Cross-cultural comparisons of traffic safety,risk perception,attitudes and behaviour
Ingunn Olea Lund,Torbjörn Rundmo *
Norwegian University of Science and Technology,Department of Psychology,7491Trondheim,Norway
a r t i c l e i n f o Keywords:Traffic safety Attitudes
Risk perception
Cross-cultural comparison
a b s t r a c t
The core aim of the prent study is to examine cultural differences in risk perception and attitudes towards traffic safety and risk,taking behaviour in the Norwegian and the Ghanaian public.An additional aim is to discuss the applicability of various traffic measures,suited for low and middle income countries in Africa.
Sample:The results of the prent study are bad on two lf-completion questionnaire surveys car-
ried out in February and March 2006.The first was a reprentative sample of the Norwegian public above 18years of age (N =247).The cond was a stratified sample of Ghanaian respondents (N =299).In Ghana the data was collected in Accra and Cape Coast.
The results showed that there is potential for further improvement of safety attitudes and risk behav-iour among Ghanaians as well as Norwegians.There were also differences in the respondents’evaluation of attitudes,risk perception and behaviour.Perceived risk and attitudes also significantly predicted risk behaviour and accidents/collisions.The implications of the results for traffic safety will be discusd.
Ó2008Elvier Ltd.All rights rerved.
1.Introduction
Traffic accidents are a major cau of fatalities world wide.In Norway,about 220people are killed in traffic every year,and more than 10000are injured (The Norwegian Directorate of Public Roads,2006).In Ghana,traffic accidents are an even greater prob-lem.The fatality rate per 10000is about 30times higher in devel-oping countries than in high-income countries (WHO,2003).In 1998,low-income countries accounted for about 85%of all traffic fatalities globally.Furthermore,96%o
f all road traffic fatalities that involve children occur in developing countries (Nantuyla and Reich,2002).The kind of traffic accidents that occur in developing countries differ from tho in high-income countries.In high-in-come countries,fatalities related to traffic accidents often involve the driver of a car.However,in developing countries many pedes-trians and pasngers in public transportation are killed.In addi-tion to causing deaths and suffering,road traffic accidents are a great burden on the countries’economy.
There are veral approaches to reducing traffic accidents.Some examples are enforcing stricter traffic regulations,improvement of roads and environment and education.However,it has been ar-gued that improvement of roads results in a reduction of traffic risk perception (Summala,1996).Improvements of roads and cars may actually cau an increa in traffic accidents becau drivers in-crea speed and are less careful.Human error may account for about 90%of all traffic accidents (Rumar,1985).It is reasonable
to assume that it would be effective to apply psychological inter-ventions in traffic safety campaigns to reduce traffic accidents.Rundmo (1999)argues that if risk perception affects behaviour,then it would be possible to change behavioural patterns through influence of risk perception.It would be interesting to examine whether similar strategies to reduce traffic accidents could effi-ciently be appli
ed cross-culturally,or whether new strategies should be developed and tested locally.It is reasonable to assume that attitudes towards and risk perception of traffic differs in dif-ferent cultures.It is also possible that there are differences across cultures in how important different factors are for behaviour.考虑英文
Studies on prevention of traffic accidents in different cultures aim to find which preventive strategy or combination of strategies are most effective in the specific cultures.Safety campaigns aimed to improve people’s attitudes towards traffic safety have been shown to be successful in high-income countries.However,Lund (2006)found that attitudes were not successful predictors of driver behaviour in Ghana.It is well recognized that prediction of behav-iour is difficult and ought to be bad on veral factors (Hines et al.,1986).In addition to attitudes,the results of previous studies indicate that risk perception is related to risk behaviour (Rundmo,1999).Rundmo (1995)found that attitudes and risk perception were related to each other,as well as behaviour.Accordingly,it would be of interest to examine whether or not there are cross-cul-tural differences in the associations between traffic safety attitudes and risk behaviour in traffic in a Norwegian and a Ghanaian public.It is important to examine whether this approach is effective in developing countries before safety campaigns using this approach are applied.In the prent study,Norway is a high-income country
0925-7535/$-e front matter Ó2008Elvier Ltd.All rights rerved.doi:10.1016/j.ssci.2008.07.008
*Corresponding author.
