Safety Science 50 (2012) 408417

Contents lists available at SciVerse ScienceDirect

Safety Science
journal homepage: www.elsevier.com/locate/ssci

What factors can predict why drivers go through yellow trafc lights? An approach based on an extended Theory of Planned Behavior
Blazej Palat , Patricia Delhomme
IFSTTAR, Laboratoire de Psychologie de la Conduite, Satory, Versailles, France

a r t i c l e

i n f o

a b s t r a c t
Red-light running is a major road-safety problem. It is rarely fully deliberate and usually occurs when a driver tries to go through a yellow light. The present research investigates drivers motivations for continuing when the light turns yellow. A survey based on an extended version of the Theory of Planned Behavior (TPB, Ajzen, 1985) was conducted in France on a sample of drivers (n = 103) with an average age of 35.6 years (range: 1875). The drivers characteristics, TPB factors, and some additional factors accounted for a signicant part of the variance in the intention to go through a yellow light (R2 = 0.73). As for the TPB factors, attitude and the descriptive norm were signicant predictors. However, facilitating circumstances were the most predictive of all factors examined. The results are discussed in view of determining how to make drivers less inclined to go through yellow lights. 2011 Elsevier Ltd. All rights reserved.

Article history: Received 13 May 2011 Received in revised form 12 August 2011 Accepted 25 September 2011 Available online 26 October 2011 Keywords: Yellow-light running Drivers motivation TPB Survey

1. Introduction The proposal from the European Parliament and the Europe Council for a directive facilitating cross-border enforcement in the eld of road safety (2008) identied red-light running as one of the greatest causes of crashes, along with speeding, drunkdriving, and not using a seatbelt. In 2007, the last year when redlight cameras in France were still in the testing phase, there were 223,612 recognized red- and yellow-light violations (Observatoire national interministriel de scurit routire, 2008). As for the legal status of red- and yellow-light running in France, Article R412-30 of The French Road Code states clearly that drivers have to stop at red light signals. The same rule applies to non-blinking yellow lights (Article R412-31). However, there is one condition in which drivers are authorized to go through a yellow light, i.e. the light turns from green to yellow when the driver is to close to brake safely before entering the intersection. Perhaps for this reason, drivers in France are still largely disrespectful of yellow-light signals (Baromtre AXA Prvention du comportement des Franais au volant vague 7, 2011). In order to make drivers abide by red- and yellow-light signals, French authorities plan to install 800 red-light cameras by the year 2012 (Observatoire national interministriel de scurit routire, 2009). Some studies have found that installing cameras at different types of intersections to deter drivers from running red lights is effective (Lum and Wong, 2002; Retting et al., 1999, 2003). On the other hand, other research has shown that red-light cameras elicit
Corresponding author. Tel.: +33 1 30 84 39 53; fax: +33 1 30 84 40 01.
E-mail address: blazej.palat@ifsttar.fr (B. Palat). 0925-7535/$ - see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ssci.2011.09.020

abrupt reactions to changing lights, including sharp braking, which increases rear-end collisions (Erke, 2009; Obeng and Burkey, 2008; Shin and Washington, 2007). Therefore, other solutions need also to be examined in order to tackle the multifaceted problem of yellow- and red-light running. Apart from the various technical measures that have been used so far (for the review, see Kennedy and Sexton, 2009; Grembek et al., 2009) and could still be enhanced, a complementary approach consisting of making drivers more consciously respectful of trafc lights is worth considering because it could trigger a behavioral change. In all cases, research on drivers motivations is required to explore the psychological factors inuencing the decision to go through or stop at a red or yellow light. A considerable amount of research has already been carried out in order to determine what circumstances and driver characteristics are strongly associated with trafc-light violations. In general, red-light running offenders can be divided into those who run red lights deliberately, and those who fail to see the light because of some kind of distraction (Green, 2003). The FHWA suggests that the latter can be caused by drowsiness, conversing with passengers, manipulating in-car devices or other electronic devices such as a cellular phone, and eating (FHWA Safety Program). When the action is deliberate, drivers might be motivated by factors such as the delay, false judgments of safety due to low trafc volume or to cases where a driver less than 2 s ahead runs the red light as well as, congestion, expectation of green when in a platoon of cars, steep downgrades, excessive speed, or a yellow signal that is too short (Bonneson et al., 2004). Shinar (1998) observed that in addition to congestion and delay, red-light running was also correlated with red- and green-light durations. This author interprets redlight running as an aggressive behavior. In a self-report-based

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417

409

study, Porter and Berry (2001) found that being in a hurry was the most important reason for red-light running, that having a passenger in ones car could deter the driver from doing so, and contrary to what had been assumed, that red-light running was not a consequence of frustration due to either trafc conditions or other drivers behavior. Finally, red-light running has been observed more frequently on Y-shaped intersections than on T-shaped or crossshaped intersections (Al-Omari and Al-Masaeid, 2003). Past research has also pointed out several driver characteristics and behaviors associated with red-light running. Being less than 30 years old, not fastening ones seatbelt, and driving in heavy trafc increased the probability of red-light running in both camera and non-camera environments (Hebert Martinez and Porter, 2006; Porter and England, 2000; Retting and Williams, 1996). Drivers with poor driving records who were driving smaller and older cars were found to run red lights more often (Retting and Williams, 1996). Young drivers not only ran red lights more often, but also violated speed limits at the same time, especially during periods of light trafc (Yang and Najm, 2007). As for going through a yellow light, young males were found to be more likely to proceed on a yellow light, and this behavior often ni et al., 1976). The distance resulted in red-light running (Konec from the intersection at the onset of the yellow light, the vehicles approaching speed, and its position in the trafc ow have been shown to be key factors of the decision to go through a yellow light, and, as a consequence, of running the red light (Elmitiny ni et al., 1976). In addition, sensation-seeking et al., 2009; Konec drivers have been found to decide more often to proceed in situations where they should stop at a yellow light (Rosenbloom and Wolf, 2002a,b). Finally, increased attentional demands were found to alter yellow-signal detection (Fort et al., 2010). To the best of our knowledge, deeper insight into drivers motivations for running a red light has been obtained in only one study (Porter and Berry, 2001). The aim of the present study was to examine drivers motivations for going through a yellow light, using an alternative approach based on the Theory of Planned Behavior (TPB, Ajzen, 1985). Because this is an exploratory study, we chose to focus on yellow-light running, which we believe is a more common behavior than red-light running. In fact, drivers most often run red lights within the period of 4 s that follows the change to red, with the median intersection entry time being less than 0.5 s (Bonneson and Zimmerman, 2004). We can assume then, that most of the time drivers actually end up running a red light when they are trying to proceed on a yellow light, hoping they can manage to get through the intersection before the signal changes from yellow to red. Therefore, the behavior is most often a voluntary act of choosing between two distinct alternatives: ni et al., 1976). stopping, or going through the yellow light (Konec Applying the TPB seems to be a useful approach for furthering our understanding of the cognitive process underpinning the decision to stop at or go through a yellow light. First, we will briey present the TPB; then we will describe our empirical research. The TPB makes it possible to predict behavior from a limited number of factors. According to this theory, behavioral intention is the direct predictor of a targeted behavior. Intention in turn is inuenced by three factors: attitude, the subjective norm, and perceived behavioral control. Attitudes stem from beliefs about probable outcomes of the targeted behavior, weighted by values attributed to those outcomes. Subjective norms pertaining to the targeted behavior depend on the perceived behavioral expectations of others and the motivation to comply with those expectations. Perceived behavioral control results from beliefs about ones ability to control the targeted behavior and the perceived power of those beliefs. Like behavioral intention, perceived behavioral control can directly predict the targeted behavior.

