher involvement of young two-wheeled riders in traffic accidents.

References Åberg, L. (1993). Drinking and driving: Intentions, attitudes, and social norms of Swedish male drivers. Accident Analysis and Prevention, 25, 289–296.

55

Agrawal, N., & Duhachek, A. (2010). Emotional compatibility and the effectiveness of antidrinking messages: a defensive processing perspective on shame and guilt. Journal of Marketing Research, 47, 263–273. Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl, & J. Beckmann (Eds.), Action-control: From cognition to behavior (pp. 11–39). Heidelberg: Springer. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211. Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned behavior: A meta-analytic review. British Journal of Social Psychology, 40, 471–499. Arnett, J. (1990). Drunk driving, sensation seeking, and egocentrism among adolescents. Personality and Individual Differences, 11, 541–546. Arnett, J. (1992). Reckless behavior in adolescence. A developmental perspective. Developmental Review, 12, 339–373. Arnett, J., Offer, D., & Fine, M. A. (1997). Reckless driving in adolescence: “State” and “trait” factors. Accident Analysis and Prevention, 29, 57–63. Baxter, J. S., Manstead, A. S. R., Stradling, S. G., Campbell, K. A., Reason, J. T., & Parker, D. (1990). Social facilitation and driver behavior. British Journal of Psychology, 81, 351–360. Benevene, P., & Scopelliti, M. (2012). Building a multi-dimensional scale on attitude towards alcohol consumptions. European Journal of Social Sciences, 34(1), 58–69. Bianchi, A., & Summala, H. (2004). The “genetics” of driving behavior: parents’ driving style predicts their children’s driving style. Accident Analysis and Prevention, 36, 655–659. Boyce, E., & Geller, E. S. (2002). An instrumented vehicle assessment of problem behavior and driving style: do younger drivers actually take more risks? Accident Analysis and Prevention, 34, 51–64. Brown, S. (1998). Associations between peer drink driving, peer attitudes to drink driving, and personal drink driving. Journal of Applied Social Psychology, 28, 423–436. Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York: Guilford Press. Brown, S. L., & Cotton, A. (2003). Risk-mitigating beliefs, risk estimates, and self-reported speeding in a sample of Australian drivers. Journal of Safety Research, 34, 183–188. Brown, I. D., & Groeger, J. A. (1988). Risk perception and decision taking during the transition between novice and experienced driver status. Ergonomics, 31, 585–597. Causse, P., Kouabenan, D., & Delhomme, P. M. (2004). Alcohol-related-accident risk perception by young drivers: Some determinants of comparative optimism. Travail Humain, 67, 235–256. Chen, C. -F. (2009). Personality, safety attitudes and risky driving behaviors – Evidence from young Taiwanese motorcyclists. Accident Analysis and Prevention, 41, 963–968. Dahlen, E. R., Martin, R. C., Ragan, K., & Kuhlman, M. M. (2005). Driving anger, sensation seeking, impulsiveness, and boredom proneness in the prediction of unsafe driving. Accident Analysis and Prevention, 37, 341–348. Dahlen, E. R., & White, R. P. (2006). The big five factors, sensation seeking, and driving anger in the prediction of unsafe driving. Personality and Individual Differences, 41, 903–915. Deery, H. A. (1999). Hazard and risk perception among young novice drivers. Journal of Safety Research, 30, 225–236. Deery, H. A., & Fildes, N. B. (1999). Young novice driver subtypes: relationship to high-risk behavior, traffic accident record, and simulator driving performance. Human Factors, 41, 628–643. Deffenbacher, J. L., Deffenbacher, D. M., Lynch, R. S., & Richards, T. L. (2003). Anger, aggression, and risky behavior: A comparison of high and low anger drivers. Behavior Research and Therapy, 41, 701–718. Deffenbacher, J. L., Oetting, E. R., & Lynch, R. S. (1994). Development of a driving anger scale. Psychological Reports, 74, 83–91. Dejoy, M. (1989). The optimistic bias and traffic accident risk perception. Accident Analysis and Prevention, 21, 333–340. Dorn, L., & Machin, M.A. (2004). Learner driving experience questionnaire. Cranfield University, UK and University of Southern Queensland, Australia. Dworkin, J. (2005). Risk taking as developmentally appropriate experimentation for college students. Journal of Adolescent Research, 20, 219–241. Elliott, M. R., Shope, J. T., Raghunathan, T. E., & Waller, P. F. (2006). Gender differences among young drivers in the association between high-risk driving and substance use/environmental influences. Journal of Studies on Alcohol, 67, 252–260. Falco, A. (2007). Road prevention and safety: An instrument for measuring locus of control in the social marketing perspective. Testing, Psychometrics, Methodology, 14, 41–61. Farrow, J. A. (1985). Drinking and driving behaviors of 16–19 year olds. Journal of Studies on Alcohol, 46, 369–374. Fergusson, D., Swain-Campbell, N., & Horwood, J. (2003). Risky driving behavior in young people: prevalence, personal characteristic and traffic accidents. Australian and New Zealand Journal of Public Health, 27, 337–342. Fuji, S., & Kitamura, K. (2003). What does a one-month free bus ticket do to habitual drivers? Transportation, 30, 81–95. Gardner, M., & Steinberg, L. (2005). Peer influence on risk taking, risk preference, and risky decision making in adolescence and adulthood: An experimental study. Developmental Psychology, 41(4), 625–635. Glick, D. C., Kronenfeld, J. J., Jackson, K., & Zhang, W. (1999). Comparison of traffic accidents and chronic disease risk perceptions. American Journal of Health Behavior, 23, 198–203. Gosselin, D., Gagnon, S., Stinchcombe, A., & Joanisse, M. (2010). Comparative optimism among drivers: An intergenerational portrait. Accident Analysis and Prevention, 42, 734–740. Gullone, E., & Moore, S. (2000). Adolescent risk-taking and the five-factor model of personality. Journal of Adolescence, 23, 393–407.

