Vf

Aa

0b

a

ARRA

KRQD

1

fpboarf

(

h0

Accident Analysis and Prevention 70 (2014) 293–300

Contents lists available at ScienceDirect

Accident Analysis and Prevention

jo u r n al homepage: www.elsev ier .com/ locate /aap

alidation of selected temperament and personality questionnairesor diagnosing drivers’ aptitude for safe driving. A Polish study

nna Łuczaka,∗, Adam Tarnowskib,1

Central Institute for Labour Protection – National Research Institute, Department of Ergonomics, Laboratory of Social Psychology, Czerniakowska 16,0-701 Warsaw, PolandUniversity of Warsaw, Faculty of Psychology, Stawki 5/7, 00-183 Warsaw, Poland

r t i c l e i n f o

rticle history:eceived 26 June 2013eceived in revised form 3 March 2014ccepted 3 April 2014

eywords:oad crashesuestionnairerivers’ aptitude

a b s t r a c t

This paper presents the results of a study aimed at validating psychological questionnaires evaluatingtemperamental and personality features. It discusses their usefulness in diagnosing drivers’ aptitude forsafe driving and working as professional drivers. Three psychological questionnaires were validated:the Formal Characteristics of Behaviour – Temperament Inventory (FCB-TI), the Eysenck PersonalityQuestionnaire – Revised and Short Scale (EPQ-R (S)) and the Impulsiveness Questionnaire (IVE). Threegroups of drivers (n = 246) aged 19–75 participated in the study. Group I (professional drivers; n = 96)and Group II (nonprofessional drivers; n = 75) had never been involved in road crashes, whereas Group III(nonprofessional drivers; n = 75) were offenders involved in fatal injury road crashes. Criterion-relatedvalidity, Cronbach’s alpha and Guttman split-half reliability coefficient were in assessing the psycho-metric properties of the questionnaires. There were some significant differences between Groups II andIII for most traits. However, contrary to expectations, higher Emotional Reactivity, Perseveration andlower Endurance as well as higher Neuroticism, Impulsiveness and Venturesomeness were determinedfor Group II than for Group III. Additionally, the temperament and personality profile of Group II turnedout to be less fitted to the profile of safe drivers than that of Group III, whose profile was actually similarto that of Group I. This seems to result from a high tendency for a positive self-presentation among Group

I and Group III (a significantly higher result on the Lie scale in comparison with Group II). The resultssuggest that if psychological tests are to decide on whether a person may be a professional driver or maydrive vehicles, the three questionnaires (FCB-TI, EPQ-R(S) and IVE) do not provide a valid diagnosis ofprofessional drivers’ aptitude because of drivers’ high tendency for positive self-presentation. However,they can be used in job counselling and in screening high-risk drivers.

© 2014 Elsevier Ltd. All rights reserved.

. Introduction

Diagnosis and certification of driving aptitude of both nonpro-essional and professional drivers as well as selection for the job arearticularly important in the field of transport and traffic. This is soecause road transport is considered to be one of the most danger-us economic sectors with respect to the number of road crashes,

nd one of the most cost-consuming means of transport for socialeasons (EU Strategy, 2007–2012). Poland has been a leader in roadatalities in Europe for many years. In 2012, the road crash severity

∗ Corresponding author. Tel.: +48 22 623 46 09; fax: +48 22 623 32 82.E-mail addresses: anluc@ciop.pl (A. Łuczak), adam.tarnowski@psych.uw.edu.pl

A. Tarnowski).1 Tel.: +48 22 554 97 77.

ttp://dx.doi.org/10.1016/j.aap.2014.04.001001-4575/© 2014 Elsevier Ltd. All rights reserved.

rate in Poland was 10 deaths per 100 crashes, whereas it was justover 3 fatalities per 100 crashes in the European Union (Zielinska,2013). In 2011, the cost of road crashes in Poland accounted forabout 1.3% of GDP (Jazdzik-Osmólska, 2012). In the same year, pro-fessional drivers in Poland caused 8.7% of road crashes, they werethe third largest occupational group with respect to accidents atwork (10%).

Practice shows that making driving accessible solely to per-sons with a certain aptitude is highly effective in lowering therisk of human error and, what follows, road transport crashes. Atthe European level, the Council Directive of 29 July 1991 on driv-ing licences (91/439/EEC) takes up this issue. Psychological testing

for drivers to diagnose their aptitude for safe driving is a possi-bility. In Poland, under the Motor Vehicle Drivers’ Act (Ustawa okierujacych pojazdami, 2011), psychological testing is obligatoryfor professional drivers and, in some cases, nonprofessional ones

2 alysis

(o

miitowcir

1

tril1rutcNaraeadCabsds

iissoueomit(tdsacssdb

taotc

94 A. Łuczak, A. Tarnowski / Accident An

after drunk- or drug-driving, after scoring over 24 penalty pointsr after being involved in a road crash).

Reliable psychological diagnostics requires, as a prerequisite, aethodology that meets precisely defined standards for the qual-

ty of psychological tests (Standards, 1999). The usefulness of testsn psychological examinations of drivers is based on various fac-ors, including their criterion-related validity and, in particular, onef its types, concurrent validity, which makes it possible to certifyhether drivers have an aptitude for driving. The validity of psy-

hological tests used in psychology and road traffic is a key factorn deciding about their usefulness in terms of lowering the risk ofoad crashes.