The TPB has made successful predictions about behaviors in a variety of domains (for the review, see Armitage and Conner, 2001). In driver psychology, studies on speeding (Delhomme et al., 2009a; Elliott et al., 2003; Goldenbeld et al., 2000; Letirand and Delhomme, 2003, 2005) or aggressive driving behavior (Parker et al., 1998) have been based on the TPB. Almost all of these behaviors can be performed as a result of mindful planning, or simply as a result of the automatic activation of attitudes about them (Ajzen and Fishbein, 2000). The same is true of yellow-light running, which can be considered as deliberate, as a time-saving strategy, or as an impulsive behavior. Clearly, drivers may already be thinking about speeding, not stopping at crosswalks, going the wrong way down one-way streets, running yellow lights, and so forth before getting into their cars when they are, for example, particularly motivated to save time (Letirand and Delhomme, 2003, 2005, 2006). On the other hand, the decision to perform those same behaviors can be made in a split-second, at the very moment when an opportunity presents itself (a straight, empty road where a driver can easily speed up; a pedestrian entering a crosswalk as the driver approaches; a one-way street ahead; a trafc light ahead that has just turned yellow, etc.). The only difference is the moment of activation of the attitudes, norms, and control beliefs about these dangerous behaviors. Adding extensions to the basic TPB factors can help to account for more spontaneous phenomena in decision making. It can also enhance the predictive value of the model when applied to a specic domain. In this study, we measured subjective risk perception, which is the key concept in several driver-motivation models (Fuller, 1984; Wilde, 2005; Ntnen and Summala, 1976). We also measured the impact of some aspects of the situational context we consider crucial to a drivers subjective control (Delhomme, 1995; Delhomme and Meyer, 1998). In fact, subjective control and risk perception are interdependent (Horswill and McKenna, 1999; McKenna, 1993). Additionally, we included affective factors assumed to support intentions such as desire and anticipated emotions (Perugini and Bagozzi, 2001; Sandberg and Conner, 2008). We examined drivers direct experience of negative outcomes of yellow-light running in order to further determine the predictive power of attitude towards yellow-light running (Fazio and Zana, 1978; Fazio et al., 1983). Another measure we made here was the descriptive norm (Kallgren et al., 2000). The descriptive norm corresponds to observed behaviors of others. It is different than the injunctive norm which corresponds to others perceived expectations concerning the behavior in question. In some cases, a descriptive norm is a better predictor than an injunctive norm (for the review, see Rivis and Sheeran, 2003). Finally, we added a measure of past behavior that have often greatly enhanced the predictive power of the TPB (Forward, 1997; Hie et al., 2010; Letirand and Delhomme, 2005; Smith et al., 2008; Wong and Mullan, 2009). Keeping in mind the research advances in the eld of disobedience of trafc lights, we designed a study aimed at examining the relationship between TPB factors, additional factors, and the intention to run the yellow light. We focused only on deliberate disrespect of yellow lights without considering unintentional yellow-light running. Given that the TPB has not yet been applied to the context of trafc-light violations, we followed Ajzens (2002) instructions in devising the questionnaire. We also gathered information about the drivers characteristics that might offer insight into the issue.

2. Method 2.1. Participants The participants were 103 automobile drivers (68 men, 35 women). The mean age of the sample was 35.6 years (range:

410

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417

1875). On average, participants had had their drivers license for 15.7 years (range: 155). The mean reported annual number of kilometers driven was 15,350 (MED = 12,500). Among the participants, 19 drivers (16.3%) reported having been involved in a car crash within the past 3 years, two of whom reported it was a crash following yellow- or red-light running. As for trafc tickets, 13 drivers (12.5%) reported having already been ticketed for running a yellow light (3 participants) or a red light (10 participants); 58.7% of all participants had not lost any points for violating speed limits, and only 4 drivers had been ticketed for driving without having fastened their seatbelt. In addition to car driving, 14 of the participants (13.6%; 12 men, 2 women) in the sample reported using a two-wheel motor vehicle (motorcycle or scooter) which they rode an average of 8928 (MED = 7500) kilometers per year. Among them, 8 drivers reported having their two-wheel drivers license for an average of 15 years. 2.2. Procedure The questionnaire was made freely available online. It was presented as an inquiry into car drivers opinions and behaviors, and as being completely condential. E-mail invitations were sent to ofce workers in a department of a middle-sized company that authorized the distribution of such invitations, and to university students who were met on university premises and asked for their e-mail addresses. Several participants were also recruited by word of mouth. In addition, a link to the questionnaire was published on several French web forums for car enthusiasts. The survey was accessible online for 3 months, from January to March 2010. It took 20 min to ll in. 2.2.1. Questionnaire preparation In addition to reviewing the related literature, we conducted a pilot study to identify the main behavioral, normative, and control beliefs about yellow-light running, as well as any other relevant factors that needed to be introduced into the questionnaire. The pilot study consisted of semi-directive interviews with 20 car drivers, 10 males and 10 females, whose average age was 32 years (range: 2553). Questions were asked about the advantages and inconveniences of yellow-light running, about what individuals or groups of individuals were important to the interviewed driver in terms of approving or disapproving of his or her running a yellow light, about personal reasons and circumstances that could make him or her stop or not stop at a yellow light, and about driving situations or road conditions that could make him or her stop or not stop at a yellow light. Among the results of the pilot study, we selected the most frequent answers, i.e., ones that represented at least 10% of all answers to a given question. For example, among the answers to the question What are the advantages of running a yellow light?, four drivers responded: A gain of a few seconds, Time gain, Makes you save time, Saves time. Such answers were considered as identical in their meaning and were considered as expressions of the same behavioral belief. They represented 20% of the answers to this question. On the basis of these answers, we created the item Running a yellow light saves time. Following the same procedure other behavioral beliefs were established: Running a yellow light avoids abrupt braking, and Being forced to stop at a yellow light is annoying. Family members, friends, and other road users were identied as the most important individuals in normative beliefs about yellow-light running. The circumstances seen as crucial for deciding whether to stop at a yellow light were: the presence of pedestrians near a pedestrian crosswalk, a school in the vicinity of the intersection, being followed too closely by another vehicle when approaching the intersection, driving at a high speed, having

a good view of the intersection, and being familiar with the intersection. In order to put participants in the situational context, a scenario was drawn up to illustrate a situation in which a yellow light is frequently run: Imagine yourself driving the car you usually use. The weather is good. It is 15 C. You need 30 min to arrive at your destination. You are 100 m from a trafc light which is green. You are driving at a speed of 50 km/h. When you are 80 m from the light, it turns yellow. There are no police at the intersection. The scenario gave no additional factors (bad weather conditions, non-usual vehicle, or risk of a ticket) that could affect the decision to stop vs. to run the yellow light. Two speed values were mentioned in the questionnaire: 50 km/h (speed limit in urban zones) in the scenario, and 70 km/h (chosen arbitrarily) in the item illustrating a facilitating circumstance consisting of approaching the intersection at a high speed. Our intention was to propose a distance from the light when it turns yellow so that accelerating would be fully deliberate in the rst condition and imposed by dynamic factors in the second. With this in mind, we considered the Gazis et al.s (1960) denition of the dilemma zone. Given the common duration of yellow lights in France (3 s in urban zones according to the Instruction interministrielle sur la signalisation routire), at 80 m, a car approaching at 50 km/h is outside the dilemma zone (between 47.99 and 32.67 m), but a car approaching at 70 km/h is inside it (between 85.15 and 49.32 m). Before putting the questionnaire online, we conducted three pre-tests followed by semi-directive interviews. 2.3. Overview of the questionnaire The questionnaire had four sections. The rst section contained the scenario. Participants had to answer the questionnaire in reference to this scenario. The second section had the structure of the TPB. The third section contained measures related to the additional factors, and the fourth section contained measures related to the drivers characteristics. Most of the measures used ve-point Likert scales (ranging from Not at all to Absolutely). The third and fourth sections also used multiple-choice items and short-answer items. The consistency of the measures was veried using Cronbachs alpha coefcient (three or more items) or the BravaisPearson correlation coefcient (two items). Item means were then calculated for measures consisting of two or more items. Cronbachs alpha is a coefcient indicating the degree of the internal consistency of measures used in psychometric tools. Theoretically, its value can vary from 0 to 1. The higher this coefcient is, the greater the consistency of the measures under consideration. It is widely recognized that values of 0.7 and above indicate satisfactory internal consistency, i.e., they measure the same entity. The BravaisPearson correlation coefcient is commonly used to assess the degree of linear correlation between two variables. When this coefcient is above 0.7, the correlation is considered strong. 2.3.1. TPB factors The measure of a drivers intention to run a yellow light was obtained from seven items (a = 0.94). Participants rated the extent to which they thought they would try to run a yellow light in situations similar to the one depicted in the scenario, and the extent to which they thought they would do so regularly. Participants also rated the probability that they would deliberately run a yellow light in this kind of situation during their next few trips in their car. They rated the extent to which they thought the idea of running a yellow light would occur to them, the extent to which they would feel like running a yellow light, and the extent to which they would be satised and pleased with themselves for having run a yellow light in this kind of situation.

