Psychology, 2010, 1: 35-44
doi:10.4236/psych.2010.11006 Published Online April 2010 (http://www.SciRP.org/journal/psych)
35
Aggression on the Road as a Function of Stress,
Coping Strategies and Driver Style
Lipaz Shamoa-Nir, Meni Koslowsky
Department of Psychology, Bar-Ilan University, Ramat Gan, Israel.
Received January 12th, 2010; revised January 19th, 2010; accepted January 20th, 2010.
ABSTRACT
According to Lazaru s and Folkman’s [1] tran sactiona l cognitive model, peop le differ in their sensitivity a nd vuln erabil-
ity to stressful events. Using questionnaire and observational techniques, the model was tested as a possible explana-
tion for aggressive driving behavior. Responses from 226 drivers who were also observed driving their cars provided
evidence for a lin k between stress and aggressive driving as well as between problem-solving strategy as a coping de-
vice in stressful situations and ho stile behaviors. In addition, analysis showed that, in general, the more years of driving
experience a driver has, the more likely he/she is to respond with instrumental rather than hostile aggression. Besides
support for the theoretical model, some of the practical applications as they related to highway safety and the preven-
tion of traffic accidents were presented.
Keywords: Dri ving Stress, Aggressive Driving, Problem-Solving Strategy, Hostile Strategy, Instrumental Stratgey
1. Introduction
Road accidents and traffic offences resulting from ag-
gressive driving have been a subject of interest to many
researchers over the years with several studies attesting
to an increase in negative outcomes. Examples range
from irritability, anger, violent reactions [2,3] and even
drivers shooting at each other during an argument such as
who saw the specific parking space first [4]. A common
explanation for these negative behaviors uses the frustra-
tion-aggression model whereby a driver who has been
blocked from getting to his/her destination expresses
frustration which may lead to some overt expression such
as harming/hurting another driver. Yet, in many situa-
tions where aggression is manifested, the so-called cause
of the frustration is not readily apparent. The present
study applies an alternative approach, Lazarus and
Folkman’s [1] transactional cognitive model, for ex-
plaining drivers’ actual reactions on the road.
According to the usual formulation, where frustration
is followed by an aggressive act [5,6] no real distinction
is made among the different types of aggression. How-
ever, Feshbach’s [7] conceptualization which distin-
guished between hostile and instrumental acts [8] seems
quite appropriate for the driving situation. Although both
types of aggression are seen as an attempt to harm an-
other person, the aim of instrumental aggression is to
gain something such as money, social status or territory,
whereas hostile aggression is mainly aimed at causing
hurt or pain. In his study on aggressive driving behavior,
Shinar [9] defined instrumental aggression as actions
taken by the driver that will aid his/her progress in driv-
ing, or help in removing or overtaking an obstacle on the
road. while hostile aggression on the road serves no pur-
pose other than harming another.
According to Shinar [10], the differentiation is not
unambiguous and many expressions of anger on the road
can be defined as either instrumental or hostile or both.
Although overlap is expected between the concepts, this
distinction can explain why there were fewer aggressive
behaviors such as driving through a red light or honking
at a driver blocking progress when the light is green
(which is sometimes considered hostile) among older
drivers as well as the greater number of such behaviors
reported among men than women [9].
While hostile aggression gives drivers a feeling of sat-
isfaction about the present difficulties in which they find
themselves, it doesn’t really solve the problem at hand.
At best, these actions help channel drivers’ anger while
producing harm to the frustrating party. Overall, the frus-
tration-aggression model, accounts for the result of the
drivers’ behavior but not for the process that leads the
driver from his/her feelings of frustration to the specific
behavioral reaction.

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 36
1.1 Road Rage and Aggressive Driving
Recently, a new term, road rage, has been introduced into
the discussion on aggressive driving. Although many
people view these terms as similar, in fact, it is likely that
the terms have specific connotations [9]. The American
National Safety Council has tried to differentiate between
them by defining aggressive driving as “movement or
activity using a vehicle that endangers or will endanger
people or property,” which is a traffic violation [11]
whereas road rage is not necessarily a traffic offence and
is seen as “an attack initiated by the driver of the car or a
passenger, on a driver of another car or its passenger,
using a car or other dangerous vehicle, this anger being
the result of an incident or event on the road during driv-
ing” [11]. Examples are tailgating, deliberately blocking
progress, honking, and even verbally or physically at-
tacking a driver [9]. The present study applies the dis-
tinction between aggressive acts for explaining these
behaviors.
1.2 Commuting Stress
Many investigators agree that driving is a complex activ-
ity, often accompanied by stress [12]. The relevant stim-
uli and responses associated with the commuting process
are a relatively new concern for stress researchers and
incorporate various environmental, personal, and situ-
ational sources [13]. Among the effects that have been
investigated here are physiological [14], psychological
[14] or organizational outcomes [15]. The commuting
stress model postulated by Koslowsky et al. [13] com-
prises several stages relating to stress-causing factors
such as distance and time, how subjective stress is con-
ceived, and how the potential negative outcomes relate to
each other. A popular type of research issue has been to
identify moderators of the stress-strain relationship. For
example, there is evidence that there are different levels
of stress associated with mode of travel. Findings by
Koslowsky and Krausz [16] showed that stress symptoms
were greater among nurses who drove their cars to work,
compared to those who commuted by public transport.
