Which theory suggests that consumers will be motivated toward achieving desirable outcomes when they anticipate positive results?

Theory of planned behavior

The theory of planned behavior (TPB), has been routinely used to examine health behaviors, including problematic alcohol use (Ajzen, 1991, 2012; Cooke, Dahdah, Norman, & French, 2016). This model suggests that the most important determinant of an individual’s behavior is their intention to perform that behavior, with three cognitive variables—attitudes, subjective norms, and perceived behavioral control—said to be the direct determinants of intention. Attitudes are relatively stable evaluative judgments about aspects of a person’s experience (e.g., an idea, a person, a behavior, etc.) that range from negative to positive and are influenced by situational factors, including observations of one’s own behavior. Attitudes represent a key explanatory variable in many theories of health behavior; research has shown attitudes predict both intention and behavior (Bem, 1967; Glasman & Albarracín, 2006; Montano & Kasprzyk, 2008). Consistent with the information above, subjective norms include two different types, descriptive and injunctive. Finally, perceived behavioral control represents the extent to which a person feels they have control over performing a desired behavior when faced with internal and external barriers and is often operationalized as self-efficacy (Bandura, 1977a, 2012).

Generally, attitudes, norms, and perceived behavior control appear to account for a large proportion of variance with respect to one’s behavioral intentions (44%) as well as in health behavior (19%) in prospective studies (McEachan, Conner, Taylor, & Lawton, 2011). With respect to the evaluation of alcohol use, a recent meta-analysis found that alcohol-related attitudes have the strongest associations with intentions (random effect size=0.62), followed by subjective norms (random effect size=0.47), and finally perceived behavioral control (random effect size=0.31); intentions, in turn, predicted drinking behavior (random effect size=0.54; Cooke et al., 2016). Thus, TPB constructs have strong predictive validity for alcohol use.

Recent research has aimed to leverage these constructs, as well as principles of cognitive dissonance, in the service of intervention. Briefly, cognitive dissonance theory suggests that individuals have an inner drive to hold their attitudes and beliefs in agreement with one another and to avoid inconsistency (Festinger, 1957; Harmon-Jones & Mills, 2019). The result of inconsistent cognitions (i.e., dissonance) has been described as an unpleasant state of arousal, negative affect, or psychological discomfort, a state that individuals are motivated to reduce or eliminate. This cognitive tension can result from discrepancies between cognitions about behaviors, perceptions, attitudes, beliefs, or feelings and may relate to the self, another person, a group, or environmental contexts (Cooper, 2007, 2012; Harmon-Jones & Mills, 2019).

Using principles of cognitive dissonance and key constructs from the TPB (attitudes & norms), an intervention was developed to reduce alcohol consumption in UK university students (Norman et al., 2018). Results indicated that participants who viewed messages targeting beliefs pertinent to TPB (attitudes and norms) that were inconsistent with their current behavior had significantly less favorable cognitions about binge drinking, consumed less alcohol, engaged in binge drinking less frequently, and had less harmful patterns of alcohol consumption during their first 6 months at university. Additional recent work aimed to target two of the key predictive variables in the TPB, attitudes and intentions, using principles of cognitive dissonance to reduce alcohol use (DiBello, Carey, & Cushing, 2018). This study, conducted with US college students, attempted to create attitude-behavior dissonance in order to make individuals feel hypocritical if they were to espouse reasons why heavy drinking is negative while engaging in the behavior themselves. A brief counter-attitudinal advocacy manipulation was adapted to the alcohol prevention context wherein individuals who reported engaging in problematic drinking responded to a writing prompt where they (a) discussed reasons why heavy drinking is harmful and (b) provided ways that students like them could avoid such alcohol related problems. This inconsistency, between their behavior and the writing activity, was used to create a feeling of cognitive dissonance in an effort to motivate behavior change. Pilot study findings indicated strong support for the feasibility and acceptability of the intervention and evidence of short-term effects in reducing drinking intentions and drinking behavior. Taken together, these findings represent an important step in the alcohol intervention field as they demonstrate the promise of the TPB as both an explanatory model of drinking behavior and provide direct targets for intervention. Consistent with the TPB, other theoretical frameworks, such as the Prototype Willingness Model, also utilize attitudes and norms as important predictors of drinking and other health related behaviors.