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417

411

Attitude was measured directly and through attitudinal beliefs. For the direct measurement, participants were asked to complete the following statement by rating ve pairs of adjectives on bipolar semantic differential scales (a = 0.89): Running a yellow light is. . .: negativepositive, harmfulbenecial, unpleasantpleasant, rudepolite, and unreasonablereasonable. Participants also rated three beliefs (a = 0.69) about yellow-light running: Yellow-light running saves time, Yellow-light running avoids sharp braking, and Being forced to stop at a yellow light is annoying. The subjective injunctive norm was measured using six items (a = 0.9). Each participant rated the extent to which his/her male and female family members, male and female friends, and other male and female drivers thought he/she should run a yellow light in this kind of situation. The descriptive norm was measured using six items (a = 0.92). Each participant rated the extent to which he/she thought that his/her male and female family members, his/her male and female friends, and other male and female drivers would run a yellow light in this kind of situation. Perceived behavioral control was measured using a direct method (four items, a = 0.65). Participants rated the extent to which yellow-light running seemed controllable for them, and the extent to which it depended on their own decision. They also rated the extent to which they thought stopping at a yellow light was easy, and the extent to which they thought it was possible. 2.3.2. Additional factors The questionnaire contained nine measures of additional factors. The inuence of passengers in the vehicle was assessed using six items (a = 0.92). Participants rated the extent to which they thought that their male and female family members, their male and female friends, and other male and female drivers, as passengers in the vehicle, would approve of their running a yellow light in this kind of situation. To measure the impact of deterring circumstances (four items, a = 0.89) and facilitating circumstances (seven items, a = 0.92), participants rated the probability that they would run a yellow light when a passenger was riding with them, if there were pedestrians near the crosswalk, if there was a school in the vicinity of the intersection, and if the intersection was considered dangerous (deterring circumstances). Likewise for if they were in a hurry, if they were driving in a trafc jam for 10 min, if another vehicle was following them too closely, if they were driving initially at 70 km/h (not 50 km/h as in the scenario), if they had a very good view of the intersection, if they were familiar with the intersection, and if they were driving at night (facilitating circumstances). As a measure of past behavior, participants reported the number of yellow lights they had run deliberately during the 3 months prior to the study. Direct experience of risk related to yellow-light running was measured using eight items illustrating dangerous consequences of yellow-light running. Participants put a check next to the items illustrating situations in which they had already been involved. Those who checked situations with a weak level of personal involvement obtained low scores for direct experience of risk, whereas those who checked situations with a high level of personal involvement obtained high scores. The following situations were presented: the participant had seen on television or heard on the radio about a crash caused by a vehicle that had run a yellow light, somebody had told the participant about having been involved as a passenger in a crash caused by a vehicle that had run a yellow light, somebody had told the participant about having been involved as a driver in a crash caused by a vehicle that had run a yellow light, somebody had told the participant about having caused a crash by running a yellow light, the participant had himself/herself witnessed a crash caused by a vehicle that had run a yellow light, the

participant had himself/herself been involved as a passenger in a crash caused by a vehicle that had run a yellow light, the participant had himself/herself been involved as a driver in a crash caused by a vehicle that had run a yellow light, and the participant himself/herself had caused a crash by running a yellow light. Perceived risk related to yellow-light running was measured using two items (r = 0.35, p < .05). Participants rated the probability of being ticketed for yellow-light running in the situation illustrated in the scenario, and the probability of causing a crash by running a yellow light in this kind of situation. Anticipated regret triggered by the negative consequences of yellow-light running was measured using two items (r = 0.45, p < .05). Participants rated the extent to which they would feel angry with themselves for having run a yellow light in the situation depicted in the scenario if it resulted in a crash or in a ticket. Respect for speed limits was measured using two highly correlated items (r = 0.73, p < .05). Participants rated how often they did not abide by speed limits, and the extent to which it seemed to them that they did so regularly. Driving without a seatbelt was measured using two items that were highly correlated (r = 0.72, p < .05). Participants rated how often they drove without a seatbelt, and to what extent it seemed to them that they did so regularly. 2.3.3. Driver characteristics Participants indicated whether they had a drivers license and if so, the date on which they obtained it. They reported the number of kilometers driven annually (or the number since obtaining their drivers license), the number of car crashes in which they had been involved in the 3 years prior to the study and whether these were caused by yellow- or red-light running, the number of times ticketed for yellow- and red-light running, the number of times ticketed for driving without a seatbelt, and the number of points on their drivers license they had lost for violating speed limits (the number of points was used as a measure of penalties for speeding given that this number depends on how much the actual speed exceeds the speed limit). They also indicated the highest school grade attended, their date of birth, and their sex. Participants who used a two-wheel motor vehicle indicated whether they had a two-wheel license and if so, when they had obtained it, and they reported their annual number of kilometers driving a two-wheel motor vehicle.

3. Results The results were analyzed rst by means of descriptive analyses, and then by an analysis of variance (ANOVA). Sex was used as the categorical predictor so as to examine differences deriving from the participants sex. Finally, to predict and explain the intention to run a yellow light, a hierarchical multiple regression model (based on the one used by Elliott et al. (2003)) was drawn up in four stages involving the drivers characteristics, TPB factors, additional factors, and past behavior. 3.1. Descriptive analysis 3.1.1. TPB factors The intention to run a yellow light was rather weak (M = 1.88, r = 0.96, min. = 1, max. = 5). Participants reported that they very rarely considered running a yellow light (M = 1.49, r = 1.04, item with the lowest score) and did not feel much like committing this violation (M = 2.2, r = 1.2, item with the highest score). Participants exhibited a rather negative attitude toward yellowlight running (M = 2.02, r = 0.83). They considered it rather

412

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417

unreasonable (M = 1.84, r = 1, item with the lowest score) and rather unpleasant (M = 2.33, r = 1, item with the highest score). Attitudinal beliefs were evaluated more or less neutrally (M = 2.59, r = 1.07). Participants thought that being forced to stop at a yellow light was not very annoying (M = 2.17, r = 1.31, most negatively rated belief) although they thought that yellow-light running could possibly save time (M = 2.82, r = 1.44, most positively rated belief). The subjective injunctive norm was considered rather unfavorable to yellow-light running (M = 2.11, r = 0.92). Female family members were perceived as the most unfavorable to yellow-light running (M = 1.67, r = 0.96), whereas male friends were perceived as the least unfavorable (M = 2.48, r = 1.28). Participants thought that others rarely run yellow lights in the kind of situation described in the scenario (M = 2.3, r = 0.9). Female family members were perceived as the most unwilling to run yellow lights (M = 1.95, r = 1), and the majority of male drivers were perceived as the least unwilling to do so (M = 2.71, r = 1.13). Perceived behavioral control over the decision to stop or not at a yellow light was quite high (M = 3.87, r = 0.79). Participants believed that yellow-light running depended on the drivers decision (M = 3.5, r = 1.37, control item with the lowest score) and that it was quite possible to stop at a yellow light (M = 4.3, r = 0.94, control item with the highest score). 3.1.2. Additional factors The inuence of passengers in the vehicle was regarded as rather unfavorable to yellow-light running (M = 2.35, r = 0.93). As passengers in the vehicle, female family members were perceived as the most unfavorable (M = 1.9, r = 1), whereas male drivers were perceived as the least unfavorable (M = 2.72, r = 1.09) to yellow-light running. Yellow-light running was considered rather improbable in the circumstances assumed to be deterring (M = 1.58, r = 0.82). Among the items involving deterring circumstances, the probability of going through a yellow light was rated the lowest when pedestrians were near a pedestrian crossing (M = 1.43, r = 0.87) and the highest when there was a passenger in the vehicle (M = 1.97, r = 1.17). Yellow-light running was considered moderately probable in the circumstances assumed to be facilitating (M = 2.7, r = 1.1). Among the items assessing facilitating circumstances, the probability of yellow-light running was rated the lowest at night (M = 2.29, r = 1.26) and the highest when the vehicle was approaching at 70 km/h (M = 3.08, r = 1.32). Participants reported having deliberately run an average of 2.85 (r = 3.51) yellow lights within the 3 months preceding the study. As for direct experience of risks related to yellow-light running, 69 drivers (66.4%) reported no previous experience of the risk of a crash. Eleven drivers (10.6%) reported having seen a crash on television or heard of a crash on the radio that was caused by
Table 1 Signicative sex differences in the measures. N = 103 Males Mean Injunctive norm Descriptive norm Inuence of passengers in the vehicle Perceived probability of ticket Perceived probability of crash Tickets for red- and yellow-light running Compliance with speed limits Points lost for speeding
*

