Most traffic accidents are the result of risky behaviors performed by drivers. Cognitive-motivational theoretical models of traffic psychology analyze how drivers perceive risk and make decisions. Experimental evidence suggests that a relatively stable cognitive bias tends to exist among drivers, leading them to determine how likely they will be to have an accident (Harré et al., 2005). In some cases, the perception of risk of the driver may be connected to the situation of assumed risk, causing the driver to behave cautiously, judiciously, etc., as predicted by the so-called Zero-Risk Model (Näätänen and Summala, 1974). In other cases, when faced with a challenging situation, drivers may adopt driving skills focused on elements such as technique, ability, and mastery. Such skills explain the subjective perception the driver maintains regarding the perceived risk of being involved in a road accident and the perceived level of risk he or she is willing to assume, as outlined in Risk Homeostasis Theory (Wilde, 1982). As contended by Task Difficulty Allostasis (Fuller, 2011), this sense of risk may vary, impacting the adjustment to the perceived difficulty of the task and the prior driving skills and experience of each individual as the difficulty perceived increases.

The scientific literature encompasses numerous studies that focus on risky driving, which has been associated with different factors: tailgating, driving under the influence of drugs or alcohol, distractions, road rage, speeding, drowsy driving, and the non-use of safety belts (Dula et al., 2011; Mãirean and Havârneanu, 2018). There are fewer studies that analyze the cognitions generated by the subject regarding the “why and how” things happen. Malle and Tate (2006), for example, studied both the reasoned explanations provided by drivers regarding their intentional behavior and how the drivers believed they could achieve the desired results. Our focus here is on a line of research analyzing alternative thoughts generated after the event has taken place or after the result has been previously obtained. This type of thinking is known as “counterfactual thinking” (Roese and Morrison, 2009; Byrne, 2016; Epstude, 2018).

These types of thoughts are important because they imagine changing certain aspects of the mental representation of reality. The cognitive process developed is the creation of imagined alternatives and a process of comparison between these alternatives and the actual results obtained (Byrne, 2016). Therefore, the focus of counterfactual thinking is on thoughts related to what might have been, or how the past might have been different had certain aspects been different (Smallman and Summerville, 2018). Counterfactual thinking is represented as conditional propositions, which contain an “if” antecedent followed by a “then” consequent. This type of conditional structure commonly reflects a causal inference: “If I had been more cautious, (then) the crash could have been avoided” (that is, driving more cautiously is enough to avoid a collision). In this case, neither the antecedent action nor the consequent result happened, so counterfactual thinking focuses on what could have been, not on what happened. We can contemplate real relationships in this causal inference that the subject regards as very probable (i.e., if I drive prudently, then I avoid a collision with other vehicles). These types of thoughts, therefore, are meant to explain the past while simultaneously preparing for the future. They involve various associations, including causal associations, and affect intentions as well as decisions. Counterfactual thoughts are, therefore, precursors in the formation of intentions to carry out future actions.

Different counterfactuals can be contemplated, depending on the following dimensions. If we consider a “direction,” two types of counterfactuals can be distinguished: upward, which reflects mental simulations where the possible results are better than those obtained in reality; and downward, which is related to mental simulations where the possible results are worse than those obtained in reality. Conversely, if we consider the “structure” dimension, a counterfactual thought could be additive, as when an event could or should have transpired [e.g., if I had respected the zebra crossing, (then) I would not have hit the pedestrian], as opposed to when the event should not have occurred (e.g., if I had not gone so fast, I wouldn’t have crashed); it would reflect a subtractive counterfactual. Finally, there is the “focus” dimension. A counterfactual thought could focus on oneself or other people or circumstances. This dimension could also focus on controllable vs. uncontrollable factors, although the factors in this dimension are typically covaried. One possibility would be that self-focused counterfactuals were controllable (e.g., If only I had paid more attention …) and other focused counterfactuals were uncontrollable (e.g., If it had not been raining…). Another possibility would be that self-focused counterfactuals were uncontrollable (e.g., If I were younger …) and other-focused counterfactuals were controllable (e.g., If there had been more surveillance on the highway …) (Epstude, 2018; Smallman and Summerville, 2018).