56

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

Harré, N. (2000). Risk evaluation, driving and adolescents: a typology. Developmental Review, 20, 206–226. Harré, N., & Sibley, C. G. (2007). Explicit and implicit self-enhancement biases in drivers and their relationship to driving violations and crash-risk optimism. Accident Analysis and Prevention, 39, 1155–1161. Harris, P., & Middleton, W. (1994). The illusion of control and optimism about health: On being less at risk but no more in control than others. British Journal of Social Psychology, 33, 369–386. Hausenblas, H. A., Carron, A. V., & Mack, D. E. (1997). Applications of the theories of reasoned action and planned behavior to exercise behavior: A meta analysis. Journal of Sport and Psychology, 19, 36–51. Horvath, P., & Zuckerman, M. (1992). Sensation seeking, risk appraisal, and risky behavior. Personality and Individual Differences, 14, 41–52. Hoyle, R. H., Stephenson, M. T., Palmgreen, P., Lorch, E. P., & Donohew, R. L. (2002). Reliability and validity of a brief measure of sensation seeking. Personality and Individual Differences, 32, 401–414. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. Irwin, C. E., & Millstein, S. G. (1986). Biopsychosocial correlates of risk-taking behaviors during adolescence: Can the physician intervene? Journal of Adolescent Health Care, 7, 82–96. Iversen, H. (2004). Associations between driver stress, locus of control and reported risk behaviour in traffic. Norwegian University of Science and Technology, Trondheim, Norway: Unpublished Ph.D. Iversen, H., & Rundmo, T. (2002). Environmental concern and environmental behavior among the Norwegian public. Journal of Risk Research, 5, 265–279. Iversen, H., & Rundmo, T. (2004). Attitudes towards traffic safety, driving behavior and accident involvement in the Norwegian Public. Ergonomics, 47, 555–572. Jessor, R. (1987). Risky driving and adolescent problem behavior: An extension of problem-behavior theory. Alcohol, Drugs and Driving, 3, 1–11. Jonah, B. A. (1997). Sensation seeking and risky driving: a review and synthesis of the literature. Accident Analysis and Prevention, 29, 651–665. Jonah, B. A., Thiessen, R., & Au-Yeng, E. (2001). Sensation seeking, risky driving and behavioral adaptation. Accident Analysis and Prevention, 33, 679–684. Jöreskog, K. A., & Sörbom, D. (2006). LISREL 8.80 for Windows [Computer software]. Lincolnwood, IL: Scientific Software International. Katwal, N., & Kamalanabhan, T. J. (2001). Anxiety, locus of control, subjective well being and knowledge of road rules correlates of accidents among public transport drivers. IFE PsychologIA, 9, 27–38. Kline, R. B. (2010). Principles and practice of structural equation modeling (3rd ed.). New York: Guilford Press. Kohn, M., & Schooler, C. (1983). Work and personality: An inquiry into the impact of social stratification. Norwood, NY: Ablex. 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, 243–255. Lardelli-Claret, P., Jimenez-Moleon, J. J., de Dios Luna-del-Castillo, J., Garcia-Martin, M., Bueno-Cavanillas, A., & Galvez-Vargas, R. (2005). Driver dependent factors and the risk of causing a collision for two wheeled motor vehicles. Injury Prevention, 11, 225–231. Lin, T. H. (2010). A comparison of multiple imputation with EM algorithm and MCMC method for quality of life missing data. Quality and Quantity, 44(2), 277. Linkenbach, J. W., & Perkins, H. W. (2005). Montana’s MOST of us don’t drink and drive campaign. A social norm strategy to reduce impaired driving among 21-34 years-olds. Washington, DC: National Highway Traffic Safety Administration, NHTSA. Machin, M. A., & Sankey, K. S. (2008). Relationships between young drivers’ personality characteristics, risk perception and driving behavior. Accident Analysis and Prevention, 40, 541–547. Mayer, R. E., & Treat, J. R. (1977). Psychological, social and cognitive characteristics of high-risk drivers: A pilot study. Accident Analysis and Prevention, 9, 1–8. McKenna, F. P. (1993). It won’t happen to me: Unrealistic optimism or illusion of control? British Journal of Psychology, 84, 39–50. McKenna, F. P., & Horswill, M. S. (2006). Risk taking from the participants’ perspective: The case of driving and accident risk. Health Psychology, 25, 163–170. McKenna, F. P., Stanier, R. A., & Lewis, C. (1991). Factors underlying illusory self-assessment of driving skill in males and females. Accident Analysis and Prevention, 23, 45–52. McKnight, J. A. (1999). Risky driving by youth. In G. Dionne, & C. Laberge-Nadeau (Eds.), Automobile insurance: Road safety, new drivers, risks, insurance fraud and regulation (pp. 243–251). Boston: Kluwer Academic Publishers. Michael, K., & Ben-Zur, H. (2007). Risk-taking among adolescents: Associations with social and affective factors. Journal of Adolescence, 30, 17–31. Moen, B. -E. (2007). Determinants of safety priorities in transport – The effect of personality, worry, optimism, attitudes and willingness to pay. Safety Science, 45, 848–863. Moller, M. (2004). An explorative study of the relationship between lifestyle and driving behavior among young drivers. Accident Analysis and Prevention, 36, 1081–1088. Moskal, A., Martin, J. -L., & Laumon, B. (2012). Risk factors for injury accidents among moped and motorcycle riders. Accident Analysis and Prevention, 49, 5–11. National Institute of Statistics (2009). Statistica degli incidenti stradali. Anno 2008 [Traffic accidents statistics, 2008]. Rome, Italy: ISTAT. Nunnally, J. C. (1978). Psychometric theory. New York: McGraw-Hill. Oltedal, S., & Rundmo, T. (2006). The effects of personality and gender on risky driving behavior and accident involvement. Safety Science, 44, 621–628. Parker, D., Lajunen, T., & Stradling, S. (1998). Attitudinal predictors of interpersonally aggressive violations on the road. Transportation Research Part F: Traffic Psychology and Behavior, 1, 11–24.