yellow-light running, and four drivers (3.9%) had heard the testimony of a person who had been involved in such a situation as a passenger. Six drivers (5.8%) had heard the testimony of a person who had been involved in such a situation as a driver, two (1.9%) had heard the testimony of a person who had caused a crash by running a yellow light, 10 (9.6%) had witnessed such a situation themselves, and one (1%) had been involved in this kind of situation as a passenger. The measure of direct experience of risk (M = 0.98, r = 1.71, min. = 0, max. = 6) was not correlated with the direct measure of attitude (r = 0.11, n.s.) nor with the belief-based measure of attitude (r = 0.15, n.s.). The risk of being ticketed for running a yellow light was rated moderately (M = 2.5, r = 1.25), as was the risk of a crash (M = 2.67, r = 1.1). Participants anticipated a rather strong feeling of regret when they were ticketed for going through a yellow light (M = 4.02, r = 1.33) and also a strong feeling of regret when they caused a crash by running a yellow light (M = 4.57, r = 1.02). Participants reported being moderately obedient of speed limits (M = 2.79, r = 1.14). More specically, they admitted speeding sometimes (M = 3.35, r = 1.21) but not very regularly (M = 2.23, r = 1.24). Participants reported that they generally drove with their seatbelt fastened (M = 1.23, r = 0.69). They very rarely drove with it unfastened (M = 1.26, r = 0.77) and did not do so regularly (M = 1.19, r = 0.72). 3.1.3. Test of sex differences Sex differences were observed on the subjective injunctive and descriptive norms (see Table 1). Females more than males had the impression that other drivers (including their friends and relatives) were in favor of yellow-light running (respectively: M = 2.55, r = 0.92 and M = 1.88, r = 0.84; F(1101) = 13.55, p < .0001, g2 = 0.12) and would run a yellow light in the situation described (respectively: M = 2.7, r = 0.8 and M = 2.09, r = 0.88; F(1101) = 12.03, p < .0001, g2 = 0.11). Furthermore, females more than males had the impression that passengers in their vehicle would be in favor of yellow-light running in the situation described (respectively: M = 2.64, r = 0.72 and M = 2.19, r = 0.98; F(1101) = 5.7, p < .05, g2 = 0.05). The probability of being ticketed for having run a yellow light was rated higher by females than by males (respectively: M = 2.94, r = 1.14 and M = 2.29, r = 1.26; F(1101) = 6.55, p < .01, g2 = 0.06). The same was observed for the probability of being involved in a crash after running a yellow light (respectively: M = 2.97, r = 0.92 and M = 2.51, r = 1.15; F(1101) = 4.13, p < .05, g2 = 0.04). Males reported having been ticketed more often than females for yellow- or red-light running (respectively: M = 0.16, r = 0.37 and M = 0.03, r = 0.17; F(1101) = 4.06, p < .05, g2 = 0.04) and speeding (respectively: M = 2.99, r = 1.17 and M = 2.39, r = 0.96; F(1101)

Females SD 0.84 0.88 0.98 1.26 1.15 0.37 1.17 1.75 Mean 2.55 2.7 2.64 2.94 2.97 0.03 2.39 0.49 SD 0.92 0.8 0.72 1.14 0.92 0.17 0.96 1.17

g2
0.12 0.11 0.05 0.06 0.04 0.04 0.06 0.06

1.88 2.09 2.19 2.29 2.51 0.16 2.99 1.32

13.55*** 12.03*** 5.7* 6.55** 4.13* 4.06* 6.97** 6.4*

p < .05. p < .01. *** p < .0001.

**

Table 2 Correlations between measurements (Intent Intention, AR_tck Anticipated regret if ticketed, AR_cra Anticipated regret if crash, Att_dir Direct measure of attitude, Att_bel Attitudinal beliefs, Inj_nor Injunctive norm, Des_nor Descriptive norm, Inf_pas Inuence of passenger, PBC Perceived behavioral control, Det_cir Deterring circumstances, Fac_cir Facilitating circumstances, Pro_tck Probability of ticket, Pro_cra Probability of crash, Dir_exp Direct experience of risk, Yrs_drv Years of driving, Kil_yr Annual no. of kilometers, Pas_beh Past behavior, Tck_run Tickets for yellow and red light running, Cra_car Number of crashes, Two_whe Two-wheel vehicle use, Spe_lim Observing speed limits, Pts_spd Points lost for speeding, Sea_blt Driving without seatbelt, Tck_blt Tickets for driving without seatbelt). Intent AR_tck AR_cra Att_dir Att_bel Inj_nor Des_nor Inf_pas PBC Intent AR_tck AR_cra Att_dir Att_bel Inj_nor Des_nor Inf_pas PBC Det_cir Fac_cir Pro_tck Pro_cra Dir_exp Yrs_drv Kil_yr Pas_beh Tck_run Cra_car Two_whe Spe_lim Pts_spd Sea_blt Tck_blt Age Sex
*

Det_cir Fac_cir Pro_tck Pro_cra Dir_exp Yrs_drv Kil_yr Pas_beh Tck_run Cra_car Two_whe Spe_lim Pts_spd Sea_blt Tck_blt Age 0.66* 0.22 0.01 0.68* 0.42* 0.50* 0.51* 0.45* 0.08 1.00 0.65* 0.09 0.18 0.01 0.15 0.06 0.22 0.03 0.01 0.14 0.11 0.05 0.48* 0.04 0.12 0.20 0.78* 0.10 0.03 0.73* 0.66* 0.58* 0.56* 0.50* 0.32 0.65* 1.00 0.04 0.23 0.15 0.11 0.17 0.48* 0.06 0.07 0.04 0.13 0.12 0.20 0.08 0.14 0.19 0.14 0.20 0.10 0.11 0.15 0.11 0.04 0.04 0.03 0.09 0.04 1.00 0.36 0.01 0.09 0.09 0.03 0.15 0.10 0.03 0.12 0.05 0.03 0.11 0.02 0.27 0.31 0.07 0.05 0.28 0.33 0.10 0.07 0.08 0.04 0.18 0.23 0.36 1.00 0.14 0.14 0.04 0.14 0.11 0.09 0.10 0.06 0.10 0.00 0.20 0.09 0.20 0.02 0.13 0.07 0.10 0.17 0.02 0.01 0.02 0.14 0.01 0.15 0.01 0.14 1.00 0.01 0.09 0.20 0.17 0.00 0.03 0.02 0.09 0.02 0.06 0.13 0.07 0.04 0.12 0.05 0.06 0.09 0.04 0.07 0.11 0.02 0.15 0.11 0.09 0.14 0.01 1.00 0.33 0.15 0.17 0.17 0.10 0.03 0.17 0.48* 0.03 0.91* 0.06 0.11 0.01 0.03 0.11 0.14 0.26 0.21 0.19 0.07 0.06 0.17 0.09 0.04 0.09 0.33 1.00 0.04 0.00 0.13 0.29 0.25 0.38 0.11 0.23 0.31 0.26 0.35 0.02 0.06 0.32 0.23 0.17 0.18 0.10 0.19 0.22 0.48* 0.03 0.14 0.20 0.15 0.04 1.00 0.09 0.05 0.04 0.14 0.22 0.05 0.04 0.22 0.06 0.07 0.07 0.11 0.12 0.14 0.08 0.02 0.06 0.20 0.03 0.06 0.15 0.11 0.17 0.17 0.00 0.09 1.00 0.18 0.09 0.15 0.06 0.04 0.05 0.14 0.20 0.04 0.04 0.08 0.11 0.04 0.15 0.11 0.12 0.08 0.01 0.07 0.10 0.09 0.00 0.17 0.13 0.05 0.18 1.00 0.02 0.09 0.17 0.14 0.07 0.12 0.04 0.20 0.07 0.09 0.06 0.02 0.09 0.07 0.06 0.06 0.14 0.04 0.03 0.10 0.03 0.10 0.29 0.04 0.09 0.02 1.00 0.12 0.02 0.10 0.32 0.07 0.18 0.12 0.17 0.15 0.13 0.19 0.06 0.06 0.13 0.11 0.11 0.13 0.12 0.06 0.02 0.03 0.25 0.14 0.15 0.09 0.12 1.00 0.32 0.10 0.05 0.03 0.25 0.15 0.10 0.06 0.14 0.08 0.08 0.06 0.08 0.00 0.05 0.12 0.05 0.10 0.09 0.17 0.38 0.22 0.06 0.17 0.02 0.32 1.00 0.02 0.23 0.13 0.24 0.23 0.06 0.08 0.25 0.14 0.25 0.33 0.29 0.06 0.48* 0.20 0.03 0.00 0.02 0.48* 0.11 0.05 0.04 0.14 0.10 0.10 0.02 1.00 0.12 0.45* 0.16 0.10 0.02 0.05 0.08 0.17 0.04 0.01 0.01 0.14 0.04 0.08 0.11 0.20 0.06 0.03 0.23 0.04 0.05 0.07 0.32 0.05 0.23 0.12 1.00 0.02 0.14 0.02 0.12 0.11 0.08 0.09 0.07 0.06 0.08 0.04 0.12 0.14 0.02 0.09 0.13 0.91* 0.31 0.22 0.14 0.12 0.07 0.03 0.13 0.45* 0.02 1.00 0.03