.1. Temperament and personality factors, and road crashes

Most studies on the causes of road crashes focus on the rela-ionship between specific traits and abilities of drivers, and crashisk. The personality factors most often selected for that purposenclude the Big Five personality dimensions. High Extraversion,ow Conscientiousness and low Agreeableness (Arthur and Graziano,996; Clarke and Robertson, 2005) are most often considered accu-ate predictors of crashes. However, study results are not alwaysnambiguous. Some showed that Extraversion correlated posi-ively with the number of heavy road crashes, whereas Neuroticismorrelated negatively (Lajunen, 2001). Others showed that higheuroticism, along with high Extraversion, low Conscientiousnessnd low Agreeableness, were related to crash risk, but only indi-ectly, as those traits had a direct impact on aberrant behaviours measured with the Driver Behaviour Questionnaire (DBQ) (Sümert al., 2005). Similarly, Taubman-Ben-Ari and Yehiel (2012) proved

direct link between personality factors and a specific type ofriving behaviour. High Neuroticism, low Agreeableness and lowonscientiousness characterised drivers admitting to a reckless andggressive driving style. For those drivers, driving involved suchenefits as a sense of power and control over a situation, and plea-ure and thrill seeking, whereas costs were defined in terms ofanger and a threat to life, potential problems, irritation and pos-ible damage to one’s personal image.

Locus of control (which here also applies to control in traffic)s another predictor of crashes and mistakes made while driv-ng (Montag and Comrey, 1987; Özkan and Lajunen, 2005). At theame time, controllability awareness turned out to be a variable thatignificantly differentiated offending drivers from non-offendingnes. Offending drivers found distinguishing between situationsnder and out of their control a problem. Possibly that is why theyvaluated all requirements as threats rather than challenges. More-ver, offending drivers were characterised by a significantly worseood (higher anxiety, depression, tiredness, confusion, dimin-

shed vitality) and substantially more indirect forms of aggressionhan non-offending drivers (Sanval et al., 2012). However, Sümer’s2003) study on professional and nonprofessional drivers (wherehe number of road crashes in which the drivers had been involveduring the previous 3 years was used as a control) showed thatuch psychological symptoms as depression and anxiety, as wells hostility and psychoticism, had a rather indirect impact on roadrashes through their effect on drivers’ aberrant behaviour as mea-ured with the DBQ. Sensation seeking had a direct impact only onpeed driving, and not on road crashes, and aggression directly pre-icted dysfunctional drinking (drinking and driving, or antisocialehaviour after drinking).

There is a link between road crashes and aggression and viola-ions in road traffic, as expressed in negativism, irritation and verbal

nd physical hostility (Li et al., 2004; Nabi et al., 2005, 2006). Onef the first studies to analyse aggression on the road comparedwo groups of taxi drivers, who had caused a different number ofrashes; it showed that drivers with more road crashes displayed a

and Prevention 70 (2014) 293–300

larger share of socially unacceptable behaviours and had more fre-quent contact with the police and courts (Tillman and Hobbs, 1949).Moreover, risk-taking propensity, expressed in such behaviours asdriving to relieve stress or racing other cars involving speed driving,were valid predictors of involvement in road crashes (Ivers et al.,2009; Patil et al., 2006).

Type A behaviour pattern (Nabi et al., 2005) is closely related tothe number of serious road crashes. One of its aspects, always beingin a hurry and under time pressure, was discovered to be significantlyconnected with the number of near crashes (Karlberg et al., 1998).

Other studies of professional drivers showed a link betweentemperament and general perceived and traffic-related stress(Waszkowska, 2009). The level of traffic-related stress was lowestin persons with a higher ability to process stimulation and act underlong-term or strong stimulation (high Briskness), lower sensitivityto emotional stimuli (low Emotional reactivity), lower tendency torepeat a reaction to a stimulus that was no longer there (low Per-severation) and a higher tendency for behaviours marked by highstimulation value (high Activity).

Studies also showed that certain configurations of tempera-ment and personality factors could constitute a special road crashrisk factor, for example, a tendency to perceive various eventsas potentially bothersome (high Perseveration and low Endurance)(Trzcinska, 1996), an emotional style of coping with stress (highEmotional reactivity and high Perseveration) (Strelau et al., 2005)and a nonintegrated personality (low Neuroticism coexisting withhigh Extraversion and high Lie) (Aneks, 2003).

1.2. The validity of personality questionnaires

The results of studies on the validity of personality questionnaireare ambiguous. A review of studies (1952–1963) on the validity ofpersonality questionnaires in job selection, which considered 14measures, showed quite high criterion-related validity. The val-ues of correlation coefficients between some of the scales at theanalysed inventories and external criteria were between .15 forexecutives in the Guilford-Martin Personnel Inventory and .55 forengineers in Gough’s California Psychological Inventory (Guion andGottier, 1965). But a meta-analysis of 117 studies (1952–1988) oncriterion-related validity of the Big Five personality dimensionsconducted from the point of view of their usability in job selec-tion indicated that the average correlations between personalityfeatures and the criterions of job selection (job proficiency, train-ing proficiency and personnel data) were low, for example, .07–.22for Conscientiousness (Barric and Mount, 1991). Also Friedman andBooth-Kewley (1987) showed that the predictive validity of person-ality features for evaluating incidence of psychosomatic disorderswas low (.1–.2). Similarly, an analysis of 494 studies indicated poorcorrelations between several dozen personality scales and job per-formance indicators (.29 and .12, depending on the method) (Tettet al., 1991), whereas Pasquier and Mazilescu’s (2009) study onthe predictive validity of a five-factor model in evaluating stu-dents’ achievements showed that only Harshness, the opposite ofAgreeableness, correlated significantly with students’ grades (−.19),although the strength of this relation expressed with the value ofCohen’s d (.39) was poor. An attempt to estimate construct valid-ity on the basis of the relationship between psychophysical andpsychophysiological indicators (e.g., electrical brain activity, event-related potentials, time and strength of motor reaction, speed ofconditioning) and extraversion, neuroticism as well as sensationseeking produced divergent results. According to Strelau (1998),this is so because of the differences in the quality of the indica-

tors. Guion and Gottier (1965) suggested that there was no hardevidence that personality questionnaires could be recommendedas a good measure in job selection. They also say that “the validityof any personality measure must be specifically and competently

alysis

dfTld

dFT(twtcdtswAV

2

2

d

•

•

•

2

og

•

•

A. Łuczak, A. Tarnowski / Accident An

etermined for the specific situation in which it is to be used andor the specific purpose or criterion within that situation” (p. 160).herefore, users of personality questionnaires should know theimitations of this kind of measures to avoid misinterpretation iniagnoses based on their results.

