Self-reported driving behaviors as a function of trait anxiety

https://doi.org/10.1016/j.aap.2008.11.004Get rights and content

Abstract

This study examined self-reported driving behaviors in 120 (Israeli) male drivers as a function of trait anxiety (TA). TA was assessed through the TA scale of the State-Trait Anxiety Inventory. For the analysis of driving behaviors, the present study used the Driver Behavior Questionnaire (DBQ) and adopted previous distinctions between four classes of behaviors within the DBQ: errors, lapses, ordinary violations and aggressive violations. Regression analyses revealed that level of TA had a significant direct positive effect on all dependent variables, suggesting riskier driving behaviors among high-anxious individuals. Significant logarithmic effects for all measures indicate that these aberrant driving behaviors increase more at increasing LTA-, than at increasing HTA-values. Consistent with the general adverse effects of anxiety on performance effectiveness, the present findings as well, are interpreted in the framework of theories which suggest that worries occupy the capacities of working memory, at the expense of the task to be performed. The positive relation between aggressive violations and TA is sought to reflect low levels of emotional adjustment among high-anxious individuals.

Introduction

Among factors which affect driving safety, the influence of some personality traits has been the subject of extensive research. Perhaps the majority of studies on the subject have investigated how driving is influenced either by sensation seeking (the willingness to take risks; Zuckerman, 1994) or by aggression. Thus, in many studies, participants who scored higher as sensation-seekers, particularly young drivers (Arnett et al., 1997, Ulleberg, 2002) reported more risky behaviors than those who scored lower on this count (Arnett et al., 1997, Iversen and Rundmo, 2002, Jonah, 1997, Jonah et al., 2001, Rosenbloom, 2003). Some of these studies and several others investigated the influence of aggression on driving behavior, reporting riskier driving among more aggressive individuals (Arnett et al., 1997, Chliaoutakis et al., 2002, Deffenbacher et al., 2003, Schreer, 2002, Sommer and Baumeister, 2002, Stucke, 2001, Ulleberg, 2002, Underwood et al., 1999).

There are relatively few studies that have investigated the influence of anxiety on driving behavior. The present study sought to examine how risky driving, as reflected by self-reported driving behavior, is affected by trait anxiety (TA).

Among the relatively few studies that have investigated the influence of anxiety on driving behaviors, using a cluster analysis of personality measures, Ulleberg (2002) assessed the effects of several of personality traits, including anxiety. This analysis yielded six subgroups of young drivers, which differed on several measures including attitudes towards traffic safety, risk perception, estimation of one's driving skills, and involvement in accidents. Two of the clusters were characterized as high-risk groups. One of these groups (comprised mostly of men) was characterized by low levels of altruism and anxiety, and high levels of sensation seeking, irresponsibility, and driving-related aggression. A second high-risk group (made up mostly of women) reported high sensation seeking, aggression, anxiety, and driving-related anger, a profile that the authors identified as one that indicates low levels of emotional adjustment. Thus, while the first group was characterized by low levels of anxiety, the second group was characterized by high levels of anxiety. Consistent with this pattern is a suggestion made by Oltedal and Rundmo (2006), who measured the effects of personality on risky driving behavior and accident involvement among Norwegian adolescents. Although generally anxiety was related to excitement seeking and to risky driving behavior, contrary to expectations, TA was only weakly related to risky driving and was unrelated to accident involvement. The authors suggested a non-linear relationship between anxiety and driving behavior. Specifically, they proposed that while an average level of anxiety does not affect driving behavior in any important way, both very low and very high levels of anxiety negatively affect driving safety, the former by promoting calmness to the extent of overconfidence and the latter by inducing excessive tension. In other words, the relationship between TA and driver behavior may be U-shaped, with higher levels of negative driver behaviors at both low and high levels of TA, due to too little concern about safety among low-anxious drivers, and due to the cognitive interference created by the anxiety among high anxious drivers.

One of the factors that might be involved in the process of reducing risky driving among anxious individuals – strongly related to confidence – is risk perception, defined as the subjective experience of risk in potentially hazardous traffic situations (Elander et al., 1993). Unsurprisingly, anxiety is associated with an increment of perceived risk (Butler and Mathews, 1987). Risk perception is indeed associated with risky behavior (Horwarth, 1988, Brown and Groeger, 1988). The general idea is that drivers who perceive a low risk of accident will drive recklessly, while those who perceive a high risk in the same situation will drive cautiously (Harre, 2000).

One of the factors that may be involved in the process of increasing risky driving in anxious individuals is consistent with the adverse effects of anxiety on performance effectiveness (Eysenck, 1996). Specifically, theorists usually have attributed such decrements to some form of attention deficit, as proposed in cognitive interference theory (Sarason, 1988) and processing efficiency theory (Eysenck and Calvo, 1992). Despite the differences between them, both theories generally propose that in anxious people attention is diverted from the task at hand by worries which preoccupy processing resources. The latter possibility received some support from a recent study (Wilson et al., 2006), which measured the effects of anxiety on simulated rally driving. Increased anxiety had a negative effect on processing efficiency as indexed by a self-report, by pupillary responses and by variability-of-gaze data. While both low- and high-trait-anxious drivers performed worse under threatening conditions, the performance of high-anxious drivers was affected to a greater extent than that of low-anxious drivers by the anxiety manipulation. The position that anxiety negatively affects safe driving is also consistent with another, even more recent, study (Taylor et al., 2007), which investigated the driving skills of fearful drivers and control (female) drivers. Fearful drivers made more errors on the driving assessment (an on-road practical evaluation with a professional instructor) than controls did. The pattern of errors was identical for both groups, suggesting that anxiety is associated with the number rather than the type of driving errors made.