Sexe 0.08 0.11 0.04 0.20 0.13 0.42* 0.42* 0.35 0.11 0.20 0.19 0.27 0.20 0.07 0.06 0.26 0.06 0.20 0.04 0.18 0.25 0.24 0.16 0.14 0.03 1.00

1.00 0.15 0.03 0.71* 0.67* 0.50* 0.51* 0.40* 0.22 0.66* 0.78* 0.14 0.31 0.02 0.04 0.11 0.35 0.07 0.04 0.20 0.12 0.15 0.23 0.10 0.02 0.08

0.15 1.00 0.54* 0.22 0.13 0.08 0.07 0.03 0.05 0.22 0.10 0.20 0.07 0.13 0.12 0.01 0.02 0.07 0.04 0.07 0.17 0.10 0.06 0.02 0.12 0.11

0.03 0.54* 1.00 0.05 0.16 0.10 0.03 0.07 0.14 0.01 0.03 0.10 0.05 0.07 0.05 0.03 0.06 0.11 0.08 0.09 0.15 0.06 0.08 0.05 0.11 0.04

0.71* 0.22 0.05 1.00 0.49* 0.57* 0.58* 0.50* 0.26 0.68* 0.73* 0.11 0.28 0.10 0.06 0.11 0.32 0.12 0.11 0.06 0.13 0.14 0.25 0.08 0.08 0.20

0.67* 0.13 0.16 0.49* 1.00 0.45* 0.41* 0.32 0.17 0.42* 0.66* 0.15 0.33 0.17 0.09 0.14 0.23 0.14 0.04 0.02 0.19 0.08 0.14 0.17 0.09 0.13

0.50* 0.08 0.10 0.57* 0.45* 1.00 0.90* 0.68* 0.32 0.50* 0.58* 0.11 0.10 0.02 0.04 0.26 0.17 0.08 0.15 0.09 0.06 0.08 0.25 0.04 0.07 0.42*

0.51* 0.07 0.03 0.58* 0.41* 0.90* 1. 00 0.75* 0.30 0.51* 0.56* 0.04 0.07 0.01 0.07 0.21 0.18 0.02 0.11 0.07 0.06 0.06 0.33 0.01 0.06 0.42*

0.40* 0.03 0.07 0.50* 0.32 0.68* 0.75* 1.00 0.21 0.45* 0.50* 0.04 0.08 0.02 0.11 0.19 0.10 0.06 0.12 0.06 0.13 0.08 0.29 0.01 0.08 0.35

0.22 0.05 0.14 0.26 0.17 0.32 0.30 0.21 1.00 0.08 0.32 0.03 0.04 0.14 0.02 0.07 0.19 0.20 0.08 0.06 0.11 0.00 0.06 0.14 0.04 0.11

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417

p < .0001.

413

414

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417

= 6.97, p < .01, g2 = 0.06), and having lost more points on their drivers license for speeding (respectively: M = 1.32, r = 1.75 and M = 0.49, r = 1.17; F(1, 99) = 6.4, p < .05, g2 = 0.06). 3.2. Regression model 3.2.1. Selection of variables Variables were selected on the basis of intercorrelations so as to eliminate possible multicolinearity problems (see Table 2). Given the high correlation between age and years since drivers license obtained, age was not introduced into the regression model. Since subjective injunctive and descriptive norms were also highly correlated, only the descriptive norm was introduced given its higher correlation with behavioral intention. The inuence of passengers in the vehicle was not included in the regression model due to its high correlation with both subjective injunctive and descriptive norms. Similarly, because deterring circumstances and facilitating circumstances were highly correlated, only facilitating circumstances were retained because of their higher correlation with behavioral intention. Finally, only the direct measure of attitude was introduced into the regression model, since it was more highly correlated with behavioral intention than attitudinal beliefs were. 3.2.2. Driver characteristics The variables introduced on the rst stage were: years since obtaining a license, annual number of kilometers driven, involvement in crashes in the three years prior to the study, number of times ticketed for yellow- and red-light running, number of times ticketed for driving without a seatbelt, number of points lost on ones drivers license for speeding, sex, and two-wheel motor vehicle use. Driver characteristics accounted for a signicant amount of the variance in intention (R2 = 0.16, sig. Fchange < .05). As indicated by the beta weights, the intention to run a yellow light was significantly predicted by the number of points lost on ones drivers license for speeding (b = 0.31, p < .01), two-wheel motor vehicle use (b = 0.29, p < .01), and annual number of kilometers driven (b = 0.26, p < .05). 3.2.3. TPB factors The variables introduced on the second stage were: direct measure of attitude, descriptive norm, and perceived behavioral control. Adding the TPB factors resulted in a substantial and statistically signicant increment to R2 (R2 = 0.59, R2 change 0:43, sig. Fchange < .0001). The intention to run a yellow light was significantly predicted by attitude (b = 0.59, p < .0001) and by the descriptive norm (b = 0.21, p < .05).

3.2.4. Additional factors The variables on the third stage were: facilitating circumstances, direct experience of risk, perceived risk of being ticketed, perceived risk of crash, anticipated regret if ticketed, anticipated regret if crash, obedience of speed limits, and driving without a seatbelt. The introduction of the additional factors resulted in a substantial and statistically signicant increment to R2 (R2 = 0.73, R2 change 0:13, sig. Fchange < .0001). The intention to run a yellow light was signicantly predicted by facilitating circumstances (b = 0.51, p < .0001). 3.2.5. Past behavior The addition of past behavior did not result in a statistically signicant increment to R2 (R2 = 0.73, R2 sig. change 0:001, Fchange < .63). 4. Discussion The aim of this survey using an extended TPB as a theoretical framework was to identify factors that intervene in making the decision to go through a yellow light. Based on our descriptive analysis, the intention to run a yellow light in the sample was weak, while reported attitude and the subjective norm about this behavior were unfavorable. However, the context was very important. When the circumstances were favorable to the violation (time pressure, a slowdown due to a trafc jam, a vehicle following too closely, a high approaching speed, good visibility of the intersection, being familiar with the intersection, and driving at night), participants indicated that they would probably consider running a yellow light. Our study showed that an extended TPB approach can help gain a more complete understanding of the motivational processes underlying yellow-light running, which was illustrated here by the large amount of explained variance in intention (R2 = 0.73, see Table 3). The present contribution to the research on obedience of trafc lights concerns the observed roles of attitude and the descriptive norm (others perceived behavior). Although relatively moderate compared to the role of the circumstances, the impact of these TPB factors in shaping peoples intention to proceed at a yellow light remains signicant. The most important point here is that, unlike context variables, one can act more easily on these factors in order to make drivers more respectful of trafc lights. In addition, some interesting ndings related to normative beliefs were obtained. Males reported perceiving more disapproval of yellow-light running from relatives, friends, and other road users than did females, and thought that their relatives, friends, and other road users were less likely to run a yellow light.