Theory of Planned Behavior and Theory of Reasoned Action

The Theory of Planned Behavior and Theory of Reasoned Action (Fishbein & Ajzen, 1975) both stipulate that an individual's behavior is determined by their intention to engage in that behavior. The original theory of Reasoned Action included four constructs: beliefs, attitudes, intentions, and behaviors. Building upon this theory, the Theory of Planned Behavior, added one additional construct of perceived behavioral control, which is defined as the perceived ability to control one's specific behaviors. The construct of intention, which is the most salient aspect of the models, is shaped by three primary elements: attitude (e.g., the way an individual feels about something or their opinion regarding a behavior), subjective norms (e.g., belief that others of importance to the individual approve or disapprove of their behaviors), and self-efficacy (e.g., the confidence one has that a behavior can be performed or completed). In the context of pediatric adherence, Grossoehme et al. (2016) examined the Theory of Reasoned Action for adherence in cystic fibrosis. Engaged spirituality was related to greater perceive utility of airway clearance (one primary treatment for cystic fibrosis) and more supportive norms for doing the treatment from close friends. This in turn predicted stronger intentions to perform airway clearance treatment, which predicted greater actual adherence to airway clearance. This study is an exemplar for examining the Theory of Reasoned Action/Planned Behavior in pediatric adherence.

Overall Conclusion

The Theory of Planned Behavior (Ajzen, 1991) and the General Aggression Model (Anderson & Bushman, 2002) are the two dominant social psychological theories applied to the study of cyberbullying perpetration in an effort to elucidate the variables and processes engaged in this form of antisocial behavior. Due to the nature of how cyberbullying perpetration is measured, both theories examine the distal (learned) mechanisms that detail cyberbullying perpetration and both can make similar predictions albeit via different levels of processing. For example, both theories detail the importance of cyberbullying attitudes in the prediction of cyberbullying perpetration; however, the routes to get from attitudes to behavior differ. For instance, GAM posits that attitudes are one of several learned knowledge structures that form and become automatized after learning aggression schemas that form one’s aggressive personality to predict (individually or interactively) aggression; whereas the Theory of Planned Behavior posits that attitudes do not directly predict cyberbullying, but do so indirectly through intentions.

The overarching criticism in applying these theories to the study of cyberbullying is that neither of these theories can adequately offer predictions that are unique to the online world to add incremental validity above and beyond predicting traditional bullying perpetration. This is not to suggest that these theories are invalid or not useful at predicting cyberbullying. However, the theoretical processes germane to both theories are not specific to the medium through which the bullying occurs. Interventions that claim to reduce cyberbullying may be able to do so by using a curriculum derived from these theories (see Part III); however, I posit that these interventions could be improved if the lack of incremental validity in applying these theories is acknowledged and applied.

Adolescents and young adults

The Theory of Planned Behavior, with a focus on intention and cues, was used to develop an adherence promotion intervention in which participants identified where and when they intended to take their medications (Brown, Sheeran, & Reuber, 2009). In this randomized controlled trial, adolescents and young adults completed a one-page questionnaire to indicate their intention to pair a medication with another activity at a specific time and place (Brown et al., 2009). Participants in the treatment group demonstrated improved electronically monitored adherence (93%) compared with controls (79%; effect size = 0.78) 4 weeks following intervention. Improvements were particularly notable for those at highest risk for nonadherence, such as those with suboptimal memory or understanding of epilepsy treatment. Although subgroup analyses were not conducted to examine the relative benefits for adolescents and young adults, the simplicity of this intervention suggests that it can be easily targeted specifically for this age group. Pakpour et al. (2015) conducted a randomized multicenter trial to evaluate a face-to-face intervention that used motivational interviewing techniques to facilitate change talk, action plans, and proactive anticipation of adherence barriers (an aspect of the Health Belief Model) as a way to enhance readiness to change and AED adherence. Participants created a specific action plan for medication taking to overcome barriers and to identify where, when, how, and how often they planned to take their medications (Pakpour et al., 2015). Results included significantly improved AED blood serum levels and higher self-reported medication adherence compared with active controls. This intervention relied on the premise that by asking participants to form an intention, adherence would improve (i.e., Theory of Planned Behavior).