1.3 Driving Behavior Styles
In studying drivers’ stress, Gulian, Matthews, Glendon,
Davis & Debney [17] argued that drivers’ stress-related
behavior depends on the driver’s appraisal of the situa-
tion, in that driving skills depend on the individual’s
ability to cope with stress. They identified five distinct
and independent categories of driving under stress and
assessed them by using the Driving Behavior Inventory
(DBI). Among the styles relevant here are “dislike of
driving” and “aggression.” Questions on “dislike of
driving” deal with anxiety, dissatisfaction and lack of
confidence, especially under difficult driving conditions.
These mainly relate to emotional stress symptoms such
as tension, and depressed mood states as a result of driv-
ing [18]. Questions on aggressive driving style deal with
feelings of anger, frustration, lack of patience and a
negative perception of other drivers who are sometimes
seen as hostile and threatening. “Aggression” questions
deal with annoyance while driving, lack of patience and
aggressive actions, especially when progress is blocked
by other drivers [19]. Research dealing with the associa-
tion between driving styles and cognitive measures of
coping found that drivers’ stress measures and resulting
behavior can be characterized by the following: drivers
who scored high on “dislike of driving” tended to cope
with stress while driving by using emotional coping
strategies (for instance self-criticism) which increased
feelings of apprehension about traffic. Drivers who
scored high on “aggression” used direct confrontation
strategies [19] which included tailgating and frequent
overtaking [18].
In addition, drivers who scored high on “aggression”
reported that they made more mistakes while driving and
committed more traffic violations such as speeding [20].
“Dislike of driving” and “aggression” were found to be
linked to processes such as cognitive assessments of cir-
cumstances involving stress and ways of coping with
them [20], including emotional reactions and reactions to
stress.
1.4 Coping with Stress
People differ in their sensitivity and reactions to stressful
situations [21]. When drivers are stressed, their aggres-
sive behavior may be easier to understand using Lazarus
and Folkman’s [1] cognitive model which describes
coping styles in stressful situations. The model suggested
by these researchers has been one of the most influential
formulations in explaining both theory and empirical
findings on coping strategies in stressful situations
[22-25]. Cognitive evaluation starts with the individual
appraising the dangers of the situation. Next, the indi-
vidual analyzes ways to cope with the situation [1] so as
to regulate emotions which may lead to modifying the
specific stress-strain link.
An individual who experiences a stressful situation can
react in one of two ways: emotion-focused coping de-
fined as decreasing emotional stress including strategies
such as abstention, blaming others, keeping distance,
selective attention, and finding something valuable in
negative events. On the other hand, problem-focused
coping includes problem-solving strategies and dealing
effectively with stress stimuli. Examples include focus-
ing on the overall problem, attempting to define the
problem, suggesting alternative solutions, considering the
alternatives, choosing one of them, and taking action.
An overlooked but interesting area is the link between
styles of coping with stress, and attitudes towards driving
and related emotions [12]. Differences in coping styles

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 37
among drivers are reflected in different attitudes towards
driving [17,18,26]. Generally, in these studies, data were
collected from questionnaires completed by participants
but aggressive reactions of drivers were not tested in real
time, i.e., on the road. In addition, the instruments for
comparing coping styles while driving were limited to
developing measures and scales to test examine stress
and copings, without examining the process of driving
while under stress.
Based on the studies in the area using stress, driving
style, and coping processes, the following specific hy-
potheses concerning aggression on the road were formu-
lated:
Hypothesis 1: Drivers who use a problem solving ap-
proach to stress will experience less perceived stress. No
relationship between emotional coping style and per-
ceived stress is expected.
Hypothesis 2: Perceived stress, coping style, individu-
ally and as an interaction term, predict who is likely to be
aggressive on the road.
Hypothesis 3: Drivers who use instrumental aggression
will manifest more stress and use more of a prob-
lem-oriented style of coping than those who use hostile
aggression while driving.
Hypothesis 4: There will be a link between perceived
stress and driving style such that perceived stress of ag-
gressive style drivers will be greater than the perceived
stress of dislike driving style drivers.
2. Method
2.1 Sample
Participants included 226 drivers (67% women) affiliated
with a university in central Israel. Mean age for the group
was 29.0 (SD = 6.73), ranging from 19-74 with an aver-
age number of years of education, 14.8 (SD = 2.92),
ranging from 8-30 years. About 49% were students, 43%
salaried employees, 4% self-employed, 3% unemployed
and less than 1% were soldiers or pensioners.
The average number of years driving was 10.27 (SD
= 8.73), ranging from 1-59 years with about 89%
saying they drove their cars almost every day The av-
erage number of kilometers driven in the middle of the
week was 186.91 (SD = 220.72) and the range was
between 1-2000 kilometers. Nearly 49% of the partici-
pants had been involved in road accidents. Of those
involved in accidents, 75% were young drivers (30 or
below). Among those who had committed a traffic
violation, about 31% had at least one or more tickets
for speeding.
During the period of observation, 31% of the drivers
displayed one aggressive behavior including 7% who
sounded a “short honk”; 1% a “long honk”; 3% “two
consecutive honks”; 9% who had “cut in” on other driv-
ers; and 12% who tailgaited.
2.2 Instruments
The State-Trait Anxiety Inventory. The Speilberger
[27] State-Trait Anxiety Inventory, as translated into
Hebrew by Teichman and Mellik [28], was used here.
Participants are asked to rank the strength of their present
feelings on a scale from 1-not at all to 4-very much. For
the present analysis, the relevant items were those that
focused on an emotional description related to stress at-
tributes that a person feels “at a given moment”, such as
serenity, safety, anger etc. A person’s anxiety level is
determined by combining the individual responses with a
higher score indicating a higher state of anxiety.