Table 3 Regression model in four stages. N = 103 Stage 1: Points lost for speeding Two-wheel vehicle use Annual kilometers driven Stage 2: Attitude toward yellow-light running Descriptive norm Stage 3: Facilitating circumstances Stage 4:
*

bStage

bStage

bStage

bStage

R2 change 0.16*

R2 0.16

0.31** 0.29** 0.26*

0.09 0.12 0.07 0.59*** 0.21*

0.09 0.17* 0.07 0.27** 0.06 0.51***

0.09 0.18* 0.07 0.43*** 0.27** 0.06 0.13*** 0.49*** 0.001 0.73 0.73 0.59

p < .05. p < .01. *** p < .0001.

**

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417

415

In line with previous research on red- and yellow-light running (Bonneson et al., 2002; Elmitiny et al., 2009; Hebert Martinez and ni et al., 1976; Porter and Berry, 2001), certain Porter, 2006; Konec circumstances were predictive of the intention to run a yellow light. Drivers intended to run a yellow light whenever such a decision could be justied (when another vehicle was following the driver too closely or when the driver was approaching at a high speed), whenever perceived risk was low (when there was good visibility of the intersection, when the driver was familiar with the intersection, and when he/she was driving at night), or whenever the driver was particularly motivated not to obey trafc lights (under time pressure or when stuck in a trafc jam). The perceived risks of causing a crash or getting a ticket for yellow-light running were relatively low. Our drivers must still have taken this into account, however, insofar as they foresaw running a yellow light only in circumstances that were particularly favorable. The observed lack of a relationship between behavioral intention and past behavior once again proves just how much the decision to run a yellow light depends on the immediate context. The role of context in triggering behavioral intentions could also be an argument for seeing yellow-light running as a habit (Chen et al., 2004; Verplanken and Aarts, 1999). Although habitual behavior is often performed without an explicit intention to do so (Wood et al., 2002, 2005), intentionality can be present to different extents in automatic and habitual behavior (Verplanken et al., 2007), as can other basic characteristics of automaticity: lack of awareness, mental efciency, and lack of control (Bargh, 1994). Given that we did not include any measures of observed or self-reported behavior in our study to test the relationship between intention and present behavior, and between past behavior and present behavior, we cannot draw any conclusions on this issue. It nevertheless seems possible that participants who had run a yellow light more than once in the past in order to avoid waiting through the red light may have developed some kind of automaticity triggered by specic context cues (for instance, visibility at an intersection when there are no other road users could automatically cause some drivers to run the yellow light). Several driver characteristics were also predictive of the intention to go through a yellow light. First of all, drivers who also drove a two-wheel motor vehicle had stronger intentions to run a yellow light. This is in line with studies that have documented motorcyclists risk-taking propensity (Chang and Yeh, 2007; Elliott et al., 2007b), but is somewhat contradictory to studies showing that experience in a two-wheel motor vehicle driving has no effect on car driving (Horswill and Helman, 2003). Future research is required to explain the mechanisms that make car drivers who switch from motorcycles to cars less respectful of yellow lights. Such research should take into consideration both risk perception and the perceptual and motor skills required for driving twoand four-wheel vehicles. A group of drivers that were generally disrespectful of the rules of the road could be distinguished in our study sample. Those who had lost more points on their drivers license for speeding were also those who had a greater intention to run yellow lights. It seems that the intention to go through yellow lights was linked to little driving experience. In line with studies on sex differences in drivers behavior (DeJoy, 1992; Turner and McClure, 2003), males in our sample perceived yellow-light running as less risky than females did, and seemed to be more disobedient of the rules of the road (trafc lights and speed limits). Compared to studies based on observation, the present study provides more insight into drivers motivations to run yellow lights. Our conclusions could be applied to developing the content of driver training courses or road-safety communication campaigns (Delhomme et al., 2009b). For example, given the impact of the descriptive norm, i.e., of imitating other people in adopting

a behavior as risky as running a yellow light, it seems important to convince drivers not to follow the example of offenders. If yellow-light running is an automatic behavior in the sense that it is habitual, it could be difcult to change it using persuasion. Apart from automaticity, it is also because habit formation inhibits the processing of information that is not relevant to the habitual behavior (McCulloch et al., 2008; Verplanken et al., 1997). However, a habitual behavior could perhaps be reduced in non-habitual contexts (Verplanken and Wood, 2006). Changes in a drivers personal life or career such as moving or getting a new job, which result in changes in daily itineraries, could offer an opportune moment for triggering this type of behavioral change. For this reason, additional research needs to be carried out in order to examine to what extent the decision to stop vs. go through a yellow light is more or less automatic. This knowledge would be useful in determining what populations should be targeted for future interventions (for example, the general population vs. motorists who move to another town). To the best of our knowledge, this study is the rst application of an extended TPB to the issue of trafc-light violations. It has some limitations, however. First of all, the sample size was small. With a bigger sample, we could expect some driver characteristics to have a greater predictive power in the intention to run a yellow light. Our conclusions remain valid nonetheless, and can be used to plan future studies on drivers motivations for obeying trafc lights. Like many other authors who have used the TPB as a theoretical framework, we relied solely on measures of behavioral intention and not on measures of actual behavior, which we have found extremely difcult to obtain. In future research, it would be necessary to improve the methodology by applying the TPB to obedience of trafc lights so as to shed more light on the relationship between intentions and future behavior. Past research has yielded several contradictory results showing that self-reports of driving can be either quite accurate (West et al., 1993) or inaccurate (Groeger and Grande, 1996) in predicting driving behavior in real settings. Some research (Lajunen et al., 1997) suggests that drivers who are motivated to keep a positive image of themselves are likely to conceal their past undesirable behaviors behind the wheel while lling in driver behavior scales. This is more likely in settings where anonymity is not guaranteed or where participants may draw some kind of benet from a positive self-presentation (Lajunen and Summala, 2003). This was obviously not the case in our study. We therefore believe that the measures obtained here reect actual driver behavior at trafc lights. However in the future, it will be necessary to take measures of driver behavior in real settings, given the highly signicant role of the immediate context in decision-making observed in this study. The importance of context variables in predicting the intention to run a yellow light could also be an argument for applying driving simulator approaches in combination with a TPB-based questionnaire (Elliott et al., 2007a), in order to investigate disrespect of trafc lights. In our study, facilitating circumstances were considered as a homogenous factor, while in reality the impact of each individual circumstance may be quite distinct. Exploring this issue could pave the way to fully understanding drivers reasons for proceeding on a yellow light or for running a red light. Last but not least, we only investigated yellow-light running. Even though the motives for yellow- and red-light running are clo ni sely related or largely the same (Elmitiny et al., 2009; Konec et al., 1976), the reasons for, and circumstances in which, a driver deliberately runs a red light (a rarer behavior) might be slightly different, e.g. when a driver rst stops at a red light, then checks if no vehicle is approaching on the cross street, and nally goes through the red light. Another such situation is when a driver starts moving right before the light changes from red to green (Yinon and Levian, 1995).