E-mail address:u.no (T.Rundmo).
Safety Science 47(2009)
547–553
Contents lists available at ScienceDirect
Safety Science
j o u r n a l ho m e p a g e :w w w.e l vier.c
om/locate/ssci
and Ghana reprents a low-income African country.The coun-tries were chon for analysis since they hold characteristics con-sidered typical for high-income and African countries.The majority of rearch on traffic safety has been conducted in high-income countries,respectively.A very few studies have examined the effectiveness of countermeasures directed at improving driver behaviour in African countries.
Norway has adopted the‘‘zero vision policy”stating that no rious or fatal road traffic accidents are acceptable.The zero vi-sion policy assumes that both authorities and individuals in traffic share the responsibility for traffic safety and have a mutual obliga-tion to each other(The Norwegian Directorate of Public Roads,The Norwegian Police Directorate&the Norwegian Society for Road Safe
ty,2002).There is a long way to go however,before the goals of the zero vision policy are achieved.In Ghana,a low-income country with a population of about22million,40%live below the poverty line.Adult literacy rate is54%(WHO,2006).There are only about300000vehicles in the country and about 1500000hold a driver licen.Exact numbers are not available (Swedish National Road and Transport Rearch Institute,2000). Although focus on traffic safety has incread in Ghana and better roads are built,there has been a lack of focus on human factors in traffic.However,as mentioned previously,safety gains are often reduced or even lost,since people often increa speed and de-crea attention when road conditions are improved(Summala, 1996).
In50%of all road traffic accidents in Ghana speeding has been the main cau of the accident(Afukaar et al.,2003).Further-more,there is a lack of widespread formalid driver education and drivers often fail to adhere to road signs.A major threat to traffic safety in Ghana is the‘‘tro-tros”(minibus),which provides unofficial public transport.The drivers minimi stopping times and compete against each other to pick up pasngers.The ‘‘tro-tros”are often involved in rious traffic accidents.The cars ud in this type of transportation are usually in bad condition and rebuilt to carry a greater number of pasngers.Accidents involving‘‘tro-tros”often occur on roads passing through rural areas.In urban areas,pedestrians are at great risk of being in-volved in traffic accidents.Pedestri
ans account for almost50% of all road fatalities in Ghana.Men,children and people between 26and45years are most at risk of being run down by a car (Afukaar et al.,2003).
We know that attitudes are great predictors of behaviour in high-income countries(Ajzen and Fishbein,1980;Ivern and Rundmo,2004).Conquently it has also been suggested that atti-tudes indirectly affect involvement in traffic accidents(West et al., 1993).Rundmo(1995,1997)found that safety attitudes were re-lated to both risk perception and behaviour.Results from Jonah (1986)suggest that risk-taking behaviour in traffic was the most important variable for involvement in traffic accidents.Accord-ingly,knowledge of how people perceive and understand risks re-lated to traffic is of importance in order to understand differences in driver behaviour.Studies carried out previously have given sup-port to the idea that traffic safety campaigns aimed to improve dri-ver behaviour would benefit from involving psychological knowledge of how individuals perceive risks related to traffic (Oltedal et al.,2004).The prent study includes veral measures of risk perception:measures of traffic risk perception,risk nsitiv-ity and risk willingness to examine whether risk perception is an effective predictor of driver behaviour.Risk perception includes perception of the probability of being involved in an accident as well as verity of conquences if an accident should take place. Risk nsitivity is the tendency to perceive all or veral risks sources
to be large and involves a tendency to focus attention in risks.Risk willingness is defined as how willing individuals are to take risks in traffic and in general.1.1.Risk perception
Risk perception is the subjective asssment of the probability of experiencing a negative event.Examining risk perception in relation to traffic safety is of interest becau it may be related to people’s driver behaviour.According to social cognition models the probability of engaging in healthy behaviour increas if the risk is perceived to be large(Rundmo,1999).Perception of risk is a social and cultural construct that reflects the values,symbols, history and ideology of people living in different cultures (Weinstein,1980).Despite their importance,psychological pro-cess related to individual risk judgements are not well under-stood(Price,2001).Individuals are often inaccurate judges of risks.Risk ratings are often biad by decision-making heuristics (Groeger and Chapman,1990).