A Checklist for Coping Styles. The questionnaire was
translated into Hebrew [29] from the original article by
Folkman and Lazarus [22] The Ways of Coping Checklist.
The questionnaire includes 43 items describing various
strategies people use in order to cope with stressful situa-
tions. The participant is asked to what degree he/she uses
each strategy when facing stressful situations. A four
factor solution for coping styles, similar to Lazarus and
Folkman, was obtained: coping focused on the problem
(12 items), coping focused on emotion (12 items),
searching for social support (8 items), and denial (5
items). Cronbach’s alpha reliability on each of the 4 fac-
tors was found to be higher than 0.74. Four factor scores
were compiled with a high score indicating that this par-
ticular strategy was used often.
Driving Behavior Inventory (DBI). The items in the
DBI [17] were translated into Hebrew. The first part of
the original questionnaire related to biographical ques-
tions such as driving experience and driving habits. The
second part consisting of 37 general stress statements
related to being on the road and reactions pertaining to
the driving experience. Gulian et al. [12] found that these
statements reflected five dimensions of stress while
driving, expressing the participant’s beliefs and reactions.
Example of items and the relevant dimension include the
following: “I overtake other cars whenever I get the
chance.”(Expression of aggression);”I am aware of dif-
ficulties on the road” (expression of alertness);”I am irri-
tated when I overtake another car”(expression of irrita-
tion when overtaking); “I feel satisfaction when overtak-
ing another car” (expression of tension when overtaking);
“Driving usually makes me frustrated” (expression of
aversion to driving-dislike driving style); “I am usually
patient when facing heavy traffic” (Expression of general
driver stress).
On the original DBI questionnaire, participants had to
mark gradations on a scale (100 mm long) showing to
what degree they agreed with the above expressions.
Matthews et al. [30] recoded the items and used the fol-
lowing scale: 1) “doesn’t describe how I feel”; 2) “de-
scribes me to a certain extent”; 3) “describes me well”; 4.
“describes me very well”. In the present study, this

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 38
scheme was used. A score was calculated for each par-
ticipant on each dimension.
The questionnaire used the back-translation procedure
discussed by Brislin [31]. Thus, an individual fluent in
both languages translated the items from English into
Hebrew, and then another translator fluent in both lan-
guages translated the items back into English. The two
translations were quite compatible and only in a few
cases was there a need to adjust a word or phrase.
Aggression Style. Based on the distinction in the lit-
erature between hostile and instrumental aggression
[7,32], two additional measures were compiled, the first
focusing on aggressive instrumental driving, which in-
cluded the following behaviors: a short honk or pushing
in front of the next driver; and a second measure for ag-
gressive hostile driving, which included the following
behaviors: a long honk, two continuous honks and tail-
gating. An individual was assigned either a value of l
(hostile aggression), 2 (instrumental aggression) or 3 (no
aggression).
2.3 Procedure
Before beginning the study, we met the parking lot man-
ager and explained to him the aims of the study and the
method to be used for gathering the data. We decided
which days driving behavior would be observed in the
parking lot and the cashiers at the entrance would dis-
tribute a questionnaire to each participant as he/she en-
tered the lot after paying the entrance fee. Every driver
was offered the questionnaire in an envelope and if any-
one asked any questions, they would be told the follow-
ing: “Read the explanation provided”. The cashiers were
also told not to force drivers to accept an envelope and to
show respect for anyone who refused to participate in the
study.
Gathering Data The questionnaires were distributed
over four days. The envelopes contained two versions of
the questionnaire: a long one with questions relating to
perception of stress, coping styles and driving styles. The
shorter version included questions relating to how stress
is perceived. The cashiers handed out the two different
questionnaires randomly.
The drivers were asked to put the completed envelopes
in a box next to the cashier. The questionnaires were
handed out to 800 drivers, of which, 237 questionnaires
were returned, a 30% response rate; 11 questionnaires
were disqualified because there was no record of those
drivers being observed. The cashiers reported that 20
drivers refused to accept envelopes. Of those who ac-
cepted the envelopes, 79 (35%) filled in the question-
naires on the spot and handed them back to the observer
or cashier at the parking lot. The rest of the question-
naires 158 (65%) were handed in and put in the box next
to the cashier or left at the psychology department.
Gathering Information from Observation The ob-
servations were done at times when the traffic was heavy
at the entrance to the parking lot and the person observ-
ing did so from the entrance to the lot without being seen
by the drivers. The observer wrote down the three middle
digits of the license plate (there was a double recording
for 37 cars so the information from the observation was
correlated with the questionnaires by age and gender
variables); the approximate status of the cars (old or new);
whether the driver was alone or with passengers; the
driver’s gender; the driver’s approximate age (seemed to
be above 30 or less than 30), and the aggressive driving
behavior used such as a short honk, a long honk, two
continuous honks, tailgating, light flashing, overtaking
and cutting in front of someone. As previous observa-
tions had indicated that the main entrance was busier
than other areas of the parking lot, the observer was sta-
tioned there. When the driver bought a parking ticket,
he/she received an envelope containing the questionnaire.
The envelope also contained particulars about the re-
searchers. The drivers were asked to complete the ques-
tionnaire no later than a half hour after entering the lot
and to leave it either with the cashier or at the psychol-
ogy department. As an incentive, all those who filled in
the questionnaire would be able to participate in a lottery
where six drivers could win free parking for one semester.