Thus, on the one hand it might be expected that high-anxious drivers will engage in more risky driving behaviors than will low-anxious drivers; yet there are also grounds to expect the opposite pattern.

In this study the assessment of general driving behavior was obtained using the Driver Behavior Questionnaire (DBQ; Reason et al., 1990). For further analysis of driving behaviors, previous distinctions were adopted, between three classes of behaviors within the DBQ: errors, lapses, and violations (Reason et al., 1990, Parker et al., 1995) and a further sub-categorization of violations into two distinct types, highway-code, or ordinary violations, and aggressive violations (Lawton et al., 1997, Gras et al., 2006).

Originally, Reason et al. (1990) identified three factors within the DBQ: violations, errors and lapses. Violations were defined as deliberate departures from behaviors believed to represent safe driving practices (e.g., disregarding the speed limit). Errors were defined as failures of observation that may be hazardous to others (e.g., failing to check one's rear-view mirror before pulling out or changing lanes). Errors also included planned actions that fail to accomplish their intended outcomes (e.g., braking too quickly on a slippery road, or steering the wrong way in a skidding vehicle). Lapses were defined as absent-minded behaviors which usually do not pose any threat to road users (e.g., attempting to drive away from traffic lights in third gear). These three basic components of the DBQ were also found by other studies (Parker et al., 1995). Lawton et al. (1997) differentiated between aggressive violations (e.g., sounding your horn to indicate your annoyance) directed at individual road users and ordinary (highway-code) violations. This distinction was supported by Gras et al. (2006) and was partially supported by Mesken et al. (2002), whose data distinguished between speeding violations and interpersonal violations. Notably, although in the literature violations have usually been associated with crash involvement, among elderly drivers high error and lapse scores have been reported to predict accident involvement (Parker et al., 2000).

As shown in this Introduction, for driving behavior as a function of the TA variable, theory and data provide a basis for a non-linear–U-shaped relationship between TA and driver behavior, with higher levels of negative driver behaviors among both low and high anxious individuals (post hoc hypothesis). No specific predictions were derived regarding the DBQ subscales.

Section snippets

Respondents

One hundred and twenty male respondents (mean age = 32.21; S.D. = 7.02; range = 22–50) volunteered to participate in the study. All of them had a full driving license (mean = 13.2; S.D. = 6.6). Approximately half of the respondents were recruited through one company and half through another (both hi-tech companies).

Measures

Data were collected using two questionnaires. TA was assessed through the TA scale of the State-Trait Anxiety Inventory (Spielberger et al., 1970). The TA scale of the STAI consists of 20

Results

The mean TA score was 33.68 (S.D. = 6.89; range = 20–49). Pearson's correlations between age and TA scores (r = −0.06) and between driver's license seniority and TA scores (r = −0.17) were both low and not significant. Likewise, the Pearson correlation between age (one missing value) and DBQ scores (r = −0.177) was low and not significant; and although the Pearson correlation between driver's license seniority and DBQ scores was significant (p < .05), it was low as well (r = −0.182). The correlation between

Discussion

For all five dependent variables including DBQ and the four subscales (errors, lapses, ordinary violations and aggressive violations), level of TA had a significant direct positive effect. For all of those measures the logarithmic effect was also significant. The linear effects are consistent with studies that have found general adverse effects of anxiety on performance effectiveness (Eysenck, 1996), as well as adverse effects of anxiety on driving performance in particular (e.g., Wilson et

Acknowledgments

The author is grateful to Elizabeth Yahnin and Julia Avrekh for collecting and coding the data, and to two anonymous reviewers for their helpful comments and suggestions on earlier drafts of this manuscript.

References (42)

  • D. Parker et al.

    Behavioural characteristics and involvement in different types of traffic accident

    Accident Analysis & Prevention

    (1995)
  • D. Parker et al.

    Elderly drivers and their accidents: the Aging Driver Questionnaire

    Accident Analysis & Prevention

    (2000)
  • T. Rosenbloom

    Sensation seeking and risk taking in mortality salience

    Personality & Individual Differences

    (2003)
  • J.E. Taylor et al.

    Driving fear and driving skills: comparison between fearful and control samples using standardised on-road assessment

    Behavior Research and Therapy

    (2007)
  • G. Underwood et al.

    Anger while driving

    Transportation Research Part F: Traffic Psychology and Behavior

    (1999)
  • D. Walton

    Examining the self-enhancement bias: professional truck drivers perceptions of speed, safety, skill and consideration

    Transportation Research Part F: Traffic Psychology and Behavior

    (1999)
  • J.R. Adams-Guppy et al.

    Speeding in relation to perceptions of risk, utility and driving style by British company car drivers

    Ergonomics

    (1995)
  • I.D. Brown et al.

    Risk perception and decision taking during the transition between novice and experienced driver status

    Ergonomics

    (1988)
  • G. Butler et al.

    Anticipatory anxiety and risk perception

    Cognitive Therapy and Research

    (1987)
  • G.S Claridge

    Origins of Mental Illness

    (1985)
  • D.E. Craig

    The relationship between depression and anxiety, and the personality traits of dependency and self-criticism

    Dissertation Abstracts International: Section B: The Sciences and Engineering

    (1996)
  • Cited by (91)

    View all citing articles on Scopus
    View full text