Deliberation: Car Use as Planned Behaviour

The theory of planned behaviour is a popular theory among researchers in the transport domain. The theory has been applied to the whole spectrum of travel modes (including car use for commuting, public transport use and walking and cycling), in a range of samples (including commuters, students and city dwellers), and using a range of research designs (including correlational research, longitudinal studies and randomised controlled trials). The abundance of research on the theory of planned behaviour means that a great deal is known about which factors from the theory are important predictors of transport choices.

Reviews of this literature suggest that behavioural intentions and perceived behavioural control are the strongest predictors of private car use (Lanzini & Khan, 2017; Gardner & Abraham, 2008). When people think it would be difficult for them to reduce their car use, they are less likely to want to reduce their car use. Attitudes toward car use, and attitudes toward more sustainable modes of transport, are also related to intentions, but generally play a more modest role. A more positive attitude toward car use is associated with more frequent car use, and less frequent use of alternative transport modes. Subjective norms only appear to have a weak association with car use.

For example, a survey study of about 200 employees of an English university used the theory of planned behaviour to explain car use for commuting to campus (Mann & Abraham, 2012). Attitudes toward car use were measured by asking questions like ‘driving is a reliable way to get to campus each day’ and ‘driving is a healthy way of getting to campus each day’. Perceived behavioural control questions included: ‘I regularly do things during my journey to and from campus that require me to have a car (e.g., drop off children)’. Subjective norms were assessed by asking people to indicate their level of agreement with statements such as: ‘My colleagues think that I should use a car to get to campus each day next week’. The authors included similar questions about alternative modes of transport (public and active transport). Past behaviour (car use in the week before the survey) was a strong predictor of car use 1 week later, as was the intention to drive and perceived behavioural control. Past behaviour, intentions and perceived behavioural control explained about 80% of the variance in car use. Attitudes and subjective norms were unrelated to car use in this study.

The TPB has also been used to examine the factors related to sustainable forms of travel (public transport, active travel). A German study (Haustein & Hunecke, 2007) found that behavioural intentions and perceived behavioural control explained 38% of the variance in mode choice. Of the two, perceived behavioural control was the strongest predictor, suggesting that people who feel capable of using sustainable travel modes are more likely to do so. Attitudes and subjective norms were unrelated to sustainable travel mode choice.

Longitudinal research also points to the importance of perceived behavioural control as a determinant of car use. One such study (Armitage, Reid, & Spencer, 2013) used a longitudinal design to examine the determinants of car use over time, which allowed an analysis of causal links. There was a 12-month interval between baseline and follow-up measurements. A total of 423 people from a mostly rural county in the north of England took part in the study. Car use was assessed by asking people how often in the past 2 weeks they had driven alone in their car (i.e., solo driving). Changes in solo driving over the 12 months were predicted by solo driving at baseline (past behaviour), and perceived behavioural control. This implies that people who drove more often at baseline were more likely to drive solo after 12 months. The more people felt able to reduce the number of solo-driving trips, the less likely they were to drive solo after 12 months.

The Theory of Planned Behaviour

The theory of planned behaviour (TPB; Ajzen, 1991) is the most widely used theory in environmental psychology. The TPB has inspired a considerable amount of research in the domain of environmentally significant behaviours, such as travel mode choice (Gardner, 2009), sustainable food choices (Arvola et al., 2008), recycling (White, Smith, Terry, Greenslade, & McKimmie, 2009) and energy conservation (Abrahamse & Steg, 2009, 2011Abrahamse & Steg, 2009Abrahamse & Steg, 2011; Ajzen, Joyce, Sheikh, & Cote, 2011). The theory provides a comprehensive framework for understanding the determinants of specific intentions and behaviours. The theory assumes that behaviour is a function of people’s intention to perform the behaviour, which in turn is determined by three sets of beliefs: behavioural, normative and control beliefs.