3. Results
3.1 Measures
As shown in Table 1, the reliability measures were sat-
isfactory for all scales. In addition, drivers were also di-
vided into an aggressive group, a participant who mani-
fested any kind of aggressive driving (without differenti-
ating between hostile or instrumental driving) and those
who didn’t.
The analyses below follow the order of the study hy-
potheses. Hypothesis 1 tested the association between
drivers’ stress and stress-coping styles. A significant
correlation was found between the problem-oriented
style of coping and levels of perceived stress, r = –.26,
Table 1. Means, standard deviations, and reliabilities for
scales
Measures M SD Cronbach’s α
Stress 1.68 54. 92.
Problem-oriented coping3.01 .45 76.
Emotion-oriented coping2.25 57. .81
Driving style (DBI):
Aggressive style 1.90 59. 75.
Dislike of driving style 2.46 44. 55.
Note. For stress & coping strategies n = 225-226
For driving style (DBI): n = 68-69

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 39
p < .01. This correlation was negative indicating that the
higher the participants’ score in problem-oriented coping
was, the less stress they felt. The correlation between
emotional coping and perceived stress was not found to
be significant.
In order to compare aggressive drivers to non-aggressive
ones for the three measures mentioned above, a multi-
variate analysis of variance (MANOVA) was used. A
significant difference was found between the two groups
of drivers Wilks’  = .915 (F (3,222) = 6.92; p < .001;
eta²-.08). The findings for the means and standard devia-
tions are reported in Table 2. The only significant dif-
ference between the two groups of drivers was in their
stress perceptions with drivers who displayed aggressive
behavior showing greater stress perceptions than those
who didn’t (M = 1.89, SD = .56 and M = 1.58, SD = .51,
respectively).
In Hypothesis 2 we argued that perceived stress, the
various styles of coping and their interaction contributed
to explaining the variance in aggressive driving. A logis-
tic regression analysis was conducted, suitable for situa-
tions in which the dependent variable was dichotomous.
The logistic regression analysis was done in four stages.
The first stage included personal traits (gender and age)
and those pertaining to driving (driving experience, in-
volvement in road accidents). In the second stage, the
level of perceived stress of the drivers was included, in
the third stage, the two measures of coping with stress
Table 2. Means (SD’s) for stress and coping strategies
comparisons by aggressive behavior
Aggressive behavior
Yes No
Measures M SDM SD F(1,224) eta²
Stress 1.89.561.58 .51 16.44***.06
Problem-oriented
Coping 3.02.353.01 .48 .07 --
Emotion-oriented
Coping 2.18.562.28 .57 1.45 --
*** p < .001
Note. n = 225-226; Yes = was aggressive; No = wasn’t aggressive
were used, problem coping style and emotion coping
style. Finally in the fourth and last stage, interaction
among measures was used.
As can be seen in Table 3, the first two stages ex-
plained 16% of the aggressive driving variance. Of the
variables in stage 1, only gender was significant. In the
second stage, stress explained an additional 5% to the
variance, F = 13.90, p < .01. In Table 4, the means for
the different measures are analyzed by aggression type.
Those drivers who were defined as aggressive (presented
aggressive behavior while driving) perceived more stress.
As the interaction term was not significant, the hypothe-
sis was only partially confirmed.
Table 3. A logistic regression analysis for aggressive/non-aggressive drivers
Measures B S.E. Wald Exp(B) R²
First Step Involvement in driving accidents .371 .297 1.565 1.087 .11**
experience in driving .018 .049 .139 .805
Age .001 .040 .001 1.132
Gender 1.430 .321 19.821*** 1.142
Second Step Involvement in driving accidents .371 .306 1.476 1.450 .16***
experience in driving –.003 .054 .003 .997
Age .017 .045 .141 1.017
Gender 1.480 .337 19.332*** 4.393
Stress –1.102 .300 13.456*** .332
Third Step Involvement in driving accidents .356 .307 1.345 1.428 17.
experience in driving .010 .054 .032 1.010
Age .008 .045 .031 1.008
Gender 1.511 .345 19.222*** 4.530
Stress –1.222 .316 14.987*** 295.
Problem-oriented Coping –.481 .373 1.660 618.
Emotion-oriented Coping .227 .291 .612 1.255
** p < .01, *** p < .001
Note. 1. n = 225-226

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 40
Table 4. Means (SD’s) for stress and coping strategies by types of aggressive behavior
Aggressive behavior
Hostile
Aggression
Instrumental
Aggression
Measures M SD M SD F(1,66) eta²
Stress 2.03 .52 1.71 .57 5.90* .08
Problem-oriented Coping 2.81 .21 3.31 .27 63.97*** .49
Emotion-oriented Coping 2.21 .51 2.15 .63 .16 --
** p < .01, *** p < .001
Note. n = 69
In Hypothesis 3, we compared drivers who displayed
instrumental aggression to those who displayed hostile
aggression. For the three measures mentioned before, a
multivariate analysis of variance (MANOVA) was con-
ducted and a significant difference was found between
the group of drivers who displayed instrumental aggres-
sion and the group of drivers who displayed hostile ag-
gression. Wilks’  = .493 (F (3.64) = 21.94; p < .001.
eta² = .50). The means and standard deviation of the three
measures of the two groups and the results of the vari-
ance analyses were done separately for each of the
measures as can be seen in Table 4.