Parker, D., & Manstead, A. S. R. (1996). The social psychology of driver behavior. In G. Semin, & K. Fiedler (Eds.), Applied Social Psychology (pp. 198–224). London: Sage Publications. Parker, D., Manstead, A. S. R., Stradling, S. G., Reason, J. T., & Baxter, J. S. (1992). Intention to commit driving violations: an application of the theory of planned behavior. Journal of Applied Psychology, 77(1), 94–101. Parker, D., Reason, J., Manstead, A., & Stradling, S. (1995). Driving errors, driving violations and accident involvement. Ergonomics, 38, 1036–1048. Paus, T. (2005). Mapping brain maturation and cognitive development during adolescence. Trends in Cognitive Science, 9, 60–68. Peck, R. C. (1993). The identification of multiple accident correlates in high risk drivers with specific emphasis on the role of age, experience and prior traffic violation frequency. Alcohol, Drugs and Driving, 9, 145–166. Pelz, D. C., & Shuman, S. H. (1973). Drinking, hostility, and alienation in driving of young men. Proceedings of the 3rd Annual Alcoholism Conference (pp. 50–74). Washington, DC: NIAAA. Perkins, H. W., Linkenbach, J. W., Lewis, M. A., & Neighbors, C. (2010). Effectiveness of social norms media marketing in reducing drinking and driving: A statewide campaign. Addictive Behaviors, 35, 866–874. Quddus, M. A., Noland, R. B., & Chin, H. C. (2002). An analysis of motorcycle injury and vehicle damage severity using ordered probit models. Journal of Safety Research, 33, 445–462. Ranney, T. A. (1994). Models of driving behavior – A review of their evolution. Accident Analysis and Prevention, 26, 733–750. Rodham, K., Brewer, H., Mistral, W., & Stallard, P. (2006). Adolescents’ perception of risk and challenge: A qualitative study. Journal of Adolescence, 29, 261–272. Romano, E., Kelley-Baker, T. -K., & Voas, R. B. (2008). Female involvement in fatal crashes: Increasingly riskier or increasingly exposed? Accident Analysis and Prevention, 40, 1781–1788. Rundmo, T., & Iversen, H. (2004). Risk perception and driving behavior among adolescents in two Norwegian counties before and after a traffic safety campaign. Safety Science, 42, 1–21. Satorra, A., & Bentler, P. M. (1994). Corrections to test statistics and standard errors in covariance structure analysis. In A. von Eye, & C. C. Clogg (Eds.), Latent variable analysis: Applications for developmental research (pp. 399–419). Thousand Oaks, CA: Sage. Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147–177. 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. Accident Analysis and Prevention, 38, 801–810. Shope, J. T., Waller, P. F., Raghunathan, T. E., & Patil, S. M. (2001). Adolescent antecedents of high-risk driving behavior into young adulthood: Substance use and parental influence. Accident Analysis and Prevention, 33, 649–658. Silbereisen, R. K., Eyferth, K., & Rudinger, E. (1986). Development as action in context. Problem behavior and normal youth development. Berlin: Springer-Verlag. Silbereisen, R. K., & Todt, E. (1994). Adolescence in context. The interplay of family, school, peers, and work in adjustment. New York: Springer. Simon, F., & Corbett, C. (1996). Road traffic offending, stress, age, and accident history among male and female drivers. Ergonomics, 39, 757–780. Simons-Morton, B., Lerner, N., & Singer, J. (2005). The observed effects of teenage passengers on the risky driving behavior of teenage drivers. Accident Analysis and Prevention, 37, 973–982. Sjöberg, L., Moen, B. E., & Rundmo, T. (2004). Explaining risk perception. An evaluation of the psychometric paradigm in risk perception research. Norway: Rotunde. Taris, T. W. (1997). Reckless driving behavior of youth: does locus of control influences perceptions of situational characteristics and driving behavior? Personality and Individual Differences, 23, 987–995. Tofighi, D., & MacKinnon, D. P. (2011). RMediation: An R package for mediation analysis confidence intervals. Behavior Research Methods, 43(3), 692–700. Tsai, V. W., Anderson, C. L., & Vaca, F. E. (2008). Young female drivers in fatal crashes: Recent trends, 1995–2004. Traffic Injury Prevention, 9, 65–69. Tsai, V. W., Anderson, C. L., & Vaca, F. E. (2010). Alcohol involvement among young female drivers in US fatal crashes: unfavourable trends. Injury Prevention, 16, 17–20. Ulleberg, P. (2001). Personality subtypes of young drivers. Relationship to risk-taking preferences, accident involvement, and response to a traffic safety campaign. Transportation Research Part F: Traffic Psychology and Behavior, 4, 279–297. Ulleberg, P., & Rundmo, T. (2003). Personality, attitudes and risk perception as predictors of risky driving behavior among young drivers. Safety Science, 41, 427–443. Van der Pligt, J. (1996). Perceived risk, comparative optimism and behavior. In J. Georgas (Ed.), Contemporary psychology in Europe: Theory, research, and applications (pp. 91–102). Seattle: Hogrefe & Huber. Vanlaar, W., Simpson, H., & Robertson, R. (2008). A perceptual map for understanding concern about unsafe driving behaviors. Accident Analysis and Prevention, 40, 1667–1673. Vanlaar, W., & Yannis, G. (2006). Perception of road accident causes. Accident Analysis and Prevention, 38, 155–161. Vassallo, S., Smart, D., Sanson, A., Harrison, W., Harris, A., Cockfield, S., & McIntyre, A. (2007). Risky driving among young Australian drivers: trends, precursors and correlates. Accident Analysis and Prevention, 39, 444–458. Vavrik, J. (1997). Personality and risk-taking: A brief report on adolescent male drivers. Journal of Adolescence, 20, 461–465. Vézina, L. (2001). High-risk drivers: A literature review. Quebec, QC: Société de l’assurance automobile du Québec. Wallén Warner, H., & Åberg, L. (2008). Drivers’ beliefs about exceeding the speed limits. Transportation Research Part F: Traffic Psychology and Behavior, 11, 376–389. Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of Personality and Social Psychology, 39, 806–820.