This article describes the results of research aimed at a vali-ating three questionnaires of temperament and personality: theormal Characteristics of Behaviour – Temperament Inventory (FCB-I), the Eysenck Personality Questionnaire – Revised and Short ScaleEPQ-R (S)) and the Impulsiveness Questionnaire (IVE), to determineheir usefulness in diagnosing drivers’ aptitude for driving and fororking as professional drivers. The main research problem was,

herefore, defined in the form of a question: Do drivers who haveaused a road crash with fatalities or injuries differ significantly fromrivers who have never caused a road crash in terms of their level ofemperament and personality traits, measured with the questionnaireselected for validation? Drivers who had caused heavy road crashesere expected to have a significantly higher Emotional reactivity,ctivity, Perseveration, Extraversion, Neuroticism, Impulsiveness andenturesomeness and a significantly lower Endurance.

. Method

.1. Questionnaires

Three psychological questionnaires were selected for vali-ation:

Formal Characteristics of Behaviour – Temperament Inven-tory (FCB-TI) (Strelau and Zawadzki, 1993); this questionnaireaddresses the energetic aspect of behaviour expressed with Sen-sory Sensitivity, Endurance, Emotional Reactivity and Activity,and to its temporal aspect expressed with Briskness and Perse-veration;Eysenck Personality Questionnaire – Revised and Short Scale(EPQ-R (S)) (Eysenck and Eysenck, 2006), in its Polish adaptation(Jaworowska, 2011a); this questionnaire evaluates three dimen-sions of personality, Extraversion, Neuroticism and Psychoticism;it also has an additional scale, Lie. In view of the low reliability ofthe Psychoticism scale, this dimension was not considered in theanalysis;Impulsiveness Questionnaire (IVE) (Eysenck and Eysenck, 2006),in its Polish adaptation (Jaworowska, 2011b); this questionnaireaddresses three other personality traits, Impulsiveness, Venture-someness and Empathy.

.2. Participants

The analysis was performed on the basis of questionnaire resultsbtained for 246 drivers of both genders, aged 19–75, from threeroups:

Group I: professional drivers “without crashes” (n = 96), profes-sional drivers who had never caused a road crash with fatalitiesor injuries, who had not caused any traffic collision in the past 2years, and who had been professional drivers for at least 2 years.The drivers had category B and C driving licences (they were taxi,truck and fleet drivers). This group of drivers participated in ourstudy as part of their mandatory (under Poland’s Drivers’ Act)periodic psychological examinations.Group II: nonprofessional drivers “without crashes” (n = 75),

drivers who had never caused a road crash with fatalities orinjuries or a road collision in the past 2 years, had a nonpro-fessional category B driving licence, and who had been activedrivers for at least 2 years. Those drivers were selected for the

and Prevention 70 (2014) 293–300 295

study in the course of telephone recruitment, at the Central Insti-tute for Labour Protection – National Research Institute, wherepart of the study was took placed; they expressed their consentto participate in the study.

• Group III: nonprofessional drivers “with crashes” (n = 75), driverswith a nonprofessional category B driving licence and at least 2years of driving experience, who in the past 2 years had causeda road crash with fatalities or injuries. Those drivers participatedin our study as part of mandatory (under Poland’s Drivers’ Act)psychological post-crash psychological examinations.

2.3. Validation method

The criterion-related validity method was selected to evaluatethe validity of the questionnaires; it is recognised as a particularlypractical for tests for selecting and classifying employees, includ-ing those for jobs requiring special aptitude (Anastasi and Urbina,1999).

The safe functioning of a person as a driver was adopted as thecriterion variable, whereas having/not having a history of causinga heavy road crash was adopted as an indicator of that criterion.

The significance of the differences between average test resultsof nonprofessional drivers “without crashes” and “with crashes”was adopted as an indicator of questionnaire validity for diagnosingpersonality risk factors.

Validity was assessed on the basis of the scores of a samplehomogenous in terms of the drivers’ age and years of work as pro-fessional drivers/driving experience: professional drivers “withoutcrashes” (n = 46), nonprofessional drivers “without crashes” (n = 75)and nonprofessional drivers “with crashes” (n = 75). Internal consis-tency and split-half reliability of the questionnaires were assessedwith Cronbach’s alpha and Guttman split-half reliability coefficient,respectively. Reliability was analysed on the basis of the question-naire scores of professional drivers “without crashes” (n = 96).

2.4. Procedure

The study took place in 16 laboratories for psychological testingof drivers in Poland, where the participants completed the ques-tionnaires. Each driver who participated in the study on a voluntarybasis (Group II) had a short conversation with the person in chargeof the testing. The drivers were informed again of the purpose of theexaminations; the course of examination was explained to them.(The aim of the study was first explained during the recruitmentprocedure on the phone.) Next, they filled in the three question-naires in random order. The examination lasted about 40 min. Thedrivers from this group were paid 22 EUR for participating in thestudy.

Drivers reporting at laboratories for psychological testing ofdrivers for their periodic or post-crash psychological examina-tions (Groups I and III) went through the standard examinationprocedure, as specified in the Polish methodology of psychologi-cal examination of drivers, with the additional temperament andpersonality questionnaires. The examination lasted about 3 h; thedrivers were not paid for their participation.

2.5. Statistical analyses

The one-sample Kolmogorov–Smirnov test was used to comparethe consistency of the distribution of the test results and nor-

mal distribution; the Kruskal–Wallis one-way analysis of varianceby ranks to compare the results of the three groups of drivers andthe Mann–Whitney U test to compare pairs of groups of drivers.The SPSS statistical package, version 14.0, was used.