Risk perception has traditionally been studied from a cognitive stance where perception is treated as a cognitive process.The psy-chometric paradigm(Slovic,1992)has been the dominant cogni-tive approach to understanding how people perceive risks. Several factors may have an effect on perceived risk.Familiarity with the risk source and perceived control over the situation are thought to have an effect on risk perception.Furthermore,risks from rare events such as aviation accidents tend
to be overesti-mated.More common risks such as traffic accidents or health haz-ards from smoking are often underestimated(Lichtenstein et al., 1978).Gleitman(1995)propod that perception of risks is re-duced in familiar situations due to habituation.By repeating behaviour or being expod to a risk veral times,the perception of the risks will decrea becau one gets familiar with it. Sjøberg(1999)found demands for risk reduction to be driven by perception of verity of conquences,not by perception of probability.
Deery(1999)defined risk perception related to traffic as the subjective interpretation involved in different traffic situations. Furthermore,higher subjective appraisal of traffic risk results in more protective behaviour in traffic such as obeying traffic regula-tions and wearing a atbelt.Accordingly,differences in driver behaviour may partly be due to differences in traffic risk percep-tion.However,if this were the ca,it would be reasonable to as-sume that,due to the high frequencies of accidents in African countries,drivers in developing countries perceive traffic risks to be lower than drivers in high-income industrialized countries.
1.2.Differences in risk perception due to gender,age and culture
Previous studies have suggested that people’s perception of risks is related to demographic variable
s such as gender and age. Most individuals tend to evaluate the probability of being involved in traffic accidents to be lower for themlves than for others.Ado-lescents and men show a tendency to be less nsitive to risks,to underestimate the probability and verity of risks caud by traf-fic situations compared to elderly people and females(Moe,1982; Spolander,1993;Parker et al.,1992;Slovic,1992;Sjøberg et al., 1996),and to perceive traffic situations less holistically(Deery, 1999).Trankle et al.(1990)found that younger males judged the risk of dangerous traffic situations as lower compared to older males.Furthermore,results from DeJoy(1992)showed that com-pared to males,females perceived a higher level of risk related to traffic.
Flynn et al.(1994)showed that,compared to males,females perceived risks of environmental health hazards to be considerably higher.Quimby(1998)concluded that adolescents were more fre-quently involved in traffic accidents than adults.Bad on thefind-ings from the literature reviewed above,the prent study expects tofind age and gender differences related to risk perception,risk nsitivity and risk willingness.
548I.O.Lund,T.Rundmo/Safety Science47(2009)547–553
1.3.Cross-cultural differences in risk perception
Studies carried out previously have shown cross-cultural differ-ences related to risk perception.Goszczynska et al.(1991)showed that there was greater variance in risk perception between coun-tries than between different regions in one country.This may be due to cultural differences related to how people perceive different risk sources.Differences in the tendency to rate risks as high or low might be related to the size of the country.Countries with a large population naturally experience and report more accidents than countries with fewer inhabitants.As a conquence,individuals living in countries with many residents often become more nsi-tive to risks due to the availability of heuristic caud by the great-er number of accidents(Goszczynska et al.,1991).
Slovic et al.(1982)suggested that risk perception is likely to vary between cultures depending on what the media cho to fo-cus on.Topics of discussion in everyday life,which cultural norms are en as important and differences in technical and legal oppor-tunities for control and regulation of risks in different cultures,af-fect how risks are perceived in a given culture.Wåhlberg and Sjøberg(2000)argued that media might have an effect on risk per-ception via availability.Information is thought to lead to stronger effects.Several replications of a study originally conducted by Slovic et al.(1980)have given valuable information about differ-ences in risk perception between countries.In t
he original study a questionnaire including90activities,substances and technolo-gies were handed out to175North American college students.Dif-ferent kinds of risks were included,for example,hand guns, nuclear reactors,chemical substances,aviation,traffic,smoking and home appliances.Respondents were asked to estimate how lethal each risk was for society.Risk was defined as risk of dying from the different hazards.Englander et al.(1986)found that Americans rated the risks from radiation and chemicals to be larger than Hungarians.The Hungarian respondents rated risks from everyday risks such as traffic,work and home-related danger and health risks as higher.The study concluded that Americans emed to be more worried about delayed effects of substances through failure of proper safety organization.