Behavioural beliefs refer to the perceived positive or negative consequences of engaging in a behaviour and people’s evaluation of these beliefs. Together, these make up a person’s attitude towards the behaviour. Attitudes reflect an individual's (positive or negative) evaluation of a specific behaviour. Attitudes towards a behaviour are generally measured via a semantic differential that includes a pair of adjectives. People indicate where they ‘sit’ on this spectrum, which is assumed to reflect their positive or negative evaluation of a behaviour (see Box 2.1). Attitudes can also be measured by asking people to indicate their level of agreement (from ‘strongly disagree’ to ‘strongly agree’) with particular statements relating to a behaviour (‘I like the idea of driving to work’, or ‘Conserving energy this winter is too much of a hassle for me’). People form many behavioural beliefs, but it is assumed that only a small number of readily accessible beliefs influence attitudes at any given time (Ajzen, 2015).

Box 2.1

Measuring the TPB variables (sample questions)

Normative beliefs include perceived expectations and behaviours of important other people, and combined with an individual’s motivation to comply with these expectations they form subjective norms. Subjective norms refer to the perceived social pressure to engage in a specific behaviour. A typical subjective norm question (see Box 2.1) would be worded along the lines of ‘People who are important to me approve of me [doing behaviour X]’, on a scale from ‘strongly disagree’ to ‘strongly agree’.

Control beliefs have to do with the extent to which people perceive there to be factors that may influence their ability to perform a behaviour. These control beliefs include factors that can facilitate or interfere with people's ability to do the behaviour, such as time, money, skills and abilities. This is known as perceived behavioural control and is also measured via a semantic differential. A sample question could be: ‘For me, to [do behaviour X] in the following week is: completely impossible to definitely possible’ (see Box 2.1).

Lastly, behavioural intentions reflect people's inclination to perform a behaviour. People who set the intention to do a specific behaviour are generally more likely to do so (‘I intend to buy organic foods next time I go grocery shopping’). Attitudes, subjective norms and perceived behavioural control predict people's behavioural intentions, and in turn, people with strong intentions are expected to engage in the behaviour when they have an opportunity to do so. In some cases, perceived behavioural control affects behaviour directly (and not indirectly via intentions).

The TPB provides insight into the barriers and enablers of environmentally significant behaviours. In general, people with more favourable attitudes towards a behaviour, with stronger endorsement for the behaviour from important others and with higher levels of perceived behavioural control will likely form stronger intentions to engage in the behaviour (see Fig. 2.1). Similarly, people with less favourable attitudes, lower levels of endorsement from other people and lower levels of perceived behavioural control are less likely to form intentions to engage in the behaviour in question. For example, the reasons why some people may refrain from reducing their meat consumption may result from a negative evaluation (attitude) of vegetarian foods, perceived social pressure to eat meat and perceived barriers to preparing vegetarian meals.

Behavioural choices can be explained in terms of perceived benefits and perceived costs. The assumption of the TPB is not necessarily that people always make rational decisions, but rather that ‘people’s intentions and behaviours follow reasonably and consistently from their beliefs no matter how these beliefs were formed’ (Ajzen, 2015, p. 127). The relative importance of attitudes, subjective norms and perceived behavioural control differs for different types of behaviours. Some behaviours, such as car use and energy conservation behaviours are more strongly correlated with perceived behavioural control, while other behaviours, such as recycling and buying organic foods are more strongly correlated with attitudes and social norms.

Organic Food Choices and the Theory of Planned Behaviour

Researchers frequently apply the theory of planned behaviour when predicting organic food consumption. Attitudes towards organic foods refer to people’s beliefs about organic foods (and their positive or negative evaluation of these beliefs), such as taste, health and the environment. Subjective norms refer to perceived social pressure to buy organic foods, that is, when important other people endorse the purchase of organic foods, people will likely have stronger intentions to buy organic foods. Perceived behavioural control refers to people’s perceptions of personal control over what they buy and eat and includes factors such as time, affordability, availability and recognition of organic products (e.g., organic labelling).

In studies on consumer food choices, perceived behavioural control is often divided into two separate variables, called perceived consumer effectiveness (PCE) and perceived availability. PCE refers to an individual’s perceived belief that her or his purchase will achieve an end goal (e.g., ‘I believe that by purchasing certain kinds of food, I can have a substantial positive impact on the environment’). Perceived availability refers to consumer perceptions of the availability of organic foods where they would normally shop for groceries.