2 alysis and Prevention 70 (2014) 293–300

3

3

3I

Gatrao.

tF1

3S

(rasb(

3

rfi((tb(

3

no

a(spneodtdd

hmt

v

3

s

Table 1Characteristics of validation groups in terms of age, years of work/experience, kilo-metres covered and penalty points in the previous year.

Variable Statistics Group I(n = 46)

Group II(n = 75)

Group III(n = 75)

Age (years) M 37.65 36.39 36.76Min 21 20 19Max 69 71 75SD 13.097 13.698 13.83

Years ofwork/experience

M 16.52 15.00 15.01

Min 3 2 2Max 42 35 54SD 11.59 10.69 11.86

Kilometres[thousands](previous year)

M 45.56 17.22 30.57

Min 3.00 0.20 2.00Max 300.00 90.00 170.0SD 47.07 15.00 36.85

Penalty points(previous year)

M 2.65 0.25 2.50

Min 0 0 0

96 A. Łuczak, A. Tarnowski / Accident An

. Results

.1. Test reliability

.1.1. Formal Characteristics of Behaviour – Temperamentnventory FCB-TI

The analysis of FCB-TI reliability, based on Cronbach’s alpha anduttman split-half reliability coefficient indicated that the reli-bility of Briskness was unacceptable (.49 and .23 respectively);he reliability of Sensory sensitivity was questionable (.66 and .61,espectively). On the other hand, the reliability of Perseveration wascceptable (.77 for both reliability coefficients) and the reliabilityf Emotional reactivity, Endurance and Activity was good (.85 and

86, .84 and .85, and .85 and .77, respectively).The values of Cronbach’s alpha for Briskness and Sensory sensi-

ivity were lower than their internal consistency according to theCB-TI manual (.77 and .73, respectively) (Strelau and Zawadzki,993).

.1.2. Eysenck Personality Questionnaire – Revised and Shortcale EPQ-R(S)

The analysis showed good reliability of all EPQ-R(S) dimensionsfor Cronbach’s alpha and Guttman split-half reliability coefficient,espectively): Neuroticism (.81 and .80), Extraversion (.83 and .87)nd Lie (.81 and .85). The reliability of EPQ-R(S) based on thecores of professional drivers was similar to its internal consistencyased on the scores for a representative Polish population (.81–.85)Jaworowska, 2011a).

.1.3. Impulsiveness Questionnaire IVEThe analysis of the IVE questionnaire indicated questionable

eliability (Cronbach’s alpha and Guttman split-half reliability coef-cient respectively) of Impulsiveness (.71 and .68) and Empathy.66 and .63), whereas the reliability of Venturesomeness was good.85 and .83). The result relating to the lowest reliability of Empa-hy in professional drivers was similar to the internal consistencyased on scores for a representative Polish population (.81–.85)Jaworowska, 2011b).

.2. Test validity

Table 1 shows the drivers’ data on age, years of work/experience,umber of kilometres covered in the previous year and the numberf penalty points.

The three groups of drivers were homogeneous in terms of theirge (Chi-squared (2) = .597, p = .742) and years of work/experienceChi-squared (2) = .884, p = .643). Professional drivers had coveredignificantly more kilometres in the previous year than non-rofessional drivers “without crashes” (Z = –5.798, p = .000) andonprofessional drivers “with crashes” (Z = −3.433, p = .001). How-ver, there was no significant difference between the two groupsf nonprofessional drivers (Z = −1.260, p = .208). Nonprofessionalrivers “without crashes” scored significantly fewer penalty pointshan professional drivers (Z = −5.118, p = .000) and nonprofessionalrivers “with crashes” (Z = −4.040, p = .000). The latter two groupsid not differ in this respect (Z = −.889, p = .374).

Both groups of nonprofessional drivers (Groups II and III) wereomogeneous in terms of gender and comprised 16 females and 59ales each. There was one female professional driver in Group I,

he other 45 were male.Tables 2–4 present basic statistics for the results scored by the

alidation groups in the questionnaires.

.2.1. Groups II and IIIAn analysis performed with the Mann–Whitney U test revealed

tatistically significant differences (p < .05) for the following traits:

Max 15 6 26SD 4.15 1.03 4.47

Formal Characteristics of Behaviour – Temperament Inventory(FCB-TI): the average result scored by nonprofessional drivers“without crashes” (Group II) was significantly higher than theresult obtained by nonprofessional drivers “with crashes” (GroupIII) for Perseveration (U = 2076.00; p = .008) and Emotional reactiv-ity (U = 1956.500; p = .003), but significantly lower for Endurance(U = 1799.500; p = .000).

Eysenck Personality Questionnaire – Revised and Short Scale(EPQ-R (S)): the average result scored by nonprofessional drivers“without crashes” (Group II) was significantly higher than theresult obtained by nonprofessional drivers “with crashes” (GroupIII) for Neuroticism (U = 1787.500; p = .000), but significantly lowerfor Extraversion (U = 2270.500; p = .040) and for Lie (U = 1831.00;p = .000).

Impulsiveness Questionnaire (IVE): the average result scored bynonprofessional drivers “without crashes” (Group II) was signifi-cantly higher than the result obtained by nonprofessional drivers“with crashes” (Group III) for Impulsiveness (U = 1874.500; p = .001)and for Venturesomeness (U = 1964.50; p = .003).

3.2.2. Groups I and IIAn analysis performed with the Mann–Whitney U test revealed

statistically significant differences (p < .05) for the following traits:Formal Characteristics of Behaviour – Temperament Inventory

(FCB-TI): the average result scored by professional drivers “with-out crashes” (Group I) was significantly lower than the resultobtained by nonprofessional drivers “without crashes” (GroupII) for Perseveration (U = 988.000; p = .000) and Emotional Reactiv-ity (U = 713.000; p = .000), but significantly higher for Endurance(U = 950.500; p = .000).