Teigen et al.,(1988)found that Norwegians perceived risks very similarly to Americans.However,Norwegians tended to perceive less risk on the majority of risk sources than Americans.However, Norwegians tended to be more concerned with drugs and narcot-ics.A replication carried out in France found that risks related to violence,high technology and chemicals were rated as more risky. The French public rated all the risk sources to be larger than the Hungarian public(Karpowicz-Lazreg and Mullet,1993).Keown (1989)also showed that there were cultural differences in risk per-ception when comparing a sample of residents in Hong Kong and Ame
ricans.However,respondents in the Hong Kong sample rated risk items such as heroine,caffeine,crime and bicycles as consider-ably higher than respondents in the North American sample.On the other hand,the North Americans evaluated risks from DNA re-arch and alcoholic beverages to be larger compared to the Hong Kong residents.
Hayakawa et al.(2000)found differences between Japane and American residents regarding risk perception of traffic accidents. Japane residents were found to evaluate the probability of traffic accidents to be larger.They also evaluated the verity of con-quences if an accident should take place to be more rious.In addition,they were more likely to think they were to blame for the accident.Rearch carried out previously has shown that ben-efits of cognitive and motivational tendencies to maintain positive illusions may apply only in some particular cultures(Heine and Lehman,1995).Residents of high-income countries have been found to evaluate personal outcomes more positively compared to residents of Asian countries.It is likely that such differences ex-ist also between African and high-income cultures.Kouabenean (1998)performed a study in the Ivory Coast on respondents with different religions,occupations and beliefs.Individuals were eval-uated on their fatalism and risk perception related to accidents. The results showed that fatalistic beliefs and mystical practices influenced people’s perception of accidents and led to more risk-taking in traffic an
d neglect of safety measures.Fatalistic beliefs may cau individuals to be resigned to risks becau they cannot do anything to reduce the risks(Sjøberg,1999).Nyland(1993), as well as Sjøberg et al.(1996),found that people with marginal status and poor living conditions tend to rate risks to be larger compared to other groups.
Individuals in developing countries are to a greater extent ex-pod to threats and must struggle for survival.Accordingly,they are likely to be more nsitive to risks.Flynn et al.(1994)showed that white males rated risks as lower compared to women and black males.This might indicate that there are smaller differences in risk perception among females and males with marginal status or living in developing countries.Due to the lack of studies examining whether risk perception related to traffic behaviour in developing countries it is difficult to generalizefindings from high-income countries.Accordingly,it is important to examine differences related to perception of traffic risks and whether risk measures work as a predictor of driver behaviour in Norway as well as Ghana.In light of the literature reviewed above,the prent study expects tofind differences between Ghanaian and Norwegian populations in risk perception of traffic related issues due to cultural differences.Ghana is a developing country,and people are more expod to risks in daily life and have to struggle for survival to a greater extent than people living in Norway. Accordingly,the prent study expects Ghanaians to rate risks in general as higher than Norwegians.
oh captain my captainSeveral studies have found a connection between reported dri-ver behaviour and involvement in traffic accidents.Hattaka et al. (1997)found lf related driver behaviour to be related to future accident involvement.This may give support to the idea that people who report less ideal attitudes and driver behaviour are more at risk of being involved in traffic accidents.The prent study aimed at examining whether there was a connection between reported dri-ver behaviour and involvement in traffic accidents in Ghana and Norway.It was expected that respondents who reported more risk behaviour in traffic had been involved in traffic accidents more fre-quently compared to respondents who reported safe behaviour.The specific aims of the prent study are the following:
Identify determinants of traffic risk perception,risk nsitivity and risk willingness in Norway and Ghana.