As we can see from the table, drivers who displayed
instrumental aggression felt more stress than those who
manifested hostile aggression. In addition, the prob-
lem-oriented coping style was greater among those drivers
who manifested instrumental aggression than those rivers
who displayed hostile aggression. A logistic regression
analysis (see Table 5) was conducted in order to see to
what degree the perception of stress and coping style
variables contributed to variance in aggressive styles of
driving. The analysis included three stages. In the first
stage, gender, age, driving experience and involvement in
road accidents was entered. In the second stage, the vari-
able expressing the degree of stress the drivers experi-
enced during the study was introduced. In the third stage,
the two measures of coping with stress (problem-oriented
coping style and emotional coping style) were introduced.
In the first stage, 8% of the variance in differences in
styles of aggression manifested by drivers was explained
with the only significant beta contribution coming from
driving experience. An ANOVA here showed that there
were significantly more drivers who manifested instru-
mental aggression (M = 11.26, SD = 9.01) than those
who manifested hostile aggression (M = 7.89, SD = 6.57),
F (1.65) = 101.11; p < .001; eta2 = .60).
Table 5. A logistic regression analysis for aggressive behavior (hostile, instrumental)
Measures B S.E. Wald Exp(B) R²
First Step Involvement in driving accidents .083 .356 .055 1.087 .08*
experience in driving –.217 .118 3.387* .805
Age .124 .087 2.014 1.132
Gender .132 .535 .061 1.142
Second Step Involvement in driving accidents .069 .369 .035 1.072 .13*
experience in driving –.175 .117 2.237 .839
Age .091 .086 1.107 1.095
Gender .076 .548 .019 1.078
Stress .946 .521 3.293* 2.576
Third Step Involvement in driving accidents .224 .589 .144 1.251 .57***
experience in driving –.100 .184 .294 .905
Age –.059 .137 .182 .943
Gender 1.391 1.103 1.591 4.019
Stress .171 .785 .047 1.186
Problem-oriented Coping –13.788 4.101 11.306*** .000
Emotion-oriented Coping –1.532 1.004 2.328 .216
* p < .05, *** p < .001
Note. 1. n = 69

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 41
In the second stage, level of stress added an additional
5% to explained variance. Interestingly, an ANOVA (see
Table 6) indicated that the drivers who manifested hos-
tile aggression manifested significantly greater levels of
stress than those drivers who manifested instrumental
aggression.
In the third stage, where coping styles were included,
an additional 44% of variance was explained, all of
which can be attributed to the problem-oriented coping
style (B = –13.788, p < .001) a careful examination of
this relationship (Table 6) shows us that drivers who
scored high in this coping style were inclined to be in-
strumentally aggressive. In stage 4, no additional signifi-
cant variance was explained. In total, 57% (p < .001) of
variance was explained by the logistic regression.
For Hypothesis 4, we examined whether there would
be differences in drivers’ stress depending on driving
style such that drivers displaying an aggressive style
would feel more stress than those drivers who dislike
driving. A Pearson’s correlation analysis showed that
there was a significant correlation between aggressive
driving style and feeling of stress, r = .38; p < .01. The
more aggressive the drivers were, the more stress they
felt. No significant correlation was found between dislike
of driving and perceived stress (p > .05).
4. Discussion
The findings supported the contention that drivers who
displayed aggressive driving behavior showed higher
levels of stress than drivers who didn’t display aggres-
sive behavior while driving. Although no link was ob-
served between stress coping style and aggression, there
was some evidence that drivers who display a high prob-
lem-oriented coping style tended to display more instru-
mental aggression than hostile aggression. Moreover,
drivers whose driving style was characterized as the dis-
like group were inclined to react emotionally when cop-
ing with stress.
Using Folkman and Lazurus’ Cognitive Model which
describes coping with stress as an ongoing process of
evaluation, we were able to explain to some extent the
process that takes place when drivers express aggression
or anger while driving especially when facing stressful
situations. Overall, stress experienced by drivers as well
Table 6. Means (SD’s) for aggressive behavior comparisons
by driving style (Aggressive, Dislike of Driving)
Aggressive behavior
Yes No
Measures M SD M SD F(1,224)eta²
Aggressive style 2.18 .68 1.74 .49 9.15** .12
Dislike of driving
style
2.34 .44 2.53 .43 2.97 --
Note. Yes = was aggressive; No = wasn’t aggressive
n = 68; ** p < .01
as their coping styles influences is associated with their
behavior on the road. This strengthened our basic as-
sumption that the frustration-aggression model used up to
now by various researchers [9,33] to explain aggressive
behavior of drivers on the roads, does not offer a suffi-
cient or consistent explanation of drivers’ aggressive
reactions. It does not fully explain the process from the
moment the driver experiences frustration to the actual
behavioral reaction.
Among the new insights into driving behavior revealed
by the data was the importance of stress perceptions and
coping styles.
4.1 Stress and Coping with Stress
Drivers who displayed aggressive behavior had higher
levels of stress than drivers who didn’t display aggres-
sive behavior. These findings were consistent with ear-
lier literature that aggressive behavior was correlated
with reports of the driving experience as a stressful
event [2,3]. Our support here of this contention is also
consistent with findings that drivers suffering from ele-
vated levels of stress tended to perceive other drivers as
a source of this emotion and causing them to react more
aggressively towards the other driver, a form of road
rage [12,17]. Nevertheless, no direct link was found
between coping style, stress and driver aggression. Ag-
gressive and non-aggressive drivers were not distin-
guished by their coping style indicating that driving
usually involves stress and that stress is a common fac-
tor that exists for all drivers [17,20]. As already re-
ported in the literature, clear stressor stimuli such as
type and length of journey [34] or lack of control in
many driving situations [13] is a common feature of
most commuting experiences. It is safe to say that as
soon as the level of stress is elevated to a certain point,
drivers are prone to act aggressively, regardless of
cause or individual style of coping with stress.