Intentions to buy organic instead of conventional foods are most strongly correlated with attitudes towards organic foods (e.g., Arvola et al., 2008; Chen, 2007; Zhou, Thøgersen, Ruan, & Huang, 2013). People who have more positive attitudes towards organic foods express stronger intentions to buy organic foods. Subjective norms (e.g., whether important other people endorse buying organic) are also often a significant predictor of behavioural intentions. Perceived behavioural control is generally not, or only weakly, related to behavioural intentions (Scalco, Noventa, Sartori, & Ceschi, 2017).

There appear to be some differences between countries in terms of how important these three variables are to people. Arvola et al. (2008) compared the importance of the variables of the theory of planned behaviour to explain intentions to buy organic apples (instead of conventional ones) in three European countries (Italy, Finland and the United Kingdom). The survey included the standard TPB measures of attitudes using a semantic differential (e.g., ‘Buying organic apples instead of conventional apples is harmful/beneficial’), subjective norm (e.g., ‘Most people I value would buy organic apples instead of conventional apples’) and perceived behavioural control (e.g., ‘If I wanted to, it would be possible for me to buy organic apples instead of conventional apples’). In all three countries, attitudes towards organic apples were positively associated with intentions to buy organic apples. Perceived behavioural control was unrelated to behavioural intentions in the three countries. But, the role of subjective norms was different in the three places: in the United Kingdom and Finland, subjective norms were associated with intentions to purchase organic apples, while in Italy, subjective norm was unrelated to intentions. A study of consumers in the Czech Republic who regularly buy organic food found that subjective norms were the most important predictor of intentions to buy organic foods (Zagata, 2012). The author attributes this to the specific context of the Czech organic market, where consumer knowledge about organic food is relatively low, and argues that in deciding whether to by organic foods, people may then rely more on what other people think about organic food.

A recent meta-analysis on the use of the theory of planned behaviour in the domain of organic food consumption indicates that it is a suitable model for explaining willingness to buy organic foods (Scalco et al., 2017), explaining between 50% and 80% of the variance in intentions to buy organic foods. However, the study authors draw attention to the fact that most studies included a measure of behavioural intentions, but not organic purchasing behaviours, and call for more research on the determinants of organic food choices. There is also relatively little research comparing the determinants of different types of organic food groups. The study findings by Arvola et al. (2008) and by Magnusson et al. (2003) – discussed above – indicate that different organic food products might be related to different determinants. More research is needed to explore the determinants of different behaviours. If a choice for certain organic products is mainly driven by environmental concerns (e.g., milk), while others are driven more by health concerns (e.g., meat), then campaigns to promote the uptake of specific organic food groups could emphasise these aspects more prominently.

Researchers have added other concepts to the theory of planned behaviour, to increase the explanatory power of the model. Most often, researchers have used the personal norm concept (i.e., moral considerations) from the norm-activation model, the self-identity concept (discussed earlier in this chapter) and the Schwartz values.

iZ Hero

Another intervention that applies the Theory of Planned Behavior is iZ Hero. This intervention is an online computer game with the goal of teaching youth about being good online citizens by demonstrating how to make safe informed online decisions. According to Khoo (2012), while children are enjoying the various aspects of the video game, iZ Hero can simultaneously provide them with the skills to mature morally and socially. iZ Hero focuses on changing cyberbullying attitudes through peer mentors for the children and teaching them about positive digital citizenship (Hutson, Kelly, & Militello, 2018). During game play, iZ Hero players navigate through a virtual world by defeating various monsters who externalize several risky online behaviors, such as cyberbullying perpetration, video game addiction, sex grooming, privacy issues, and online pornography. By defeating these monsters, the iZ Hero character gains power and statute in the virtual world all the while teaching youth about such behaviors. This intervention applies the Theory of Planned Behavior through changing attitudes toward myriad risky online behaviors, to reduce subsequent behavior.

Liau, Park, Gentile, Katna, Tan, and Khoo (2017) sampled 440 Singaporean youth who completed a pretest battery of questionnaires and then played iZ Hero either before (experimental group) or after (control group) the posttest battery. Results showed that the youth in the experimental group had a larger decrease in scores on the cyberbullying and offline meeting scale than those in the control group. Indeed, inspection of Fig. 17.2 shows a sharper decrease in this measure; however, iZ Hero addresses many additional online topics and the results from the same validation study showed a stronger decrease for those in the experimental condition (compared to those in the control) for attitudes toward offline meetings, attitudes toward online disclosure, and attitudes toward playing less video games (Liau et al., 2017).