Examine differences in traffic risk perception,risk nsitivity and risk willingness due to gender and age in the two countries.
Investigate whether adding traffic risk perception,risk nsitiv-ity and risk willingness in addition to attitudes increas explains the variance in driver behaviour in Norway and Ghana.
Examine whether there is a relation between reported driver behaviour and accident history in the tw
o countries.
2.Methods
2.1.Sample and procedure
A lf-completion questionnaire survey was carried out in Norway and Ghana.In Norway a total of247respondents replied to the questionnaire.The respon rate was23%.Of the respon-dents145(59%)were females and101(41%)were males.Respon-dents’age ranged from18to90years(M=41.26years,SD=17.91).
I.O.Lund,T.Rundmo/Safety Science47(2009)547–553549
From the Norwegian sample220had a drivers’licen(90%).In Ghana a total of299respondents replied to the questionnaire and the respon rate was85%.One hundred and fortyfive(50%) females and143(50%)males were included in the Ghanaian sam-ple.The respondents’age varied between17and95years (M=31.39years,SD=12.16).Of the respondents99had a driv-ers’licen(34%).The data collection in Ghana was carried out in Accra and Cape Coast.Accra is the capital of Ghana and has about 2million inhabitants,whereas Cape Coast is a smaller town with abo
ut82thousand inhabitants.Supervisors,master students from the Norwegian university of Technology and Science,reprenta-tives and master students in Accra and Cape Coast were responsi-ble for gathering data in collaboration with the author of the prent study.Norwegian and Ghanaian students approached respondents in different areas of the two cities to make sure indi-viduals from different social layers were included in the sample. Individuals who did not write were interviewed.Individuals who did not speak English were interviewed by Ghanaian students. 2.2.Questionnaire
Attitude towards traffic safety was measured by a questionnaire previously validated and found suitable for measuring such atti-tudes(Ivern and Rundmo,2004).The scale ranged from‘‘strongly agree”to‘‘strongly disagree”.Traffic risk perception was measured by20items(Rundmo and Fuglem,2000).Twelve items measured probability of injury due to different traffic accidents.The respon
alternatives ranged from‘‘very high probability”to‘‘no probabil-ity”.Eight items measured perception of verity if an accident should take place.The respon alternatives ranged from‘‘very -vere/fatal”to‘‘minimal”.Risk nsitivity was measured by a total of33items.Seventeen items measured perceived probability of experiencing various risk sources.Sixteen items measured per-ceived proba
bility of experiencing a health injury due to various risks.A modified version of Fischoff et al.(1978)was ud for this measure.Respons were given on afive-point scale ranging from ‘‘very probable”to‘‘very low probability”.Risk willingness was measured by six items.The items were developed to measure how willing people were to take risks in various situations.An-swers were given on a scale ranging from‘‘most willing”to‘‘would not do it”.The respondents were also asked about their accident history.Three items were included.A total of24items were in-tended to measure driver behaviour(Ivern and Rundmo,2004). Respons to statements were made using afive-point evaluation scale.Respon alternatives ranged from‘‘very often”to‘‘never”.
2.3.Statistical analysis
To examine the dimensional structure of risk perception related to probability of traffic accidents and conquences of traffic acci-dents as well as risk nsitivity and risk willingness,principal com-ponents analysis with direct rotation was applied.Multivariate analys of variance(MANOVA)were applied to analy differ-ences in factor dimensions due to country,gender and age.Multi-ple block-regression using the enter method was applied tofind to what degree a model consisting of risk perception in addition to attitudes was able to predict behaviour in traffic.MANOVA analysis was carried out in order to e whether there was a relation be-tween reported driver behaviour and acci
dent history.
3.Results
3.1.Risk perception,risk nsitivity and risk willingness
The internal consistency in the indices measuring risk percep-tion,risk nsitivity and risk willingness was judged to be satisfac-tory(e Table1).The dimensional structure of the indices was examined previously(e Lund(2006)for further details).How-ever,for two of the dimensions the reliability was a bit lower than ideal.This is probably due to the fact that both the dimensions consisted of only two indicators.