The connection between counterfactual thoughts and emotions is established under the framework of “what could have been,” compared to the current moment (Suls and Wheeler, 2000; Mandel, 2003; Markman et al., 2008; Epstude and Roese, 2011; Tasso et al., 2017; Epstude, 2018). Seta et al. (2015) reported that when alternative worlds are better than the results experienced, emotion and cognition may be more closely linked to alternative realities than to the results obtained. Nevertheless, as a general rule, when the reality is compared with its counterfactual alternative, the emotions can be amplified (Giguère et al., 2014; Byrne, 2016). If the discrepancies between the result obtained and the desired result are minor, positive emotions may appear (e.g., relief, satisfaction, and sympathy). Conversely, greater discrepancies may result in the appearance of negative emotions (e.g., regret, guilt, and shame). In particular, under the latter condition, individuals may demonstrate repetitive, intrusive, and negative cognitions, referred to as ruminative thoughts, which may amplify negative affective states, making it difficult to solve problems (Spellman and Mandel, 1999; Wade et al., 2008; Cavalera et al., 2018).

Specifically in the case of upward counterfactual thoughts when the antecedent focuses on a personal choice (e.g., “If I had not drunk, I would have avoided road collisions”), moral emotions of shame or guilt appear if the established criteria and norms have not been met in the performance of the appropriate behaviors (Zeelenberg et al., 1998; Etxebarria, 2003; Tangney et al., 2007; Giguère et al., 2014). However, these moral emotions have their differences. Guilt appears when someone perceives that he or she is responsible for the failure in a situation considered controllable (Tangney and Dearing, 2003). The emotion of guilt facilitates the appearance of reparative behaviors to modify the result obtained (Tangney et al., 2014). Instead, when someone experiences shame, he or she attributes this to his or her failures, and not to his or her behavior in situations where there is no perceived control (Tracy and Robins, 2007). When a person experiences shame, he or she forms negative judgments about his or her abilities, generating a desire to flee and disappear (Tangney et al., 2014). Thus, a negative evaluation of own behavior and feelings of remorse or regret may be associated with guilt, while a negative evaluation of the self and feelings of helplessness or insignificance may be associated with shame (Tangney and Dearing, 2003; Schmader and Lickel, 2006; Tracy and Robins, 2007).

Different authors (Niedenthal et al., 1994; Roese and Olson, 1997) have studied the role of counterfactual thinking in reference to guilt and shame. They argue that guilt is usually associated with upward counterfactual thoughts in the form of “If I hadn’t overtaken the car, I would not have crashed,” where the alternative picks up a specific behavior, as a cause of the unwanted result. Shame, on the other hand, emphasizes the cause in the proposed alternatives, questioning the abilities of the person and focusing on the need for a change in personal behavior to undo the result. One upward counterfactual thought associated with shame, where the person has no ability, could be “If I weren’t such a distracted person, I would not have had this crash.” Finally, it should be noted that downward counterfactual thinking is linked to feelings such as satisfaction or relief. This type of thinking can fulfill a function of mood repair, as it tends to make the person feel better (Roese and Olson, 1995; Sanna, 1998; Sanna et al., 2001).

The process of self-regulation is consistent with this approach. It is triggered when an individual considers his or her present status and compares it to a more desirable one in which behavior and emotion reflect feedback control (Carver and Scheier, 2016). In their theory of the self-regulation of behavior, Scheier and Carver (1985) contend that optimists strive to achieve a desirable outcome, which they believe themselves capable of attaining. In contrast, pessimists consider the outcomes to be unattainable and will, consequently, either desist or fail to commit to the actions required to attain the target outcome. Sundry studies have focused on analyzing how different future scenarios are constructed in the minds of optimists and pessimists (Sharot, 2011; Garrett and Sharot, 2017). It is important to consider the difference between optimism and pessimism as cognitive expectations associated with future events, and counterfactual thinking as a cognitive process involving the imagination of alternative outcomes to past events.

This leads us to the focus in this study on the conception of optimistic vs. pessimistic expectations, as proposed by Carver and Scheier (2016), which indicates that dispositional optimistic expectations generally focus on the results to be obtained. This study will consider a fundamental aspect of the human behavior factor in road safety, namely, the personality traits of the drivers and their disposition, in particular, for optimism and pessimism (del Valle and Mateos, 2008; del Valle, 2019; del Valle and Matus, 2019). Similar to the study of Giguère et al. (2014), we believe that counterfactual thoughts are decisive in the study of the self-regulatory functions of guilt and shame. The consideration that counterfactual thoughts create different scenarios or alternatives in the achievement of the proposed objectives can provide insight into the causal attribution that subjects make of the result of the current situation. Different authors (Markman et al., 1995; Roese and Olson, 1997; Nasco and Marsh, 1999) point out, after obtaining a negative result, upward counterfactual thoughts are more frequent than downward counterfactuals because the upward counterfactuals reflect what the participants could have done to obtain better results. The functionality of downward counterfactual thoughts is related to the justification of the results obtained since they imagine it could have been worse.