Eysenck Personality Questionnaire – Revised and Short Scale (EPQ-R (S): the average result scored by professional drivers (GroupI) was significantly lower than the result obtained by nonpro-fessional drivers “without crashes” (Group II) for Neuroticism(U = 838.000; p = .000), but it was significantly higher for Extraver-sion (U = 1320.500; p = .030) and for Lie (U = 892.000; p = .000).

Impulsiveness Questionnaire (IVE): the average result scored

by professional drivers (Group I) was significantly lower thanthe result obtained by nonprofessional drivers “without crashes”(Group II) for Impulsiveness (U = 1146.500; p = .003).

A. Łuczak, A. Tarnowski / Accident Analysis and Prevention 70 (2014) 293–300 297

Table 2Basic statistics for results scored by the validation groups in Formal Characteristics of Behaviour – Temperament Inventory FCB-TI.

Drivers Statistics Briskness Perseveration Sensory sensitivity Emotional reactivity Endurance Activity

Professional drivers “withoutcrashes” (Group I)

n 46 46 46 46 46 46

M 17.00 8.37 15.61 3.67 14.91 11.13Mdn 17.00 8.50 16.00 3.00 17.00 11.50SD 2.20 3.67 3.03 3.87 4.63 3.82Min 8 2 8 0 5 4Max 20 16 20 15 20 18

Nonprofessional drivers“without crashes” (Group II)

n 75 75 75 75 75 75

M 15.65 11.43 14.83 8.56 10.85 9.67Mdn 16.00 12.00 16.00 9.00 11.00 10.00SD 3.62 4.00 3.68 4.73 5.11 5.08Min 6 1 4 0 1 0Max 20 19 20 18 20 19

Nonprofessional drivers “withcrashes” (Group III)

n 74 74 75 73 73 74

M 16.62 9.74 15.69 6.33 13.86 10.28Mdn 17.00 10.00 16.00 6.00 14.00 10.50SD 2.31 4.15 2.83 4.31 4.42 3.94Min 8 1 6 0 2 0Max 20 20 20 20 20 17

Table 3Basic statistics for results scored by the validation groups in Eysenck PersonalityQuestionnaire – Revised and Short Scale EPQ-R(S).

Drivers Statistics Neuroticism Extraversion Lie

Professional drivers“without crashes”(Group I)

n 46 46 46

M 1.72 9.15 8.20Mdn 1.00 9.50 9.00SD 2.07 2.34 2.74Min 0 4 2Max 10 12 12

Nonprofessionaldrivers “withoutcrashes (Group II)

n 75 75 75

M 4.25 7.88 5.85Mdn 3.00 8.00 6.00SD 3.21 3.05 2.42Min 0 0 1Max 12 12 11

Nonprofessionaldrivers “withcrashes” (Group III)

n 75 75 75

M 2.40 8.95 7.48Mdn 2.00 10.00 8.00SD 2.41 2.40 2.71

3

atprpnp

3

md

Table 4Basic statistics for results scored by the validation groups in Impulsiveness Ques-tionnaire IVE.

Drivers Statistics Impulsiveness Venturesomeness Empathy

Professional drivers“without crashes”(Group I)

n 46 46 46

M 3.80 6.70 11.78Mdn 3.00 6.50 12.00SD 2.61 4.32 2.91Min 1 0 5Max 12 16 17

Nonprofessionaldrivers “withoutcrashes (Group II)

n 74 74 74

M 5.62 7.85 12.30Mdn 6.00 7.00 12.00SD 3.40 4.56 3.33Min 0 0 4Max 14 15 18

Nonprofessionaldrivers “withcrashes” (Group III)

n 74 74 74

M 3.86 5.66 12.57Mdn 3.00 5.00 13.00SD 2.83 3.98 3.15Min 0 0 5Max 11 15 18

Min 0 1 0Max 10 12 12

.2.3. Groups I and IIIAn analysis performed with the Mann–Whitney U test revealed

statistically significant difference (p < .05) only for Emotional Reac-ivity in the FCB-TI questionnaire. The average result scored byrofessional drivers (Group I) was significantly lower than theesult obtained by drivers “with crashes” (Group III) (U = 1008.000;

= .000). For other temperament and personality traits, there wereo significant differences between professional drivers and non-rofessional drivers “with crashes”.

.3. Temperament and personality profiles

Separate temperament and personality profiles were deter-ined for each group on the basis of the traits that significantly

ifferentiated them. Figs. 1–3 illustrate the profiles.Fig. 1. Temperament profile based on Formal Characteristics of Behaviour – Tem-perament Inventory FCB-TI in the three groups of drivers.

298 A. Łuczak, A. Tarnowski / Accident Analysis

Fa

ihnapnplmeposcmehat

Iaowpatqm

Fg

ig. 2. Temperament profile based on Eysenck Personality Questionnaire – Revisednd Short Scale EPQ-R(S) in the three groups of drivers.

Nonprofessional drivers “without crashes” (Group II) signif-cantly differed from the other groups. They had significantlyighest Perseveration, Emotional Reactivity, Neuroticism, Impulsive-ess, and significantly lowest Endurance and Extraversion as wells the lowest score on the Lie scale. Therefore, when set againstrofessional drivers and nonprofessional drivers “with crashes”,onprofessional drivers “without crashes” turned out to be mostrone to excessively focus on the past, its minute analysis and long-

asting emotional reactions to its events and past decisions, reactingore easily and stronger to trivial events; they also more often felt

motional tension, marked by higher lability, lower maturity andoorer emotional control, and had a greater decrease in the levelf task performance under stress. At the same time, in compari-on with the other two groups, nonprofessional drivers “withoutrashes” turned out to be significantly less resilient to the require-ents of difficult situations and less resistant to unfavourable

nvironmental factors. The results also suggested that the groupad a significantly higher tendency to undertake risky behaviours,lso those connected with failing to consider the consequences ofheir own behaviours.

