ContentslistsavailableatScienceDirect
Accident
Analysis
and
Prevention
j o u r n a l ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / a a p
A
conceptual
framework
for
road
safety
and
mobility
applied
to
cycling
safety
Paul
Schepers
a,b,∗,
Marjan
Hagenzieker
b,c,
Rob
Methorst
a,
Bert
van
Wee
d,
Fred
Wegman
b,caMinistryofInfrastructureandtheEnvironment,TheNetherlands bSWOVInstituteforRoadSafetyResearch,TheNetherlands
cDelftUniversityofTechnology,CivilEngineeringandGeosciences,TheNetherlands dDelftUniversityofTechnology,TransportandLogisticsGroup,TheNetherlands
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:Received7December2012
Receivedinrevisedform17February2013 Accepted12March2013 Keywords: Roadsafety Bicycle Cyclingsafety Landuse Infrastructure Models Conceptualmodel Framework
a
b
s
t
r
a
c
t
Scientificliteraturelacksamodelwhichcombinesexposuretorisk,risk,andtherelationshipbetween
them.Thispaperpresentsaconceptualroadsafetyframeworkcomprisingmutuallyinteractingfactors
forexposuretoriskresultingfromtravelbehaviour(volumes,modalsplit,anddistributionoftraffic
overtimeandspace)andforrisk(crashandinjuryrisk).Theframework’sthreedeterminantsfortravel
behaviourarelocationsofactivities;resistances(generalizedtransportcosts);needs,opportunities,and
abilities.Crashandinjuryrisksaremodelledbythethree‘safetypillars’:infrastructure,roadusersand
thevehiclestheyuse.Creatingalinkintheframeworkbetweenriskandexposureisimportantbecause
ofthe‘non-linearrelationship’betweenthem,i.e.risktendstodecreaseasexposureincreases.
Further-more,‘perceived’risk(atypeoftravelresistance)playsaroleinmodechoice,i.e.theperceptionthata
certaintypeofvehicleisunsafecanbeadeterrenttoitsuse.Thispaperusestheoriestoexplainhow
theelementsinthemodelinteract.Cyclingisanareawheregovernmentstypicallyhavegoalsforboth
mobilityandsafety.Toexemplifyapplicationofthemodel,thepaperusestheframeworktolinkresearch
oncycling(safety)tolanduseandinfrastructure.Themodel’svalueliesinitsabilitytoidentifypotential
consequencesofmeasuresandpoliciesforbothexposureandrisk.Thisisimportantfromascientific
perspectiveandforpolicymakerswhooftenhaveobjectivesforbothmobilityandsafety.
© 2013 Elsevier Ltd. All rights reserved.
1. Introduction
Thispaperintroducesaconceptualframeworkforroadsafety andmobilityandappliesittocyclingsafetytoexemplifyits appli-cation.Oneofthemajorproblemsofroadsafetyresearchisthat mostof itdoesnot havea strongtheoreticalbasis. The lackof suchabasismakesitdifficulttodesignsuitablestudiesand inter-pretfindings(Elvik,2004).Currentroadsafetymodelsarefocused mainlyonrisk.Trafficandtransportliteratureoffersmodelsfor travelbehaviourthathelptoexplainexposuretorisk.Tothebest ofourknowledge,thereisnoframeworkthatcombinesexposureto risk(resultingfromtravelbehaviour)andriskincurrentscientific literature.Suchaframeworkwouldbeusefulforbothroadsafety researchersandpolicymakersforidentifyingthepotentialeffects ofmeasuresandpolicies.Becauseroadcrashesresultfromboth exposuretorisk(hereafterreferredtoasexposure)andrisk,amodel
∗ Correspondingauthorat:MinistryofInfrastructureandtheEnvironment,The Netherlands.Tel.:+31887982457.
E-mailaddress:paul.schepers@rws.nl(P.Schepers).
comprisingbothfactorsaswellastheinteractionsbetweenthem wouldhelpresearchersacquireabroaderinsightintopotentially relevantsafetyeffects.Thispaperpresentsaconceptualframework forroadsafetyincorporatingfactorsfordeterminingexposureand risk,andtherelationshipbetweenthesetwo.