A. Falco et al. / Journal of Safety Research 46 (2013) 47–57

57

Wong, J. -T., Chung, Y. -S., & Huang, S. -H. (2010). Determinants behind young motorcyclists’ risky riding behavior. Accident Analysis and Prevention, 42, 275–281. World Health Organization (2007). Youth and road safety in Europe. Policy briefing. Copenhagen, Denmark: WHO Regional Office for Europe. Yagil, D. (1998). Gender and age differences in attitudes toward traffic laws and traffic violations. Transportation Research Part F: Traffic Psychology and Behavior, 1, 123–135. Yagil, D. (2001). Reasoned action and irrational motives: A prediction of drivers’ intention to violate traffic laws. Journal of Applied Social Psychology, 31, 720–740. Yu, J., & Williford, W. R. (1993). Alcohol and risk/sensation seeking: Specifying a causal model on high-risk driving. Journal of Addictive Diseases, 12, 79–96. Zuckerman, M. (1994). Behavioral expression and biological bases of sensation seeking. New York, NY: Cambridge University Press.

Alessandra Piccirelli hais a Ph.D. in Psychological Science and works as a postdoctoral researcher at the University of Padova, FISPPA Department, Applied Psychology Section - Department FISPPA, Section of Applied Psychology.

Alessandra Falco is Adjunct Professor of Work and Organizational Psychology, and Marketing and Communication at the University of Padova, - FISPPA Department FISPPA, Section of Applied Psychology Section.

Nicola Alberto De Carlo is Full Professor of Work and Organizational Psychology, and Marketing and Communication at the University of Padova, FISPPA Department, Applied Psychology Section - Department FISPPA, Section of Applied Psychology.

Damiano Girardi hais a Ph.D. in Psychological Science and works as a post-doctoral researcher at the University of Padova, FISPPA Department, Applied Psychology Section Department FISPPA, Section of Applied Psychology. Laura Dal Corso is Adjunct Professor of Work and Organizational Psychology at the University of Padova, FISPPA Department, Applied Psychology Section - Department FISPPA, Section of Applied Psychology.