Professional drivers turned out to have lowest Perseveration,mpulsiveness and Neuroticism and significantly highest Endurance,s well as the highest score on the Lie scale. Therefore, on the basisf the results, it is possible to describe the group, in comparisonith nonprofessional drivers “without crashes”, as significantly lessrone to focus on the past, more balanced and emotionally mature,

nd better able to react adequately in difficult situations, such ashose provoking long-term or strong stimulation. Moreover, theuestionnaires showed that they were significantly less prone toood swings and depression, regained their emotional balance

ig. 3. Temperament profile based on Impulsiveness Questionnaire IVE in the threeroups of drivers.

and Prevention 70 (2014) 293–300

after strong stimulation much faster and were significantly lessprone to undertake risky behaviours without thinking about theirpossible consequences. However, their higher score on the Lie scalesuggested that professional drivers had a greater need for a positiveself-presentation than nonprofessional drivers “without crashes”.

The temperament and personality profile of nonprofessionaldrivers “with crashes” is also similar to the profile generated forprofessional drivers with respect to their result on the Lie scale,which was significantly higher than that scored by nonprofessionaldrivers “without crashes”. The only trait significantly differentiat-ing professional drivers from drivers “with crashes” was their levelof Emotional Reactivity, which was significantly lower for profes-sional drivers.

3.4. Temperament and personality profiles, and crash risk

An additional analysis identified in the validation group driverswith a trait profile connected with individual risk factors. Tennonprofessional drivers “without crashes” (5.1%) had a tendencyto perceive various events as potentially burdensome. Seventeendrivers (8.7%; 1 professional driver, 4 drivers “with crashes” and12 nonprofessional drivers “without crashes”) had a non-effective,emotional strategy for managing stress. 12 drivers (6.1%; 4 nonpro-fessional drivers “without crashes”, 8 drivers “with crashes”) had anon-integrated personality.

4. Discussion

The results of our analysis of the three psychological question-naires evaluating temperament and personality traits, and checkingtheir validity for diagnosing an aptitude for safe driving, cannotbe interpreted unambiguously. It was assumed that the validity ofthe questionnaires would be measured with the criterion validitymethod, where the safe functioning of a person as a driver wasthe criterion variable, while having/not having a history of causinga road crash with fatalities or injuries was the criterion indica-tor. The considerable differentiation in the questionnaire resultsbetween nonprofessional drivers “with crashes” and nonprofes-sional drivers “without crashes” was an indicator of questionnairevalidity. The results scored by drivers “with crashes” and those“without crashes” differed significantly for eight out of the 13 fea-tures measured with the questionnaires. However, the nature of thedifferentiation was unexpected. The temperament and personalityprofile of drivers who had caused heavy road crashes satisfied therequirements for drivers better than the profile of drivers “withoutcrashes”. Nonprofessional drivers “with crashes” turned out to havelower Perseveration, lower emotional reactivity and higher strength;they were less neurotic, less impulsive and less venturesome thannonprofessional drivers “without crashes”. However, a significantlyhigher score on the Lie scale in the EPQ-R(S) questionnaire makesit possible to conclude that their better driver’s profile resultedfrom their considerable need for a positive self-presentation of thetraits that a safe driver should have. Actually, the driving style ofdrivers “with crashes” was less safe, as evidenced by their signifi-cantly higher number of penalty points, which are equivalent to thenumber of recorded cases of the traffic offences they committed,in comparison with drivers “without crashes” with a comparabledriving experience and number of kilometres travelled the previousyear.

A stronger need for a positive self-presentation in drivers “withcrashes” in comparison with drivers “without crashes” could have

resulted from the different situations in which both groups of non-professional drivers were examined. Drivers “with crashes” wereobliged to participate in a psychological examination; they did soon the basis of a referral issued under the Motor Vehicle Drivers’ Act.

alysis

Mibduitwds

epttaasi

p2ldttbamta

5

Fsoadad

podpii

udsaseipi

tccfs

A. Łuczak, A. Tarnowski / Accident An

oreover, the result of the psychological examination was a factorn the decision whether drivers with an crash history would getack their suspended driving licence. In contrast, nonprofessionalrivers “without crashes” participated in the examination as vol-nteers and the results were of no consequence to them. However,

t is worth pointing out that most cases of “risk” profiles relatedo a temperament-related risk factor or non-integrated personalityere found in nonprofessional drivers, including nonprofessionalrivers “with crashes”, despite their strong motivation for a positiveelf-presentation.

For professional drivers, like for drivers “with crashes”, thexaminations were mandatory. They were part of their periodicsychological examinations, the result of which was the basis ofhe decision whether or not to extend the validity of their licenceo work as a driver. This is possibly the reason why there was

significantly higher tendency for a positive self-presentationmong professional drivers, too, in comparison with nonprofes-ional drivers “without crashes”, for whom the result was of generalnterest only; they could learn about their aptitude for driving.

It is worth noting that the tendency for a positive self-resentation is common in self-descriptive methods (Owsley et al.,003). According to Owsley et al., personality traits may impact the

evel of the drivers’ honesty when disclosing information on theirriving behaviours. That is why empathic persons feel a strong needo be honest and admit their errors. On the other hand, if they admitheir errors, they do not want to disclose their proneness to riskyehaviours. Moreover, relying solely on self-descriptive methods innalysing links between drivers’ behaviour and the risk of a crashay lead to inflated results (Dorn et al., 2010), as data collected in

his way may be partially subject to bias, measurement errors andrtefact (Sümer, 2003).

. Conclusions

The results of our analysis of three psychological tests, i.e.,CB-TI, EPQ-R(S) and IVE, lead to a conclusion that they are notuitable for job selection. They do not provide a reliable diagnosisf the required aptitude of drivers, in view of a considerable risk ofn impact of a strong motivation for positive self-presentation inrivers undergoing mandatory psychological examinations aimedt determining whether they can or cannot work as professionalrivers.