Toexplaintheframework anditsusefulness,it isappliedto thequestion of how cycling safety is affectedby land useand infrastructure characteristics(roadnetworks,roadsections, and intersections).Cyclingisanareawheregovernmentstypicallyhave targetsforbothmobilityandroadsafety.Governmentspromote cyclingasitisanenvironmentallysustainablemodeoftransport andisassociatedwithpublichealthbenefits(seee.g.DeHartog etal.,2010;Heinenetal.,2011).Adaptinglanduseand infrastruc-tureis ameansfor governments toimprovecycling safetyand increasebicycleuse.However,researchregardingcyclist mobil-ityandcyclistsafetyisnotyetwellconnected.Therearereview studiesthatdescribehowthebuiltenvironmentaffectsbicycleuse (e.g.Heinenetal.,2010)andhowroadfactorsaffectcyclingsafety (e.g.Reynoldsetal.,2009),butonlyfewlinkbothfactors.Thislink isimportantforpolicymakersandresearch,becauseignoringone ofthetwofactors,ortheinteractionbetweenthem,mightleadto 0001-4575/$–seefrontmatter © 2013 Elsevier Ltd. All rights reserved.
anover-orunderestimationofthesafetyeffectsofcandidatepolicy options,andresultin‘erroneous’policies.
Section2 ofthis paper describestheconceptualroad safety framework,andSection3itsapplicationtocycling(safety)andits linkwithlanduseandinfrastructurecharacteristics.Bothsections includesubsectionsforrisk,travelbehaviourandtheirinteraction. Section4exemplifiesthemodel’sapplicationbyapplyingittothe low-costmeasureofconvertingone-waytotwo-waycycletracks, withtheresultsdiscussedinSection5.
2. Aconceptualroadsafetyframework
Consistentwith Asmussenand Kranenburg (1982),our con-ceptualframeworkcontainsfactorsdeterminingexposuretorisk (resultingfromtravel behaviour),crashrisk,and injury risk(or injuryseverity).It combinesVanWee’s(2009)passenger trans-portmodelforexposuretoriskwiththemodelofthethreetraffic safetypillarsforrisk(Othmanetal.,2009)(seeFig.1).This sec-tionintroducestheconceptualroad safetyframework. Afteran explanationof thedecisions underlyingits development, three subsections focus on travel behaviour (Section 2.1.1), and risk (Section2.1.3), andtherelationshipbetweenexposureand risk (Section2.1.2). Section2.1.4 describes thedemarcation needed to reduce the framework’s complexity. Section 2.2 uses theo-riesandconceptstoexplain theinteractionoftheframework’s elements.
Aconceptualframeworkormodelisanabstractionor simplifi-cationofrealitytohelpusbetterunderstandrealworldsystems, facilitatecommunication and integrateknowledge across disci-plines(Heemskerketal.,2003;Ford,2009).Thesegoalsarebest servedbymodelswithalimitednumberoffactors,suchastheVan Wee(2009)modelandthethreetrafficsafetypillars(Othmanetal., 2009).Bothmodelshavemutuallyinteractingfactors,recognizing thataccidentscanresultfromcombinationsofinteracting vari-ables.Incontrast,crash-phasemodelssuchasHeinrich’sDomino Theory(Heinrich,1931)assumethataccidentsresultfromaseries ofeventsorcircumstancesandarethuspreventableby eliminat-ingone ofthecausesin thelinearsequence. AccordingtoToft etal. (2012),because accidentsoftenresult fromcombinations ofmutuallyinteractingvariables,modellingapproachesforcrash researchneedtoshiftfromlinear models(suchas crashphase models)tonon-linearmodels(suchasthemodeldescribedinthis paper).
Peopleareexposed toriskin trafficbecausetheytravel and becausetherearedangerspresentintraffic.Asyet,wehavenot yetmanagedtoachievedanger-freetravel.Themeasuresusedin theroadsafetyliteratureforexposuretoriskaredirectlylinked totravelbehaviour,e.g.kilometrestravelledandAnnualAverage DailyTraffic(AADT) (whatthebest measureis dependsonthe issuebeingstudied,seeHakkertandBraimaister, 2002). There-fore,travelbehaviourandexposuretoriskhavebeencombined intheframeworkinonebox.Similarly,crashandinjuryriskare putinoneboxalthough botharegenerallyacceptedasdistinct dimensionsoftheroadsafetyproblem(Rumar,1999).Thisisdone becausethelinkstootherelementsinthemodelaresimilar,and itreducesthemodel’scomplexity.Themodeldoesnotincludethe post-crashphasein which,forinstance,theemergency medical systemisrelevanttotheinjuryrisk.Separateboxesforcrashand injuryriskwouldhavetobeinsertedifelementsrelatingtothe post-crashphaseweretobeincludedinthemodel.Themodelis notachronologicallyorganizedcrash-phasemodel,butthere is orderinthesensethattraveldecisionstakenbeforetraffic par-ticipation (the focus of the upper part of the model) result in exposuretoriskduringtrafficparticipation(thelowerpartofthe model).