6.2.3 Behavior-specific beliefs, attitudes, and habits

Whether individuals choose to engage in a specific behavior may depend on how they weigh its costs and benefits. The theory of planned behavior (TPB, Ajzen, 1991) offers a useful framework. TPB proposes that intentions to engage in a behavior are the outcome of three factors: attitudes about the behavior, subjective norms (i.e., social pressure), and perceived behavioral control (i.e., perceived ability to enact the behavior). Each of these, in turn, is influenced by specific beliefs. Attitudes form in response to beliefs about the consequences of the behavior, subjective norms are based on beliefs about what valued peer groups think, and perceived behavioral control arises from beliefs about the feasibility of a behavior and one’s personal capabilities. TPB has successfully explained variance in a wide range of climate-related intentions and behaviors including purchase of energy-efficient light bulbs (Harland, Staats, & Wilke, 1999), public transportation use (Bamberg, Ajzen, & Schmidt, 2003), and interest in adopting solar panels (Korcaj, Hahnel, & Spada, 2015; Wolske et al., 2017). Variables from DOI theory nicely complement TPB for explaining interest in innovative energy technologies such as solar PV and alternative fuel vehicles. Similar to behavioral beliefs and attitudes in TPB, DOI posits that an innovation is more likely to be adopted the more it is perceived to have favorable characteristics. These include a relative advantage over prior practices, low complexity to learn and use, compatibility with existing values and routines, the ability to adopt on a trial basis, and observable evidence that others have adopted the innovation successfully.

Numerous types of beliefs have been associated with households’ financial investments to reduce energy consumption (Balcombe, Rigby, & Azapagic, 2013; Kastner & Stern, 2015). These include perceptions about expected net consequences for the household (e.g., financial costs and benefits, convenience, changes to the comfort and esthetics of one’s home, independence in energy supply, and changes in social status) as well as consequences to others or the environment (e.g., limiting climate change, reducing dependence on foreign fossil fuels, and improving the environment for future generations). How people evaluate these consequences may also be a function of their underlying dispositions. For example, individuals with strong proenvironmental personal norms may be less deterred by the inconvenience of certain actions.

Habits and interventions to change them are a familiar topic for psychological research. However, as Table 6.1 makes clear, they are relevant only to daily energy-using behaviors (type D), which together have RAER of less than 20% of all the potential of the behaviors in the table. Among these behaviors, the ones with the greatest technical potential concern travel routines, which have not been studied much by psychologists and likely are highly dependent on contextual factors such as the availability of alternative forms of transportation, as well as on intrapersonal factors.

2.4.6 Individual Differences in Preferred Task Demand and Difficulty

Accumulating evidence reveals that drivers vary in their individual dispositions to adopt a particular level of task demand. In a study with Steve Stradling’s group at Napier University (project HUSSAR—high unsafe speed accident reduction), we interviewed a national sample of British drivers, and in part of this we presented respondents with a picture of a single carriageway rural road and asked them about two speeds: What speed would they normally drive and what speed would put them right at the edge of their safety margin? There was wide variation in preferred speed: 81% of the sample were distributed over a range of nearly 30 mph (36–64 mph). Furthermore, 7% indicated a speed lower than 36 mph and 11% a speed higher than 64 mph. There was similarly wide variation in what speed they thought would put them right at the edge of their safety margin. A majority (61%) said that a speed less than 65 mph would do so. Twenty-two percent said a speed of 65–74 mph, 11% said a speed between 75 and 84 mph, and 6% said a speed of 85 mph or faster would do so (Stradling et al., 2008).

Despite this wide individual variation in preferred and in edge-of-safety margin speeds, there was a consistent relationship between the two speeds: Edge-of-safety margin speed represented a 14% increase over preferred speed. Furthermore, feelings of risk and stress did not vary with speed chosen: The feeling of risk was similar, whether one was a slow or a fast driver on the same segment of roadway. This suggests that despite variations in speed choice, perceived task difficulty may have been much more equivalent among drivers.