It also seems that FCB-TI, EPQ-R(S) and IVE may be used in thesychology of transport as screening tests in special-risk groupsf drivers (drivers who have caused crashes, drivers engaging inrink- or drug-driving, drivers scoring more than the 24-penalty-oints limit). However, we should bear in mind that a result

ndicating a “risk” profile should be confirmed with additionalnformation (e.g., other tests and specialist consultations).

Additionally, it seems that there are no contraindications againstsing those tests in career guidance and counselling with a view toetermining a profile of various traits, skills and abilities of work-eeking individuals, changing the nature of their career or planning

future path of their career and identifying those areas of profes-ional activity to which a given profile is best suited. Psychologicalxaminations used in career guidance and counselling do not resultn a certification of one’s aptitude for, or contraindications against, aarticular job; they inform, educate or both on aptitude for specific

ndividual careers.It is worth emphasising that questionnaire methods, based on

he subjects’ self-description of traits or behaviours, require a

autious interpretation of the results, especially if psychologicalertification of one’s aptitude for driving or working as a driverollows them. The results of subjective methods should then beupplemented with objective methods.

and Prevention 70 (2014) 293–300 299

6. Limitations

Examinations in which drivers are aware of the consequencesof poor results are bound to be stressful. Professional drivers andnonprofessional drivers “with crashes” in this study were stronglymotivated to make the best possible self-presentation. This canmake the results unreliable. Results could be more reliable if driverstook the same tests in a safe environment, i.e., an environmentfree from stress that the results would affect their certification ofaptitude or lack of aptitude for driving or working as a driver. How-ever, due to the regulations of Poland’s Data Protection Law, it wasimpossible to contact drivers who had caused fatal road crashes toask them to participate in a study like that.

This study was retrospective. Therefore, it is worth consider-ing possible changes in temperament, especially on Neuroticism,Venturesomeness and Impulsiveness dimensions in Group III, due totraumatic events like being an offender in a fatal road crash. Thesechanges could be a significant factor influencing the results of thequestionnaires.

Acknowledgements

The authors thank the drivers who participated in the examina-tions and the psychologists who administered the tests.

This study was prepared on the basis of the results of a researchtask carried out within the scope of the second stage of theNational Programme “Improvement of safety and working condi-tions” partly supported in 2011–2013 — within the scope of stateservices — by the Ministry of Labour and Social Policy. The CentralInstitute for Labour Protection – National Research Institute wasthe Programme’s main co-ordinator.

References

Anastasi, A., Urbina, S., 1999. Testy psychologiczne (Psychological Testing]). Pra-cownia Testów Psychologicznych Polskiego Towarzystwa Psychologicznego,Warszawa.

2003. Aneks nr 1 do Metodyki psychologicznych badan kierowców (Annex No. 1 tothe Methodology of Psychological Examination of Drivers). Instytut TransportuSamochodowego, Zakład Psychologii Transportu Drogowego, Warszawa.

Arthur Jr., W., Graziano, W.G., 1996. The five-factor model, conscientiousness anddriving accident involvement. J. Pers. 64 (3), 593–618.

Barric, M.R., Mount, M.K., 1991. The Big Five personality dimensions and job perfor-mance: a meta-analysis. Pers. Psychol. 44 (1), 1–26.

Clarke, S., Robertson, I.T., 2005. A meta-analytic review of the Big Five personality fac-tors and accident involvement in occupational and non-occupational settings. J.Occup. Organ. Psychol. 78, 355–376.

Dorn, L., Stephen, L., af Wåhlberg, A., Gandolfi, J., 2010. Development and vali-dation of a self-report measure of bus driver behaviour. Ergonomics 53 (12),1420–1433.

2007–2012. EU Strategy, Available at: http://ec.europa.eu/social/main.jsp?catId=151(retrieved 14.02.13).

Eysenck, H.J., Eysenck, S.B.G., 2006. Manual of the Eysenck Personality Scales (EPSAdult) including EPQ-Revised (EPQ-R), EPQ-R Short Scale, Impulsiveness (IVE)Questionnaire. Hodder and Stoughton Educational, London, UK.

Friedman, H.S., Booth-Kewley, S., 1987. The disease-prone personality: a meta-analytic view of the construct. Am. Psychol. 42, 539–555.

Guion, R.M., Gottier, R.F., 1965. Validity of personality measures in personnel selec-tion. Pers. Psychol. 18, 135–164.

Ivers, R., Senserrick, T., Boufous, S., Stevenson, M., Chen, H.Y., Woodward, M., Norton,R., 2009. Novice drivers’ risky driving behavior, risk perception, and crash risk:findings from the DRIVE study. Am. J. Public Health 99 (9), 1638–1644.

Jaworowska, A., 2011a. Kwestionariusze osobowosci Eysencka. EPQ-R. EPQ-R wwersji skróconej. Polskie normalizacje (Eysenck’s Personality Questionnaires.EPQ-R. EPQ-R Short Scale. A Polish standardisation). Pracownia Testów Psycho-logicznych Polskiego Towarzystwa Psychologicznego, Warszawa.

Jaworowska, A., 2011b. Kwestionariusz Impulsywnosci IVE. Impulsywnosc,Skłonnosc do ryzyka, Empatia. Polska normalizacja (IVE Impulsiveness Ques-tionnaire. Impulsiveness, Venturesomeness, Empathy. A Polish standardisation).Pracownia Testów Psychologicznych Polskiego Towarzystwa Psychologicznego,

Warszawa.