A determining variable that mediates between optimistic and pessimistic expectations and counterfactual thoughts is the perceived control that drivers believe they have because it primarily functions as a risk moderator (Girotto et al., 1991; Helweg-Larsen and Shepperd, 2001; Klein and Helweg-Larsen, 2002). Several studies have found that drivers who overestimate their driving skills and their perceived degree of control are more inclined to drive dangerously while maintaining a more optimistic view on the risk of having an accident (Sümer et al., 2006; Fernandes et al., 2007; Harre and Sibley, 2007). The degree of control or perceived controllability is a perception of an individual of his or her capacity, available resources, or opportunities to obtain positive results or to avoid negative ones through his or her behavior (Thompson, 1981). Similar to the study of Epstude and Roese (2008), we contend that people exaggerate the personal control they believe they have over a given situation. People who generate counterfactual thoughts tend to report factors that they believe could be manipulated, controlled, or altered in some way (Byrne, 2016). There are events in daily life that are controllable (e.g., use of the seat belt) and others that are uncontrollable (e.g., traffic jams). When people generate counterfactual thoughts, they tend to imagine how the results would have been different if they had behaved differently, albeit, to a greater extent, in situations, they believe they control (Girotto et al., 1991). These controllable situations have a greater likelihood of eliciting upward (i.e., “it could have been better”) than downward counterfactuals (Roese and Olson, 1995), while upward counterfactuals enhance retrospective control perceptions (McMullen et al., 1995; Nasco and Marsh, 1999). For example, Davis et al. (1995) have noted that the “controllability effect” appears in traffic accidents involving the loss of a loved one. Drivers that have not caused the accident focus on alternatives to their behavior rather than on the alternative actions of the driver who caused the accident (“If we had not gone by car that day, we would not have had the accident”).

In the context of driving, perceived high control can inflate estimation of the drivers of their ability, whereby both the optimism and the perceived controllability of the event maintain a close relationship (Harris, 1996; White et al., 2004). Optimistic biases of an individual are expressed in their perception of personal risk, so that, in the event of an accident, the individual tends to attribute it more to external factors (e.g., rain, a blowout), as opposed to internal factors related to driving (Harré et al., 2005). Optimistic offender drivers in their driving habits make the decision to carry out risky behaviors, even knowing that these involve high probabilities of risk (e.g., increasing speed, crossing an amber traffic light, not respecting the distance of security, or checking the mobile). McKenna (1993) indicated that people believe themselves to be less likely to suffer a road accident than others when they are driving (i.e., personal control). When they are passengers, the chances of having an accident are no different than those of other people. The illusion of control is what prompts them to attribute accident-free driving to their ability rather than the effect of external factors (Hammond and Horswill, 2002).

The last variable is related to drivers who have committed traffic violations, such as driving over the alcohol limit, etc. When these offending drivers lose points on their licenses, they must attend reeducation courses to recover some or all of the points lost. In Spain, these courses are known as “intervention, awareness, and road reeducation courses in the license points system, the Directorate General for Traffic (DGT—Dirección General de Tráfico).” Various studies have been conducted in Spain with drivers who have lost their license (Cuervo and Villanueva, 2015; Valero et al., 2017; Faílde et al., 2018; Padilla et al., 2018; Martí-Belda et al., 2019). We have conducted previous studies focusing on this group of offender drivers, and we have analyzed a type of anticipatory thinking, called pre-factual thoughts. del Valle (2019) has obtained a cognitive profile of optimistic offender drivers, indicating that they believe themselves to be more likely to achieve their desired outcomes regardless of the driving conditions, considering themselves to be more skillful drivers than their peers. In another study, del Valle and Matus (2019) reported similar findings, indicating that optimistic offender drivers consider themselves more likely to achieve their desired outcomes and do not question their skills or resources. They believe their results would be even better were they assisted by external factors. However, this group does not record better results than the other groups, so this thinking is not realistic.

There are no conclusive studies in the scientific literature that analyze alternative thoughts related to “what could have been” (i.e., counterfactual thoughts) and the emotions of guilt and shame in optimistic offender drivers. Therefore, this study aimed to analyze the extent to which the controllability of the situational context in a simulated driving task performed by optimistic offender drivers who attend reeducation courses influenced their use of internal and external factors in counterfactual thoughts and negative emotions (i.e., guilt and shame). As regards this objective, we consider the extent to which optimistic offender drivers, under conditions of induced control, are different from non-offenders drivers due to the effects of the internal and external factors on counterfactual thinking and in the experiences of guilt and shame.