3.2.Differences in risk perception,risk nsitivity and risk willingness
manifestdestinyTable2shows that there was an overall‘‘effect”of country on risk perception related to probability of traffic accidents(Wilks’k 0.88,p<0.001),as well as judgement of verity of conquences if an accident should take place(Wilks’k0.99,p<0.034).Risk n-sitivity(Wilks’k0.68,p<0.001)was highly significant.Ghanaians perceived the probability of being involved in a traffic accident as greater than Norwegians.Furthermore,Ghanaians judged the con-quences to be more vere compared to Nor
wegians.In general, Ghanaian respondents were found to be more nsitive to risks than Norwegian respondents.However,Norwegians were more nsitive to risks related to habits of health,such as not getting en-ough exerci,food,smoking and drinking alcohol.
After establishing differences between the two countries related to risk perception,the next step was to examine differences in risk perception related to age and gender parately for Norway and Ghana.Table3shows the result from four parate multivariate analys of variance(MANOVA)where risk perception,risk nsi-tivity and risk willingness were the dependent variables.Age and gender werefixed factors.In Norway,gender affected risk percep-Table1
Number of items,total inter-item correlation and Chronbach’s alpha for all risk measures
Measures Number of
items
Mean corrected item total
男生穿衣搭配技巧
correlation
Chronbachs’
a
Risk perception:probability of traffic accidents
Accidents due to car
traffic
感恩母爱50.620.817
Role in traffic40.750.882 Collision with a living
creature
20.400.554
Risk perception:conquences of traffic accidents
Accidents due to car
traffic
50.500.735 Collision with living
creature
20.360.518
Risk nsitivity
War30.810.906 Dias60.760.913 Leisure activities30.510.691 Health precautions50.480.721 Health habits40.590.781 Nature catastrophes50.820.818 Weapons and motor
vehicles
60.560.806
Risk willingness
In traffic40.480.652
In general20.740.887
Table2
Multivariate analysis of variance for dimensions of risk perception of probability and conquences of traffic accidents,risk nsitivity and risk willingness by country
Wilks k F d f p Country Risk perception,probability0.8821.1830.0000* Risk perception,conquences0.99 3.4020.034**
Risk nsitivity0.6827.8570.000*
Risk willingness0.99 1.5820.207 *p<0.001.
**p<0.05.
550I.O.Lund,T.Rundmo/Safety Science47(2009)547–553
tion of conquences of traffic accidents(Wilks’k0.95,p<0.01) and risk nsitivity(Wilks’k0.92,p<0.05).Females perceived risks to be larger on both dimensions compared to males.The gen-der differences related to perceived risk were insignificant in Ghana.In Norway,age contributed to significant differences in risk nsitivity(Wilks’k0.70,p<0.001)and risk willingness(Wilks’k 0.81,p<0.001).Adolescents were least nsitive to risks and most willing to take risks.In Ghana,there were differences in risk per-ception of conquences of traffic accidents(Wilks’k
0.96, p<0.05),risk nsitivity(Wilks’k0.86,p<0.01)and risk willing-ness(Wilks’k0.93,p<0.001)due to age.Ghanaian adolescents perceived conquences from traffic accidents to be smaller,they were less nsitive to risks in general and were more willing to take risks compared to young and old adults.
3.3.Attitudes,risk perception,risk nsitivity and risk willingness related to safe driver behaviour
Table4shown that we were able to explain53%of the total var-
kush
iance in risk behaviour in traffic in the Norwegian sample.The identicalfigure for the Ghanaian sample was19%.Attitudes to-wards traffic safety,risk willingness as well as age and gender were the significant predictor variables when applying the Norwegian data in the analys.Perceived risk and risk nsitivity were found to have a modest effect in both the analys.However,in Ghana attitudes as well as the other predictor variables emed to have
a modest association with risk behaviour in traffic.