2.1. Descriptionoftheframework 2.1.1. Travelbehaviour
Travelbehaviourliteraturecommonly distinguishesbetween trafficvolumes,modalsplitand distributionoftrafficover time andspace(VanWee,2009).VanWee(2009)developedamodel forpassengertransportthatcontainselementsdeterminingtravel behaviour:locationsofactivities,transportresistances (general-izedtransportcosts),andneeds,opportunitiesandabilities.People travelbetweenLocationsofactivitiestoperformactivitiessuchas living,working,andshopping.Traveltakesmoneyandtimeand incurs non-monetary’costs suchas discomfort,which together makeupTravelresistance.Perceivedrisk,whichisalsoatypeof resistance,ismodelledexplicitlybyanarrowfromRisktoTravel resistance.Besideslocationsandtravelresistance,travelbehaviour is also affected by Needs, opportunities, and abilities (NOA); for instance theneed for active travel,the possessionof a driving licenseandcar,orthephysicalfitnessneededtowalkandcycle.All threecategories(locations,resistance,andNOA)areinfluentialin alldirections.Travelbehaviourdecisionssumuptotrafficvolumes, modalsplit,andthedistributionoftrafficovertimeandspace(Van WeeandMaat,2003).Traveldecisionstakenbyindividualsbefore trafficparticipationhave alsobeencalled‘strategic andlifestyle decisions’(Michon,1985;Hatakkaetal.,1999):modechoiceand movingtoanewhome,etc.Thesedecisionsresultinexposureto riskduringtrafficparticipation.Behaviourduringtraffic partici-pationhasbeendescribedastacticaland operationalbehaviour (Michon,1985).
2.1.2. Thelinkbetweenexposureandrisk
Themodelcomprisesanarrow fromExposuretorisktoRisk, becauseexposureaffectsrisk.Mostempiricalstudiesshowthatrisk decreasesasexposureincreases(Elvik,2009).AnarrowfromRisk toExposuretoriskisincludedtoindicatethattrafficparticipants areexposedtorisksonlytotheextentthatrisksarepresent.The modelalsoincludesafeedbackloopfromRisktoTravelresistance. Riskmayaffectperceivedriskwhich,inturn,cancausetravellers toshifttoothermodesorevenavoidtrips(Heinenetal.,2010;Van Weeetal.,inpress).
2.1.3. Risk
Crash risk resultsfrom interaction betweenthree elements, sometimes called the ‘three traffic safety pillars’: road user(s), vehicle(s), and infrastructure (e.g. Othman et al., 2009). Simi-larly,epidemiologistsusethetermshost,agent,andenvironment (Haddon, 1980). Note that Haddon’sdefinition of environment alsoincludesthesocialenvironment.Single-vehiclecrashesmay involveonly one vehicleand one roaduser,whereas ‘conflicts’ involvean interactionbetween severalvehicles androad users (foramoredetailedmodelthatincludestheinteractionbetween roadusers,seeHoutenbos,2008).Dependingontheenergythatis exchangedbetweenroadusers,vehiclesandinfrastructure,crashes mayresultininjurieswithvaryinglevelsofseverity.Crashesmay befatalwhenforcestransferredtovictimsexceedtheir biomechan-icaltolerance.Thistolerancedependsonage,healthstatus,stature andothercharacteristicsofroadusersinvolvedinacrash(Corben etal.,2004).Theframeworkprovidesfortwo-wayarrowsbetween RiskontheonehandandInfrastructure,VehiclesandRoadUserson theotherhand.Theskillsandcapabilitiesofroadusers,andthe qualityofvehiclesandinfrastructurecanbeimproved,e.g.forroad users–educationandrequirementssuchaslicenceagelimitsand healthrequirements(ElvikandVaa,2009).Reversingthedirection ofthearrows:highrisksmayleadtopoliciestoreducetheserisks, e.g.EuroNCAPforcars(EuroNCAP,2012)andEuroRAPforroads
Fig.1.Conceptualframeworkforroadsafety,includingexposureandrisk(sectionsdescribingthetheoriesarereferredtoinparenthesis).
(EuroRap,2011)(foreffectivenessstudies,seee.g.LieandTingvall, 2002;VlakveldandLouwerse,2011).