Researchers attempting to identify the circumstances
under which drivers choose to use violence against other
drivers in order to solve problems on the road may want
to consider the stress variable as a probable main or con-
tributing cause. Our claim here is that drivers experienc-
ing elevated levels of stress tended to blame other drivers
and one way of dealing with the stress was to behave
aggressively, if not violently, towards them. In the study,
although drivers entering the parking lot were all ex-
posed to the same conditions, aggressive tendencies were
reported mainly among those who perceived stress.
Those drivers who score high on the problem-oriented
coping style tend to solve problems through instrumental
aggression, which is not meant to harm people and can
even be considered as a “healthy” way of coping with
stress while driving. This assumption is supported by the
regression analysis which showed that the coping with
stress variable had less effect on aggressive behavior.

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 42
Therefore, the problem-oriented coping style served as a
sort of moderating variable between perceived stress
while driving and aggression. Moreover, the negative
correlation between the task-oriented coping style and
perceived stress appears to indicate that drivers with this
kind of coping style are not inclined or tempted to react
violently, but rather choose behavior that mitigates their
feeling of stress.
The assumptions underlying the examination of driver
style and its relationship to coping with stress and ex-
pressions of aggression while driving were partially up-
held. The question is whether there is a link between the
driving styles categorized as “aggressive driving” and
“dislike of driving” and styles of coping [1] and aggres-
sive response. In correlation analyses, a connection was
found between aggressive driving style and perceived
stress. The higher the drivers scored on the aggressive
driving style measure, the greater the feeling of stress. In
our observations, we noted that the more aggressive
drivers were indeed those with an aggressive driving
style score. These findings are also compatible with an-
other finding, namely, drivers who reacted aggressively,
as compared to the non-aggressive ones, reported ex-
periencing higher perceived levels of stress. It should be
noted that in spite of the obvious connection between
aggressive driving style and high levels of stress, result-
ing in aggressive driving, there are no field or empirical
studies that have dealt with these associations. In parallel,
we found that those drivers who are averse to driving
cope with stress emotionally. This finding is a replication
of previous reports where drivers with high levels of
driving aversion preferred emotional reaction to stress
rather than behavioral reactions. These drivers reported
feeling worried about driving and handling the traffic but
coped with the stress of driving by using emotional cop-
ing strategies, such as self-blame or self-criticism [20].
Because they are inclined to blame themselves, it would
seem they prefer an internalized cognitive-emotional
reaction and reject an overt negative behavior that may
not be considered as effective.
4.2 Theoretical Contribution
In addition to using the cognitive model of coping with
stress [1] to explain the influence of stress on drivers’
reaction, this study has provided a specific, theoretical
contribution in defining aggressive behavior while driv-
ing. The research literature lacks a clear definition of
road anger or aggressive driving and it is difficult to dis-
tinguish between various aggressive expressions while
driving. Since a consistent and comprehensive definition
of aggressive driving is missing [35], lack of order and
an inability to test hypotheses characterize the field.
By dividing aggression into two types or categories, it
is possible to portray drivers using measures of stress and
coping styles. The first type includes aggressive behavior
that acts as a practical and deliberate solution to a prob-
lem on the road, whether by avoiding the situation or by
hurting others. The second type includes hostile behav-
iors for the purpose of getting rid of anger or fury which
are not connected to the problem. A partial answer to the
query whether road rage is a useful [36] or redundant [35]
phrase was provided here. It would appear that hostile
behaviors described in the present study include some of
the actions that typify “road rage.” Such hostile behavior
is purposely meant to hurt other drivers and is different
qualitatively from instrumental behaviors.
We think that this study makes an important contribu-
tion in clarifying both the process and outcome of the
driving experience. By providing definitions and appro-
priate categorizations, it is now possible to begin “talking
the same language.” It is worthwhile exploring other
avenues doing research in the future on drivers’ tenden-
cies to behave aggressively and to recognize them as
such. In spite of the connection between the drivers’
evaluation and the aggression they express, it is still not
clear whether the drivers’ tendency towards aggressive
driving influences the choice they make to express ag-
gression (instrumental or hostile) while driving. If it does,
how is it expressed (the level of aggression, frequency,
etc)?
4.3 Some Applications
The above findings may also have some important im-
plications for road safety and prevention of road acci-
dents, particularly concerning aggressive behaviors as
providing a possible underlying basis for explaining why
certain drivers tend to be involved in traffic violations or
road accidents. In another vein, results here can be ap-
plied in the compilation of training programs on road
safety focusing on the human factor and the psychology
of driving rather than on the traditional areas of training
and prevention of road accidents: teaching road skills;
regulations, infrastructures etc. It is not sufficient to fo-
cus on legislation or obeying the laws. The findings pre-
sented here may indicate a pressing need to focus on
psychological aspects of the driving experience and ways
for channeling the perceived stress into less negative
consequences. Ineffective, hostile solutions can be com-
pared to more effective instrumental ones with the goal
of modifying behaviors that can lead to road accidents.
