Anger is related to aggression in traffic, which in turn may lead to risky driving (e.g. Lajunen, Parker & Stradling, 1998; Deffenbacher, Huff, Lynch, Oetting & Salvatore, 2000; Parker, Lajunen & Summa!a, 2002; Deffenbacher, Lynch, Filetti, Dahlen & Oetting, 2003). Anger may also lead directly to risk taking behaviour. Although evidence for the link between anger and risk exists in areas outside traffic (Lerner & Keltner, 2001; Lerner, Gonalez, Small & Fischoff, 2003), the research that has been carried out in the area of traffic is mostly correlational and therefore leaves space for alternative explanations. For example, Arnett, Offer and Fine (1996) showed that people who are in an angry state exceed the speed limit to a greater degree than people who are not angry. This might imply that an angry state leads to speeding behaviour. An alternative explanation is that certain types of people are more inclined to both experience anger and to exceed the speed limit. For example, Sensation Seeking (defined by Zuckerman (1994) as "the need to seek novel, varied, complex and intense sensations and experiences") has been associated with both anger (Zuckerman, 1994; Iversen & Rundmo, 2002) and risk (Jonah, 1997; Jonah, Thiessen & Au-Yeung, 2001; Iversen & Rundmo, 2002; Roberti, 2004). In the present study, the effects of anger on cognitive processes related to risk will be investigated. Two types of emotion effects on performance can be distinguished. Emotion may consciously trigger actions or action tendencies (Frijda, 1986), as is the case in aggression, or emotions may cause a bias in cognition (Clore & Gasper, 2000). A number of cognitive processes may be affected by emotions, for instance memory (Parrot & Spackman, 2000) and social judgement (Forgas, 1995, 1998). However, not all of these processes are related to risk taking behaviour and as such relevant for traffic safety. Two cognitive processes that are related to risk can be distinguished: optimism bias and illusion of control. Optimism (or optimistic) bias refers to the extent to which people are biased about their chances of getting involved in good or bad events. The term is also referred to as unrealistic optimism and was first described by (Weinstein, 1980). In two studies, Weinstein showed that people rate their chances to experience positive events as higher than average and their chances to experience negative events as lower than average. It is important to keep in mind that optimism bias can only be determined at group level. An individual person may rate his or her risk as lower than average and may actually be correct. However, on a group level, it is unlikely that the risk of the majority is below average. Also in the area of traffic, research shows that people tend to overestimate their driving skills and underestimate their risk of getting involved in a car crash. Several studies (Svensson, 1981; Svensson, Fischhoff & MacGregor, 1985; Dejoy, 1989;) showed that drivers are optimistically biased: they rate themselves as more skillful and safer than the average driver, and they rate their accident likelihood as less than the average driver. Perceived controllability of the accident is a strong predictor of optimism in these studies. Also McKenna (1993) found that the optimism bias disappears when taking perceived control into account. So, people think they are better drivers only when they think they have control over the situation. In general, people tend to overestimate the degree of control they can exert on a situation. Langer (1975) called this phenomenon the illusion of control. In her studies, the concept of illusion of control refers to situations in which the control is really absent. However, in traffic related studies, the concept is mostly used to refer to situations where the control is genuine, but the benefits of the control are illusory. Horswill and McKenna (1999a,b) found differences in preferred speed when people were asked to imagine they were driving themselves, and when people were asked to imagine they were a passenger. As drivers, people accepted higher speeds than as passengers, presumably because as passengers they thought their hypothetical driver was less able to cope with higher speed than themselves. The control as a driver is genuine, but this control does not necessarily lead to a decreased risk level. Several studies have shown relations between emotions and these cognitive processes. Dewberry and Richardson (1990) and Dewberry, Ing, James, Nixon and Richardson (1990) showed an inverse relationship between anxiety and optimism bias: people who were more anxious about negative life events were less inclined to be unrealistically optimistic. The authors conclude that anxiety reduces optimism, and as anxiety is an example of negative affect, the authors furthermore generalise that not only anxiety, but also negative affect in general, reduces optimism. Several studies have examined the effect of specific emotions, as opposed to general positive or negative affect, on cognitive processes. Some older studies investigated the relationship between emotions and illusion of control. Alloy and Abramson (1979) and Alloy, Abramson and Viscusi (1982) found that depressed persons are less vulnerable to the illusion of control than non-depressed persons. Lemer and Keltner (2001) found opposing effects offear and anger on risk perception: angry respondents rated situations as less risky than fearful respondents. Hemenover and Zhang (2004) showed that anger is related to optimistic evaluations, which is contrary to the conclusion of Dewberry that general negative affect reduces optimism. The last two studies used the appraisal tendency framework (Lemer & Keltner, 2000) to explain emotion-specific effects on cognitive processes. According to this framework, people who are in a specific emotional state are likely to interpret other (not necessarily related) events in line with the emotions. For example, as anger is associated with a high level of perceived control, angry people will rate situations as more controllable and therefore less risky than non-angry individuals. Sad and fearful respondents will rate situations as less controllable and therefore as more risky, whereas happy persons will display the same pattern as angry respondents. The general aim of the study reported here was to investigate the effect of anger on cognitive bias in traffic. An experimental design was used in which subjects either did or did not receive an emotion induction procedure. All subjects made judgements of traffic situations and the performance of the two groups was compared. The emotions that people experience and express in certain situations are different from person to person. As mentioned before, sensation seeking has been shown to be related to the experience of anger (Zuckerman, 1994). Therefore, in the current study the sensation seeking scale is included. It is a personality scale that aims to measure the tendency of people to engage in exciting or thrilling activities. Within the sensation seeking scale, four subscales can be distinguished: Thrill and adventure seeking, experience seeking, boredom susceptibility and disinhibition. It is expected that the emotion manipulation procedure works best with people who score high on sensation seeking. Another reason to include sensation seeking in the experiment, is that relations have been shown between sensation seeking and risk taking behaviour (Horvath & Zuckerman, 1993; Heino, 1996; Jonah et al., 2001).
Abstract