3.4.Driver behaviour and involvement in traffic accidents
In Norway there was a significant relation between reported driver behaviour and involvement in traf
fic accidents(e Table 5).As can be en no such connection was found in Ghana.How-ever,frequency analysis showed that38%of Ghanaian drivers had been involved in traffic accidents where they were injured, 46%had been involved in traffic accidents where others were in-jured and78%with vehicle damage.In comparison,only18%of the Norwegian drivers had been involved in traffic accidents where they were injured,27%had been involved in accidents where oth-ers were injured and42%with vehicle damage.4.Discussion
The results of this study found differences between Norway and Ghana related to traffic risk perception,risk nsitivity and risk willingness.This supports Goszczynska et al.(1991)who sug-gested that there were major differences between countries in per-ception of risks.Ghanaians were expod to a more hazardous traffic environment than Norwegians.Conquently,traffic accidents have been the most frequent cau of injury related fatal-ities in this country(Afukaar et al.,2003).The results showed that Ghanaians perceived the probability of being involved in traffic accidents to be larger than Norwegians.The verity of con-quences if an accident should happen was also considered to be greater by Ghanaians.Ghana has aboutfive times as many inhab-itants as Norway and naturally more accidents are reported.This is a possible explanation to why Ghanaians are more nsitive to traf-fic risks than Norwegians.Furthermore,in Norway,most fatalities due to traf
fic accidents involve the driver of a motor vehicle.In Ghana the number of cars increas fast.Many victims of traffic accidents are pedestrians or pasngers.Accordingly,Ghanaian residents may perceive traffic risks to be larger becau they have
gstarTable3
Multivariate analysis of variance for dimensions of risk perception of probability and conquences of traffic accidents,risk nsitivity and risk willingness by gender and age
Wilks
k
F df p
Gender Norway Risk perception,probability0.96 2.4330.066 Ghana0.97 1.8230.144
Norway Risk perception,
conquence 0.95 5.9620.003**
Ghana0.990.1120.894 Norway Risk nsitivity0.92 2.5670.015*** Ghana0.980.5270.820 Norway Risk willingness0.98 1.4120.246 Ghana0.990.3820.680 Norway Risk perception,probability0.980.6460.646 Ghana0.95 1.9560.070
Norway Risk perception,
conquence 0.96 2.2940.058
Age Ghana0.96 2.4340.047*** Norway Risk nsitivity0.70 5.71140.000*
Ghana0.86 2.14140.010*
Norway Risk willingness0.8112.8140.000*
Ghana0.93 4.7540.001*
*p<0.001.
**p<0.01.
***p<0.05.Table4
Predictors of driver behaviour in Norway and Ghana
Variables entered Norway
Beta
F-
change
Ghana
Beta
F-
change Attitudes30.48** 1.01 Attitudes towards rule violations0.4860.050
Attitudes towards responsibility in
traffic
0.0120.098
Attitudes towards drinking and
driving
0.2320.185脸盆
Risk Perception,probability of traffic
accidents
1.79 1.98 Collision with car/car running off the
road
À0.453À0.156
Role in traffic0.008À0.079
Collision with a living creatureÀ0.070À0.031
Risk perception,conquences of
traffic accidents
À0.071À0.247
Collision with a car/car running off the
road
À0.0400.010
Collision with a living creature
Risk willingness19.82**0.33
In traffic0.3680.074
In general0.045À0.114
operation是什么意思Risk nsitivity 1.4 1.17 War0.0180.364
DiasÀ0.119À0.053
Leisure activities0.0370.133
Health precautionsÀ0.157À0.099
Habits of health0.106À0.075
Nature catastrophes0.097À0.398
Weapons and motor vehiclesÀ0.057À0.217
Age groups0.106 3.76*0.045 1.60 GenderÀ0.1600.234
Adjusted R2=0.49(Norway),0.10(Ghana).
*p<0.05.展望未来英语
**p<0.001.
Table5
Multivariate analysis of variance for involvement in accidents with person injury and/ or vehicle damage by driver behaviour
Country Wilks k F df p Driver behaviour Norway0.313 1.381080.017*
Ghana0.2820.99930.517 *p<0.05.
I.O.Lund,T.Rundmo/Safety Science47(2009)547–553551