416

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417 Elliott, M.A., Armitage, C.J., Baughan, C.J., 2007a. Using the theory of planned behaviour to predict observed driving behaviour. British Journal of Social Psychology 46, 6990. Elliott, M.A., Baughan, C.J., Sexton, B.F., 2007b. Errors and violations in relation to motorcyclists crash risk. Accident Analysis and Prevention 39, 491499. Elmitiny, N., Yan, X., Radwan, E., Russo, C., Nashar, D., 2009. Classication analysis of drivers stop/go decision and red-light running violation. Accident Analysis and Prevention 42, 101111. Erke, A., 2009. Red light for red-light cameras? A meta-analysis of the effects of redlight cameras on crashes. Accident Analysis and Prevention 41 (5), 897905. Fazio, R.H., Zana, M.P., 1978. On the predictive validity of attitudes: the roles of direct experience and condence. Journal of Personality 46 (2), 228244. Fazio, R.H., Powell, M.C., Herr, P.M., 1983. Toward a process model of the attitudebehavior relation: accessing ones attitude upon mere observation of the attitude object. Journal of Personality and Social Psychology 44 (4), 723735. FHWA Safety Program. <http://safety.fhwa.dot.gov/intersection/resources/ fhwasa10005/brief_6.cfm> (19.07.11). Fort, A., Martin, R., Jacquet-Andrieu, A., Combe-Pangaud, C., Foliot, G., Daligault, S., Delpuech, C., 2010. Attentional demand and processing of relevant visual information during simulated driving: a MEG study. Brain Research 1363, 117 127. Forward, S., 1997. Measuring attitudes and behavior using the theory of planned behavior. In: Rothengatter, T., Carbonell Vaya, E. (Eds.), Trafc and Transport Psychology: Theory and Application. Pergamon Press, Amsterdam, pp. 353365. Fuller, R., 1984. A conceptualization of driving behavior as threat avoidance. Ergonomics 27 (11), 11391155. Gazis, D., Herman, R., Maradudin, A., 1960. The problem of the amber signal light in the trafc ow. Operations Research 8 (1), 112132. Goldenbeld, C., Levelt, P.B., Heidstra, J., 2000. Psychological perspectives on changing driver attitude and behavior. Recherche-Transports-Scurit 67, 65 81. Green, F., 2003. Red Light Running, Research Report APR 356. ARRB Transport Research Ltd. Grembek, O., Zhou, K., Li, I., Li, M., Zhang, W.B., 2009. Red-light Running Collision Avoidance: Final Report. California PATH Research Report UCB-ITS-PRR-2009-1, Institute of Transportation Studies, University of California, Berkeley. Groeger, J.A., Grande, G.A., 1996. Self-preserving assessment of skill? British Journal of Psychology 87, 6179. Hebert Martinez, K.L., Porter, B.E., 2006. Characterizing red light runners following implementation of a photo enforcement program. Accident Analysis & Prevention 38 (5), 862870. Hie, M., Moan, I.S., Rise, J., 2010. An extended version of the theory of planned behavior: prediction of intentions to quit smoking using past behavior as moderator. Addiction Research & Theory 18 (5), 572585. Horswill, M.S., Helman, S., 2003. A behavioral comparison between motorcyclists and a matched group of non-motorcycling car drivers: factors inuencing accident risk. Accident Analysis and Prevention 35 (4), 589597. Horswill, M.S., McKenna, F.P., 1999. The effect of perceived control on risk taking. Journal of Applied Social Psychology 29, 377391. Instruction interministrielle sur la signalisation routire. <http:// www.securiteroutiere.gouv.fr/IMG/pdf/3-6_2009_01_IISR_6e_200908_cle27bbb8.pdf> (03.06.11). Kallgren, C.A., Reno, R.R., Cialdini, R.B., 2000. A focus theory of normative conduct: when norms do and do not affect behavior. Personality and Social Psychology Bulletin 26 (8), 10021012. Kennedy, J., Sexton, B., 2009, Literature Review of Road Safety at Trafc Signals and Signalised Crossings, Published Project Report (PPR436), Transport Research Laboratory. ni, V.J., Ebbesen, E.B., Konec ni, D.K., 1976. Decision process and risk taking in Konec trafc: driver response to the onset of yellow light. Journal of Applied Psychology 61 (3), 359367. Lajunen, T., Summala, H., 2003. Can we trust self-reports of driving? Effects of impression management on driver behaviour questionnaire responses. Transportation Research Part F 6, 97107. Lajunen, T., Corry, A., Summala, H., Hartley, L., 1997. Impression management and self-deception in trafc behaviour inventories. Personality and Individual Differences 22 (3), 341353. Letirand, F., Delhomme, P., 2003. Augmenter la valeur prdictive de la thorie du comportement plani par lvaluation de plusieurs options comportementales: Une application du comportement de vitesse sur route. Nouvelle Revue de Psychologie Sociale 2 (3), 319331. Letirand, F., Delhomme, P., 2005. Speed behavior as a choice between observing and exceeding the speed limit. Transportation Research Part F 8 (6), 481492. Letirand, F., Delhomme, P., 2006. Accessibilit des croyances associes aux deux options dune alternative comportementale. Une piste explorer pour mieux expliquer le comportement de vitesse des jeunes conducteurs? Cahiers internationaux de Psychologie Sociale 66, 314. Lum, K.M., Wong, Y.D., 2002. A study of stopping propensity at matured red light camera T-intersections. Journal of Safety Research 33 (2), 355369. McCulloch, K.C., Aarts, H., Fujita, K., Bargh, J.A., 2008. Inhibition in goal systems: a retrieval-induced forgetting account. Journal of Experimental Social Psychology 44, 857865. McKenna, F.P., 1993. It wont happen to me: unrealistic optimism or illusion of control? British Journal of Psychology 84, 3950. Ntnen, R., Summala, H., 1976. Road User Behavior and Trafc Accidents. North Holland Publishing Company, Amsterdam.

4.1. Conclusion In the present study, the intention to run a yellow light depends strongly on the situational context, so that it can be argued that this behavior is to some extent habitual or automatic. Drivers who intend to run a yellow light perceive this violation less negatively than others do, and are convinced that others commit it as well. Male drivers especially are perceived as being less respectful of yellow lights. These results can be used to conceive arguments aimed at convincing drivers to abide by yellow- and red-light signals during driver training courses or road-safety communication campaigns. There are also a number of relationships between speeding and yellow-light running, as well as between prior twowheel vehicle use and yellow-light running while driving a car. Additional research could help explain what actually happens when a road user switches from a motorcycle to a car that makes him disrespect yellow-light signals. In order to design and implement effective countermeasures, it would also be necessary to verify to what extent yellow-light running is intentional vs. automatic. References
Ajzen, I., 1985. From intentions to actions: a theory of planned behavior. In: Kuhl, J., Beckman, J. (Eds.), Action-control: From Cognitions to Behavior. Springer, Heidelberg, pp. 1139. Ajzen, I., 2002. Constructing a TPB Questionnaire: Conceptual and Methodological Considerations, Internet Website of the University of Massachusetts. <http:// www.people.umass.edu/aizen/pdf/tpb.measurement.pdf> (20.10.09). Ajzen, I., Fishbein, M., 2000. Attitudes and the attitude-behavior relation: reasoned and automatic processes. In: Stroebe, W., Hewstone, M. (Eds.), European Review of Social Psychology. John Wiley & Sons, pp. 133. Al-Omari, B.H., Al-Masaeid, H.R., 2003. Red light violations at rural and suburban signalized intersections in Jordan. Trafc Injury Prevention 4 (2), 169172. Armitage, C., Conner, M., 2001. Efcacy of the theory of planned behavior: a metaanalytic review. British Journal of Social Psychology 40, 471499. Bargh, J.A., 1994. The four horsemen of automaticity: Awareness, intention, efciency, and control in social cognition. In: Wyer, R.S., Jr., Srull, T.K. (Eds.), Handbook of Social Cognition, Basic Processes, vol. 1. Lawrence Erlbaum Associates, Hillsdale, pp. 140. Baromtre AXA Prvention du comportement des Franais au volant vague 7. <http://www.tns-sofres.com/points-de-vue/ F2A3C4C7FCF442B7B8B7730A746C9481.aspx> (22.04.11). Bonneson, J., Zimmerman, K., 2004. Development of Guidelines for Identifying and Treating Locations with a Red-light Running Problem. FHWA/TX-05/0-4196-2. Texas Transportation Institute, The Texas A&M University System, College Station. Bonneson, J., Zimmerman, K., Brewer, M., 2002. Engineering Countermeasures to Reduce Red-light Running. FHWA/TX-03/4027-2. Texas Transportation Institute, The Texas A&M University System, College Station. Bonneson, J., Zimmerman, K., Brewer, M., 2004. Red-light Running Handbook: an Engineers Guide to Reducing Red-light-related Crashes. FHWA/TX-05/0-4196P1, Texas Transportation Institute, The Texas A&M University System, College Station. Chang, H., Yeh, T., 2007. Motorcyclist accident involvement by age, gender, and risky behaviors in Taipei, Taiwan. Transportation Research Part F 10, 109122. Chen, C., Grling, T., Kitamura, R., 2004. Activity rescheduling: reasoned or habitual? Transportation Research Part F 7, 351371. Commission of the European Communities, 2008. Proposal for a Directive of the European Parliament and of the Council Facilitating Cross-boarder Enforcement in the Field of Road Safety, COM(2008)0151, Brussels. DeJoy, D.M., 1992. An examination of gender differences in trafc accident risk perception. Accident Analysis & Prevention 24 (3), 237246. Delhomme, P., 1995. Evaluation de ses propres capacits de conduite et activit de conduite. Revue Transport et Scurit 48, 3951. Delhomme, P., Meyer, T., 1998. Control motivation and young drivers decision making. Ergonomics 41 (3), 373393. Delhomme, P., Fernandez, V., Paran, F., 2009a. Exploitation de la vague 3 de lenqute MARC. Vol. 1: Croyances vis--vis de la vitesse et risque routier (Final report, long version, convention DSCR (CNSR)-INRETS. N CV04000060). INRETS, Arcueil. Delhomme, P., Dedobbeleer, W., Forward, S., Simoes, A., Adamos, G., Areal, A., Chapp, J., Eyssartier, C., Loukopoulos, P., Nathanail, T.G., Nordbakke, S., Peters, H., Pinto, M., Ranucci, M.-F., Sardi, G.M., Trigoso, J., Veisten, K., Walter, E., 2009b. Manual for Designing, Implementing, and Evaluating Road Safety Communication Campaigns (Campaigns and Awareness Raising Strategies in Trafc Safety). Belgian Road Safety Institute, Brussels. Elliott, M.A., Armitage, C.J., Baughan, C.J., 2003. Drivers compliance with speed limits: an application of the theory of planned behavior. Journal of Applied Psychology 88, 964972.