4.4 Limitations and Future Research
In spite of the study outcomes, generalizing the results to
other populations is limited for a number of reasons. The
drivers were observed as they entered the parking lot.
This is a situation which doesn’t necessarily represent
drivers’ behavior while driving or in other situations.
This situation limited the possible range of aggressive
behaviors. For example, observations taken during the
day did not enable observations such as “light flashing”

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 43
or “high beaming”. It is not surprising that certain be-
haviors such as overtaking were not feasible and were
not observed. In addition, every “participant” in the study
was only observed once (as he/she drove into the parking
lot). Inferences here are limited because one observation
may not be representative of his/her driving behavior.
An improvement of the methodology in the future
could be to measure the exact waiting time of each driver
at the entrance to the parking lot, in order to be sure that
all the participants felt the same amount of frustration
and stress. In other words, we suggest measuring the
waiting time of each driver from the time he/she reaches
the parking lot to the time he/she goes past the entrance.
This measure could be used to assess the level of stress
caused by the circumstances and it would be able to dif-
ferentiate the drivers’ behaviors more successfully. Pre-
sumably the last driver in line would be more frustrated
than the driver at the head of the line because the latter
would have to wait less time.
In addition, because it is known that aggressive driving
is influenced by stressful situations, and by various situ-
ational factors which increase stress levels, such factors
might increase or lessen stress while driving, and this
should continue to be examined. One possibility would
be not to settle for a general measure of stress but rather
to carry out a number of measures of the stress variable,
and to differentiate between stress factors related to the
driver’s personality and situational factors. It is reason-
able to assume that a driver who lives far away from the
university, and has to travel, will be under more stress
than someone who doesn’t have to travel far.
A comment about the observational technique is in
order. Even though one person carried out the observa-
tions, and thus observer reliability/consistency was rela-
tively high, it is possible that the person who observed
the lot did not notice various behaviors inside or outside
the cars, such as “hand gestures”, or “swearing.” There-
fore, exact observation techniques should be used, such
as taking a picture of the drivers or having a number of
people observing the lot.
In conclusion, though we succeeded in showing that
the coping with stress model examined in the study is an
effective tool for better understanding driving and coping
styles, it is not clear to what degree the chosen situation
was a source of stress for the study participants, and
whether their feelings of stress were caused by other
factors not related to the observed situation. Other per-
sonal, as well as situational, variables need to be consid-
ered in the future so as to provide a more realistic picture
of the process leading to aggressive behavior.
REFERENCES
[1] R. S. Lazarus and S. Folkman, “Stress, Appraisal and
Coping,” Springer, New York, 1984.
[2] D. A. Hennessy and D. L. Wiesenthal, “Gender, Driver
Aggression and Driver Violence: An Applied Evalua-
tion,” Sex Roles , Vol. 44, No. 11-12, 2001, pp. 661-676.
[3] D. A. Hennessy and D. L. Wiesenthal, “Driving Venge-
ance and Willful Violations: Clustering of Problem Driv-
ing Attitudes,” Journal of Applied Social Psychology, Vol.
35, No. 1, 2005, pp. 61-79.
[4] P. Marsh and P. Collet, “The Car as a Weapon,” Et cetera,
Vol. 44, No. 2, 1987, pp. 146-151.
[5] L. Berkowitz, “Whatever Happened to the Frustration-
Aggression Hypothesis?” American Behavioral Scientist,
Vol. 21, 1978, pp. 691-707.
[6] J. Dollard, L. W. Doob, N. P. Miller, O. H. Mowrer and R.
R. Sears, “Frustration and Aggression,” Yale University
Press, New Haven, 1939.
[7] S. Feshbach, “The Function of Aggression and the Regu-
lation of Aggressive Drive,” Psychological Review, Vol.
71, 1964, pp. 257-272.
[8] B. G. Rule, “The Hostile and Instrumental Functions of
Human Aggression,” In de Wit, J., Hartup, W.W., Eds,
Determinations and Origins of A ggressive Behavior, Mou-
ton, The Hague.
[9] D. Shinar, “Aggressive Driving: The Contribution of the
Drivers and the Situation,” Transportation Research, Vol.
1, No. 2, 1998, pp. 137-160.
[10] D. Shinar, “Road Aggression: As Highway Frustrations
Grow, Drivers Become More Reckless,” Psychology In-
ternational, Vol. 10, No. 3, 1999, pp. 1-3.
[11] R. Martinez, National Highway Traffic Safety Admini-
stration, Washington, DC, 17 July 1997.
[12] E. Gulian, G. Matthews, A. I. Glendon, D. R. Davis and L,
M. Debney, “Dimensions of Driver Stress,” Ergonomics,
Vol. 32, 1989, pp. 585-602.
[13] M. Koslowsky, A. N. Kluger and M. Riech, “Commuting
Stress: Causes, Effects and Methods of Coping,” Plenum
Press, New York, 1995.
[14] R. W. Novaco, D. Stokols and J. Campbell, “Transporta-
tion, Stress and Community Psychology,” American
Journal of Community Psychology, Vol. 7, 1979, pp. 361-
380.
[15] N. Nicholson and P. M. Goodge, “The Influence of Social,
Organizational and Biographical Factors on Female Ab-
sence,” Journal of Management Studies, Vol. 13, 1976,
pp. 234-254.
[16] M. Koslowsky and M. Krausz, “On the Relationship be-
tween Commuting, Stress Symptoms and Attitudinal
Measures: A LISREL application,” Journal of Applied
Behavioral Science, Vol. 29, No. 4, 1993, pp. 485-492.