Jazdzik-Osmólska, A., 2012. Metodologia i wycena kosztów wypadków drogowychna sieci dróg w Polsce na koniec 2011 r (Methodology and Assessmentof the Cost of Road Accidents on the Network of Poland’s Roads as ofthe End of 2011). Instytut Badawczy Dróg i Mostów, Warszawa, Available

3 alysis

K

L

L

M

N

N

O

Ö

P

P

S

1

00 A. Łuczak, A. Tarnowski / Accident An

at: http://www.krbrd.gov.pl/download/pdf/koszty wypadkow drogowych nasieci drog w Polsce na koniec roku 2011.pdf (retrieved 05.06.13).

arlberg, L., Undén, A.L., Elofsson, S., Krakau, I., 1998. Is there a connection betweencar accidents, near accidents, and Type A drivers? Behav. Med. 24 (3), 99–106.

ajunen, T., 2001. Personality and accident liability: are extraversion, neuroticismand psychoticism related to traffic and occupational fatalities? Pers. Individ.Differ. 31 (8), 1365–1373.

i, F., Li, C., Long, Y., Zhang, C., Hennessy, D., 2004. Reliability and validity of aggressivedriving measures in China. Traffic Inj. Prev. 5 (4), 349–355.

ontag, I., Comrey, A.L., 1987. Internality and externality as correlates of involve-ment in fatal driving accidents. J. Appl. Psychol. 72, 339–343.

abi, H., Consoli, S.M., Chastang, J.F., Chiron, M., Lafont, S., Lagarde, E., 2005. Type Abehavior pattern, risky driving behaviors, and serious road traffic accidents: aprospective study of the GAZEL cohort. Am. J. Epidemiol. 161 (9), 864–870.

abi, H., Consoli, S.M., Chiron, M., Lafont, S., Chastang, J.F., Zins, M., Lagarde, E.,2006. Aggressive/hostile personality traits and injury accidents: an eight-yearprospective study of a large cohort of French employees – the GAZEL cohort.Psychol. Med. 36 (3), 365–373.

wsley, C., McGwin Jr., G., McNeal, S.F., 2003. Impact of impulsiveness, venture-someness and empathy on driving by older adults. J. Safe. Res. 34, 353–359.

zkan, T., Lajunen, T., 2005. Multidimensional Traffic Locus of Control Scale (T-LOC):factor structure and relationship to risky driving. Pers. Individ. Differ. 38 (3),533–545.

asquier, D., Mazilescu, C.A., 2009. Predictive validity of a personality questionnairebased on the Five Factors Model: subjacent mediations? Cogn. Brain Behav.Interdiscip. J. 13 (2), 165–177.

atil, S.M., Shope, J.T., Raghunathan, T.E., Bingham, C.R., 2006. The role of personalitycharacteristics in young adult driving. Traffic Inj. Prev. 7 (4), 328–334.

anval, N., Chatteriee, S., Dasqupta, M., Chatteriee, S., 2012. Aggression and related

psychosocial correlates among offending and non-offending drivers: a thera-peutic intervention. SIS J. Proj. Psychol. Ment. Health 19, 97–106.

999. Standards for Educational and Psychological Testing. American EducationalResearch Association, American Psychological Association, National Council onMeasurement in Education, US.

and Prevention 70 (2014) 293–300

Strelau, J., Zawadzki, B., 1993. The Formal Characteristics of Behaviour – Tempera-ment Inventory (FCB-TI): theoretical assumptions and scale construction. Eur. J.Pers. 7, 313–336.

Strelau, J., 1998. Psychologia temperamentu (Psychology of Temperament).Wydawnictwo Naukowe PWN, Warszawa.

Strelau, J., Jaworowska, A., Wrzesniewski, K., Szczepaniak, P., 2005. Kwestionariuszradzenia sobie w sytuacjach stresowych CISS. Podrecznik (CISS Stress Hand-ling Questionnaire. A Manual). Pracownia Testów Psychologicznych PolskiegoTowarzystwa Psychologicznego, Warszawa.

Sümer, N., 2003. Personality and behavioral predictors on traffic accidents: testinga contextual mediated model. Accid. Anal. Prev. 35, 949–964.

Sümer, N., Lajunen, T., Özkan, T., 2005. Big Five personality factors as the distalpredictors of road accident involvement. In: Underwood, G. (Ed.), Traffic andTransport Psychology. Theory and Application. Proceedings of the ICTTP 2004.Elsevier, Amsterdam, pp. 215–227.

Taubman-Ben-Ari, O., Yehiel, D., 2012. Driving styles and their associations withpersonality and motivation. Accid. Anal. Prev. 45, 416–422.

Tett, R.P., Jackson, D.N., Rothstein, M., 1991. Personality measures as predictors ofjob performance: a meta-analytic review. Pers. Psychol. 44, 703–742.

Tillman, W.A., Hobbs, G.E., 1949. The accident-prone driver, a study of the psychiatricand social background. Am. J. Psychiatry 106, 321–331.

Trzcinska, M., (Unpublished MA thesis) 1996. Wpływ temperamentu i stresu na stanzdrowia. Adaptacja Inwentarza Zdrowia GHQ (General Heath Questionnaire –D. Goldberga) (The Impact of Temperament and Stress on Health. An Adapta-tion of D. Goldberg’s General Health Questionnaire). Uniwersytet Warszawski,Warszawa.

Ustawa z dnia 5 stycznia 2011 r. o kierujacych pojazdami (Motor Vehicle DriversAct of 5 January 2011). Dziennik Ustaw (J. Laws) No. 30, item 151.

Waszkowska, M., 2009. Temperament a poczucie stresu w ruchu drogowym (Tem-

perament and perceived stress in road traffic). Med. Pr. 60 (2), 137–144.

Zielinska, A., 2013. Przełomowy rok 2012. Analiza danych statystycznych o wypad-kach drogowych w Polsce i Unii Europejskiej (Analysis of statistical data onroad accidents in Poland and the European Union). Bezpieczenstwo Ruchu Dro-gowego 1, 3–9.