B. Palat, P. Delhomme / Safety Science 50 (2012) 408417 Obeng, K., Burkey, M., 2008. Explaining crashes at intersections with red light cameras: a note. Transportation Research Part A 42 (5), 811817. Observatoire national interministriel de scurit routire, 2008. La scurit routire en France, Bilan de lanne, 2007. La Documentation franaise. Observatoire national interministriel de scurit routire 2009. <http://www2. securiteroutiere.gouv.fr/ressources/bilan/2009/sources/index.htm> (09.06.11). Parker, D., Lajunen, T., Stradling, S., 1998. Attitudinal predictors of interpersonally aggressive violations on the road. Transportation Research Part F 1, 1124. Perugini, M., Bagozzi, R., 2001. The role of desires and anticipated emotions in goal directed behaviors: broadening and deepening the theory of planned behavior. British Journal of Social Psychology 40, 7998. Porter, B.E., Berry, T.D., 2001. A nationwide survey of self-reported red light running: measuring prevalence, predictors, and perceived consequences. Accident Analysis and Prevention 33 (6), 735741. Porter, B.E., England, K.J., 2000. Predicting red-light running behavior: a trafc safety study in three urban settings. Journal of Safety Research 31 (1), 18. Retting, R.A., Williams, A.F., 1996. Characteristics of red light violators: results of a eld investigation. Journal of Safety Research 27 (1), 915. Retting, R.A., Williams, A.F., Farmer, C.M., Feldman, A.F., 1999. Evaluation of red light camera enforcement in Oxnard, California. Accident Analysis and Prevention 31 (3), 169174. Retting, R.A., Ferguson, S.A., Hakkert, A.S., 2003. Effects of red light cameras on violations and crashes: a review of the international literature. Trafc Injury Prevention 4 (1), 1723. Rivis, A., Sheeran, P., 2003. Descriptive norms as an additional predictor in the theory of planned behavior: a meta-analysis. Current Psychology 22 (3), 218 233. Rosenbloom, T., Wolf, Y., 2002a. Signal detection in conditions of everyday life trafc dilemmas. Accident Analysis and Prevention 34, 763772. Rosenbloom, T., Wolf, Y., 2002b. Sensation seeking and detection of risky road signals: a developmental perspective. Accident Analysis & Prevention 34, 569580. Sandberg, T., Conner, M., 2008. Anticipated regret as an additional predictor in the theory of planned behavior: a meta-analysis. British Journal of Social Psychology 47 (4), 589606. Shin, K., Washington, S., 2007. The impact of red light cameras on safety in Arizona. Accident Analysis and Prevention 39 (6), 12121221. Shinar, D., 1998. Aggressive driving: the contribution of the drivers and the situation. Transportation Research Part F 1, 137160. Smith, J.R., Terry, D.J., Manstead, A.S.R., Louis, W.R., Kotterman, D., Wolfs, J., 2008. The attitudebehavior relationship in consumer conduct: the role of norms,

417

past behavior, and self-identity. Journal of Social Psychology 148 (3), 311 334. The French Road Code. <http://www. legifrance.gouv.fr/ afchCode.do; jsessionid= 1CD161703F48F4DA3229481304451 B13.tpdjo02v_3?idSectionTA= LEGISCTA 000006177124&cidTexte= LEGITEXT000006074228&dateTexte=20110708> (08.07.11). Turner, C., McClure, R., 2003. Age and gender differences in risk-taking behavior as an explanation for high incidence of motor vehicle crashes as a driver in young males. International Journal of Injury Control and Safety Promotion 10 (3), 123 130. Verplanken, B., Aarts, H., 1999. Habit, attitudes, and planned behavior. Is habit an empty construct or an interesting case of goal-directed automaticity? In: Stroebe, W., Hewstone, M. (Eds.), European Review of Social Psychology, vol. 10. Wiley, Chichester, pp. 101134. Verplanken, B., Wood, W., 2006. Interventions to break and create consumer habits. Journal of Public Policy & Marketing (Special issue: Helping consumers help themselves: improving the quality of judgments and choices) 25 (1), 90103. Verplanken, B., Aarts, H., van Knippenberg, A., 1997. Habit, information acquisition, and the process of making travel mode choices. European Journal of Social Psychology 27, 539560. Verplanken, B., Friborg, O., Wang, C.E., Tramow, D., Woolf, K., 2007. Mental habits: metacognitive reection on negative self-thinking. Journal of Personality and Social Psychology 92, 526541. West, R., French, D., Kemp, R., Elander, J., 1993. Direct observation of driving, self reports of driving behaviour, and accident involvement. Ergonomics 36 (5), 557567. Wilde, G.J.S., 2005. On the homeostasis of risk. In: Hennesy, D., Wisenthal, D. (Eds.), Contemporary Issues in Road User Behavior and Trafc Safety. Vova Science Publishers, New York, pp. 311. Wong, C.L., Mullan, B.A., 2009. Predicting breakfast consumption: an application of the theory of planned behavior and the investigation of past behavior and executive function. British Journal of Health Psychology 14 (3), 489504. Wood, W., Quinn, J.M., Kashy, D.A., 2002. Habits in everyday life: thought, emotion, and action. Journal of Personality and Social Psychology 83 (6), 12811297. Wood, W., Tam, L., Guerrero Witt, M., 2005. Changing circumstances, disrupting habits. Journal of Personality and Social Psychology 88 (6), 918933. Yang, D., Najm, W.G., 2007. Examining driver behavior using data gathered from red light photo enforcement cameras. Journal of Safety Research 38 (3), 311321. Yinon, Y., Levian, E., 1995. Presence of other drivers as a determinant of trafc violations. The Journal of Social Psychology 135 (3), 299304.