[17] E. Gulian, A. I. Glendon, G. Matthews, D. R. Davies and
L. M. Debney, “Exploration of Driver Stress Using
Self-Reported Data,” In T. Rothengatter and R. de Bru-
ineds, Road User Behaviour: Theory and ResearchVan
Gorcum, Assen/Maastricht, 1988.
[18] G. Matthews, “Cognitive Processes in Driver Stress,” In
Proceeding of the 1993 International Congress of Health
Psychology, International Congress of Health Psychology,

Aggression on the Road as a Function of Stress, Coping strategies and Driver Style 44
Tokyo, 1993, pp. 90-93.
[19] G. Matthews, L. Dorn, W. H. Thomas, D. R. Davis, A. I.
Glendon and R. G. Taylor, “Driver Stress and Perform-
ance on a Driving Simulator,” Human Factors, Vol. 40,
No. 1, 1998, pp. 136-149.
[20] G. Matthews, P. A. Desmond, L. Joyner, B. Carcary and
K. Gilliland, “A Comprehensive Questionnaire Measure
of Driver Stress and Affect,” In: Vaya, E.C. and Rothen-
gatter, J.A. Eds., Traffic and transport psychology: The-
ory and application, Pergamon, Amsterdam, 1997, pp.
317-324.
[21] M. C. W. Peeters, B. P. Buunk and W. B. Schaufeli, “A
Micro-Analytic Exploration of the Cognitive Appraisal of
Daily Stressful Events at Work: The Role of Controllabil-
ity,” Anxiety, Stress, and Coping, Vol. 8, 1995, pp. 127-
139.
[22] S. Folkman and R. S. Lazarus, “If it Changes it must be a
Process: A Study of Emotion and Coping during Three
Stages of a College Examination,” Journal of Personality
and Social Psychology, Vol. 48, No. 1, 1985, pp. 150-
170.
[23] S. Folkman and R. S. Lazarus, “Coping as a Mediator of
Emotion,” Journal of Personality and Social Psychology,
Vol. 54, 1988, pp. 466-475.
[24] S. Folkman, R. S. Lazarus, C. Dunkel-Schetter, A.
Delongis and R. J. Gruen, “Dynamics of a Stressful En-
counter: Cognitive Appraisal, Coping and Encounter
Outcomes,” Journal of Personality and Social Psychol-
ogy, Vol. 50, 1986, pp. 992-1003.
[25] E. J. Peacock, P. T. P. Wong and S. T. Reker, “Relations
between Appraisals and Coping Schemes: Support for the
Congruence Model,” Canadian Journal of Behavioral
Science, Vol. 25, pp. 65-80.
[26] A. I. Glendon, L. Dorn, G. Matthews, E. Gulian, D. R.
Davies and L. M. Debney, “Reliability of the Driving Be-
haviour Inventory,” Egronomics, Vol. 36, 1993, pp. 719-
726.
[27] C. D. Spielberger, “The Manual of State-Trait Anxiety
Inventory,” Consulting Psychologists Press, Palo Alto,
1983.
[28] Y. Teichman and H. Mellik, “State-Trait Anxiety Inven-
tory,” Tel-Aviv University, Ramot Press, Tel-Aviv, 1976.
[29] A. Avitzur, “Personal and Social Resources, Post-Trau-
matic Influence after Participating in War,” M.A. Thesis,
Bar-Ilan University, Ramat Gan, 1987.
[30] G. Matthews, P. A. Desmond, L. Joyner, B. Carcary and
K. Gilliland, “A Comprehensive Questionnaire Measure
of Driver Stress and Affect,” In: Vaya E.C. and Rothen-
gatter J.A. Eds., Traffic and Transport Psychology: The-
ory and Application, Amsterdam: Pergamon, 1997, pp.
317-324.
[31] R. W. Brislin, “Translation and Content Analysis of Oral
and Written Material,” In: Triandis, H.C. and Berry, J.W.
Eds., Handbook of Cross-Cultural Psychology: Method-
ology, Allyn and Bacon, Boston, Vol. 2, 1980, pp. 389-
444.
[32] B. A. Bettencourt, A. Talley, A. J. Benjamin and J. Val-
entine, “Personality and Aggressive Behavior under Pro-
voking and Neutral Conditions: A Meta-Analytic Re-
view,” Psychological Bulletin, Vol. 132, 2006, pp. 751-
777.
[33] A. N. Doob and A. E. Gross, “Status of Frustrator as an
Inhibitor of Horn-Honking Responses,” The Journal of
Soci al Psyc h ology, Vol. 76, 1968, pp. 213-218.
[34] M. Koslowsky and M. Krausz, “On the Relationship be-
tween Commuting, Stress Symptoms and Attitudinal
Measures: A LISREL Application,” Journal of Applied
Behavioral Science, Vol. 29, No. 4, 1993, pp. 485-492.
[35] C. Dula and S. E. Geller, “Risky, Aggressive, or Emo-
tional Driving: Addressing the Need for Consistent
Communication in Research,” Journal of Safety Research,
Vol. 4, No. 5, 2003, pp. 559-563.
[36] J. Yu, P. C. Evans and L. Perfetti, “Road Aggression
among Drinking Drivers: Alcohol and Non-Alcohol Ef-
fects on Aggressive Driving and Road Rage,” Journal of
Criminal Justice, Vol. 32, No. 50, 2004, pp. 421-430.