Project HUSSAR also confirmed, on the basis of a 12-year literature review, the national survey, and four focus groups, earlier findings by Musselwhite (2006) that there are four distinguishable groups of drivers. We have labeled them low risk threshold, high risk threshold, opportunistic, and reactive (Fuller, Bates, et al., 2008). Risk threshold in this context refers to the upper limit of task difficulty a driver will accept (i.e., the smallest separation between perceived task demand and capability).

Low risk threshold drivers comply with speed limits, reduce their speed if they realize they are traveling faster than the speed limit, and are unlikely to change their driving behavior in a 30 mph (50 km/h) zone as a result of momentary influences, including if they are in a hurry. They are typically older, more experienced, and represent approximately 40% of male and female drivers.

In marked contrast, high risk threshold drivers have positive attitudes to high-risk behavior and a thrill-seeking and expressive use of their car (Machin & Sankey, 2008), often as part of a youth subculture that exploits driving as a recreational activity that is functionally related to their life situation (Møller & Gregersen, 2008). They drive at higher speeds; commit more, and more extreme, speed limit violations and other forms of dangerous driving behavior; and have more convictions. Not surprisingly, they are more involved in collisions. Members of this group are typically young, inexperienced, and male, and they are poorly calibrated. They represent approximately 14% of drivers.

The origins of the driving style of at least some members of this group may date back to early childhood. In a seminal paper by Vassallo et al. (2007), which was concerned in part with identifying longitudinal precursors of high-risk driving behavior, three clusters of drivers were identifiable at ages 19 and 20 years who differed reliably in their engagement with risk-related driving behaviors, such as excessive speeding, drink driving, drug driving, driving when fatigued, and not using seat belts. Members of the high-risk group, which comprised 7% of their sample of 1135 young adults, were mainly male (77%) and were found to have been involved in more speeding offenses and collisions. Compared with others, they were more antisocial in behavior and choice of friends, more aggressive, more irresponsible, showed less empathy, and were more likely to engage in maladaptive coping (e.g., multisubstance use). However, particularly intriguing in their findings was that the characteristics of antisocial behavior and aggressiveness differentiated between the groups as early as ages 5–8 years and persisted throughout later childhood and adolescence. Does this finding imply that we can identify certain types of high-risk driver as soon as they are old enough to go to school? If so, what implications might this have for early intervention?

Opportunistic drivers do not pursue high speed for its own sake, unlike the high risk threshold drivers. They tend to adjust their speed to the conditions rather than to the speed limit, and they will exceed the limit if they believe it is safe to do so. They exploit opportunities to get ahead. Approximately 23% of drivers can be labeled as primarily opportunistic, and they are more likely to be male than female. The latter, on the other hand, are more likely to be reactive drivers. This group is not persistently concerned with making good progress and tends to avoid unsafe high speed and dangerous overtaking. However, such drivers can be strongly influenced by their emotional state, driving faster if annoyed or angry or under time pressure. Consistent with this is the finding in a questionnaire study by Björklund (2008) that women drivers report more irritation than men when impeded or exposed to reckless driving, and evidence presented by Lustman and Wiesenthal (2008) that female drivers report more aggression than men when feeling low levels of anger in similar scenarios.

Dispositional influences on driver risk threshold, and therefore speed choice (potentially), are partly captured by the social and cognitive variables that form the core elements of the theory of planned behavior (TPB), notably intentions, attitudes, and perceived social norms. However, correlations between measures of these variables and measures of actual behavior are not particularly strong, perhaps explaining approximately 25% of the variance in the behavioral variable (Åberg & Wallén Warner, 2008), and it is perhaps self-evident that such a conceptual approach cannot provide a comprehensive model of dispositional influence and most certainly not an account of real-time speed decisions by drivers. Thus, Paris and Van den Broucke (2008) conclude in an evaluation of TPB that actual speeding behavior can only partially be predicted from TPB concepts and that

the cognitive determinants of safe driving as identified by the TPB need to be complemented by other factors, including less “conscious” cognitive factors such as personal identity and habit formation, as well as external factors, such as cues to action, reinforcers, or the design of roads. (p. 179)