First, different ANOVAs were used to analyze whether the established groups presented significant differences in the LOT and the PANAS questionnaires. Second, additional ANOVAs were carried out to analyze whether the study groups presented differences in the total number of counterfactual thoughts and to analyze the mistakes made during the simulator task. Third, a correlation analysis was applied to analyze the existence of significant relationships between errors committed and negative emotions (i.e., guilt and shame). Fourth, to analyze whether the study groups in the proposed conditions presented differences in two variables—counterfactual factor (upward and downward) and negative emotions (i.e., guilt and shame)—two multivariate repeated measures analysis of variance (MRM-ANOVA) were carried out.

The SPSS v.25 Statistics software was used to perform the analysis. A level of significantly lower than α = 0.05 was the reference established to consider the differences significant. When significant differences were present, all the analyses calculated the size effect.

We have analyzed to what extent the optimistic offender drivers, under conditions of induced control, differed from non-offender drivers in the effect of internal and external factors on counterfactual thinking and emotions (i.e., guilt and shame). It was found that optimistic offenders in both high- and low-control conditions used upward counterfactuals linked to external resources more than the other groups. When these participants generated these thoughts, they indicated that they could have obtained better results if there had been external factors involved. In relation to the downward counterfactuals, these participants had a different profile, depending on the control conditions. In a low-control condition, they were linked to a greater extent with internal factors. The belief that things could have been worse if they had not done something highlights protective functionality on an emotional level. With respect to emotions, we observed that this group recorded the lowest scores for negative emotions (i.e., guilt and shame) in both conditions and compared to the rest.

Consistent with previous studies (del Valle, 2019; del Valle and Matus, 2019), optimistic offenders overestimated their overall level of success. We believe that counterfactual thinking can have a dysfunctional implication. When drivers generate upward counterfactual thinking in explaining their failures or unwanted results, they overestimate their abilities (Petrocelli et al., 2012). Different studies have indicated how optimistic offender drivers have a greater tendency to qualify as more skilled compared with other drivers (Dogan et al., 2012; Horswill et al., 2013; Stephens and Ohtsuka, 2014; Mãirean and Havârneanu, 2018). These participants focus on external aspects of the situation (“If it hadn’t snowed, then I could have avoided the accident”) to justify their unwanted results, as noted by Girotto et al. (2007). Most of the causal attributions involved in upward counterfactuals tend to diminish the extent of the problem (i.e., if you had only had more time.) instead of considering other possibilities (i.e., inadequate knowledge or a misunderstanding of the problem) (Sherman and McConnell, 1995).

We contend that it is precisely this self-assessment that these drivers make of the abilities that lead them to ignore, or at least underestimate, the negative feedback provided by the environment (Allison et al., 1996; Petrocelli et al., 2012). The fact that these drivers ignore what happened, thinking about what “could have been,” may explain why these drivers have the lowest values in negative emotions (i.e., guilt and shame), both in low- and high-control conditions, as we will be commenting on in due course.

In relation to downward counterfactuals, the optimistic offenders in the low-control condition note how the results obtained could have been worse if they had not done something. These drivers also used this thinking to justify the mistakes made. In addition, these thoughts could have an affective function, helping people to feel better, as they reduce the potentially negative impact of considering an unattained result (Epstude and Roese, 2008; Sweeny and Vohs, 2012; Byrne, 2016). Similar to the study of Smallman and Summerville (2018), we think that one of the functions of counterfactuals is to provide reasons for poor performance, facilitating different excuses. This type of counterfactual implicitly denies the existence of any possibility of improvement, and it can reduce the motivation to change and improve. In these cases, counterfactuals can justify self-esteem of an individual while concurrently reducing the motivation for successive improvement (McCrea, 2008).

As mentioned earlier, optimistic offenders recorded the lowest values of negative emotions (i.e., guilt and shame). Tangney and Dearing (2003) have commented that emotions such as guilt depend on the negative judgment of a person of his or her action. This emotion tends to appear in situations in which a failure is perceived; there is a perception of controllability in his or her actions, and, therefore, the driver is attributed internal responsibility for it (e.g., “If I had not had that drink, the collision would have been avoided”). Some authors (Fedewa et al., 2005) have commented that guilt can encourage actions to amend the result generated; on the one hand, these drivers do not feel guilt, and, on the other, they attribute responsibility for the result to external aspects (e.g., If the pedestrian had not crossed the road, the accident would have been avoided). The fact that these drivers, in conditions of low control, demonstrate significant relationships between a higher number of errors committed and a higher level of shame leads us to suggest that it is shame, with the non-acceptance of internal limitations or the ability of the driver, which might explain that these drivers evade their responsibilities and indicate that they could have obtained better results if there had been external factors involved (i.e., upward counterfactuals). Fee and Tangney (2000) indicated that shame proneness was related to procrastination tendencies, whereas guilt proneness was not. These authors indicated that procrastination was a means of self-protection.

Optimistic non-offender drivers recorded a different profile than the group previously commented. We have found that downward counterfactual thoughts require special attention because they are the ones that recorded differences depending on the control conditions (as shown in Figure 1). McMullen and Markman (2000) have commented that if a downward counterfactual conjures a negative affect by considering the real possibility that the outcome could have been far worse, it could serve as an admonition to change the behavior of one. We have observed that in situations of higher control when these drivers thought things could have been worse had they not taken different measures (i.e., downward counterfactual), they also experienced more guilt than shame. Following the theories of counterfactual reasoning posited by Mandel (2005), subjects may focus not only on how the events could have been avoided but also on how they were caused. The fact that these subjects report greater guilt than shame may explain the regret they may experience when, despite controlling the situation, they perceive their failures and responsibility (Tangney and Dearing, 2003; Mandel and Dhami, 2005).

Under conditions of low control, we can see how drivers recorded higher levels of shame and guilt, experienced simultaneously (as shown in Figure 1). Similar to the study of Tangney et al. (2007), we contend that guilt becomes maladaptive when merged with shame. Alicke et al. (2008) have reported the assertion among counterfactual reasoning theorists that the emotional response of an observer to an event is amplified when accompanied by the knowledge that an outcome could have been averted. Under situations of low control, in which bad results are also obtained, it is not functional to think that the results would have been worse if the driver had not carried out actions to avoid negative results (i.e., downward counterfactuals), and these could have been avoided. Furthermore, under this condition, downward counterfactual thoughts did not show the functionality indicated by Sanna (1998), Sanna et al. (2001). They gave this thought a restorative mood function, which could also be used to feel better, or as a feeling of relief.

In the group of pessimistic non-offenders, upward counterfactual thoughts varied, depending on the conditions (as shown in Figure 1). When these drivers generated this type of thoughts, they attributed the result to their abilities under conditions of low controllability (i.e., an internal, stable, and uncontrollable factor). Under this condition, they tended to feel shame. Similar to the study of Tracy and Robins (2007), we understand that shame appears when the responsibility for the result is focused on the ability of a person, not on a specific behavior “If I were a more skilled driver, I would not have had this crash.” This alternative clearly emphasizes an internal factor, which underpins a negative evaluation of the self-image of the driver. This emotion may evoke more reticent behavior, which serves to distance the driver from the situation (Tangney and Dearing, 2003). Various authors (Sanna, 1998; Nasco and Marsh, 1999; Branscombe et al., 2003) have reported that the preparative function of the upward counterfactuals is observed among individuals whose self-efficacy is high, but not low. Markman and Miller (2006) have found that people with moderate depression did not obtain benefits when considering how they could have prevented the negative results. This leads us to suggest that upward counterfactuals can become dysfunctional since they would foster the emotion of shame, as shown in Figure 2.

As we have been commenting, the functionality of counterfactual thoughts and negative emotions appears under situations with unfavorable outcomes, which elicit a more causal than the favorable type of reasoning (Boninger et al., 1994). The negative emotions in optimistic offenders do not vary under the different control conditions. This is not the case in the other two groups. Similar to the study of Murrar et al. (2019), we believe that shame can negatively influence behavior, while guilt can do so positively. The drivers show the influence of shame under the condition of low control on both optimistic and pessimistic non-offenders. In turn, the role of guilt appears in the high-control condition of these groups. The level of guilt is heightened when people imagine how a change in their actions could have produced a different result; for example, “I would not have been fined if I had not speeded.” In contrast, the level of shame is heightened when they imagine how a change in their personality could have produced a different result; for example, “I would not have had the accident if I were not so clumsy” (Niedenthal et al., 1994). As it is understandably difficult to change personality traits, shame is, therefore, described as a maladaptive emotion since it promotes withdrawal and hinders the reparation process, while guilt is seen as an adaptive emotion, as it motivates repair (Tangney et al., 2014; Carpenter et al., 2016).

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by University of Salamanca. The patients/participants provided their written informed consent to participate in this study.

The author confirms being the sole contributor of this work and has approved it for publication.

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.