2.1.4. Demarcation
Inlinewithsystemstheorytheframeworkdepicts safetyas anemergentpropertythatariseswhensystemcomponents inter-act, but the components are also affected by the environment (Leveson,2004).Similarly,ourframeworkhasseveralinputsfrom theenvironmentsuchasdemographics,fuelprices,technological developments,etc.In theinterestsofreducingtheframework’s complexitytheseexternalinfluencesarenotconceptualized. Sim-ilarly,theframeworkdoesnotdepictafeedbackloopfromcrashes andinjuriestoNOAtoindicatetheeffectofinjuriesonabilities. Relationshipsthatmayexistbetweenthemodel’sexposureand risk elements (e.g.between Infrastructure, Travelresistance and betweenRoadusersandNOA)areexcludedforthesamereasonand toemphasizetheimpactofdifferencesintiming.Traveldecisions takenbeforetrafficparticipationresultinexposuretoriskduring trafficparticipation.
2.2. Theories
Thissectionbrieflydescribesintermsoftheoriesandconcepts howtheelementsintheframeworkinteract:travelbehaviour the-ories(Section2.2.1),theoriesexplainingthelinkbetweenexposure and risk (Section2.2.2), crashrisk theories (Section2.2.3), and injuryrisktheories(Section2.2.4).
2.2.1. Travelbehaviourtheories
Thedominanttheoryforexplainingtravelbehaviouris (ran-dom) utility maximization (McFadden, 1974). This holds that peoplemaximizetheirutility,e.g.atripismadeifthe(expected) benefitsofperforminganactivityatalocation(‘locationsof activ-ities’)exceedthe(expected)time,costandeffortoftravel(‘travel resistance’).Alternativemodelsofboundedrationalityhavebeen developedwhich, withoutcompletelyabandoningtheideathat reasonunderliesdecision-makingprocesses,tendtobemore psy-chologicallyplausible.Forexample,ProspectTheoryaccountsfor decisionheuristicssuchaslossaversion(KahnemanandTversky, 1979;Van deKaa,2010).Regret theoryholdsthat peoplewish toavoidtheregretthatanon-chosenalternativeturnsouttobe moreattractivethanthechosenone(Chorusetal.,2008).The The-oryofPlannedBehaviourholdsthatattitudestowardsbehaviour,
subjectivenormsandperceivedbehaviouralcontroltogethershape an individual’s behavioural intentions and behaviours (Ajzen, 1985).Decidingtomakeatripmayalsodependonneeds(e.g. driv-ingasastatussymbol),opportunities(e.g.havingarailwaystation nearbytogobytrain),andabilities(e.g.beinghealthyenoughto cycle)(seee.g.Vleketal.,1997).
Thetheoriesmentionedsofarhelpexplainthelinksbetween determinants fortravel behaviour(needs,resistance, locations). Theyalsohelpexplaintheotherlinksbetweenthefactorsinthe upper partof theframework. For instance peoplewho greatly appreciatealargecity’sculturalandsocialactivities(needs)will prefer livingin a largecity (location). People desiring safeand fasttravel (travelneeds/preferences) mayseeka dwelling near alargerailwaystation(location).Atheorythathelpstoexplain thelinkbetweenlocationsofactivitiesandNOAisthatof time-spacegeography.Itexplainsthemovementofindividualsinthe spatial-temporalenvironmentwiththeconstraintsplacedonthem bythesetwofactors(Hägerstrand,1970).Forinstance,tobeable toworkwithcolleaguesoreatfamilydinnerstogetherrequires severalpeopletobeatthesameplaceatthesametime. Opportuni-tiestogoshoppingdependonopeninghours,etc.Therelationship betweenlocationsofactivitiesandresistancescanbeexplainedby the‘theoryofconstanttraveltimebudgets’,whichholdsthat,atan aggregatelevel(e.g.thecountryorstatelevel),averagedailytime spentontravelisfairlyconstant(MokhtarianandChen,2004).For example,thismeansthatifanewmotorway,railway,orcyclepath isopenedwhichreducestraveltimes(i.e.decreasedresistance), somepeoplemayconsiderchangingresidentiallocationor desti-nationssuchasthejoblocation.Constanttraveltimebudgetscan beexplainedbyutilitytheory.Besidesseekinganoptimalbalance betweentimeforactivitiesandrelatedtravel,peoplecomparethe marginaldisutilityofextratraveltimeoradditionaltripswiththe marginalbenefitsofrelatedactivities(VanWeeetal.,2006).
2.2.2. Theoriesexplainingthelinkbetweenexposureandrisk Theframeworkdepicts arelationshipbetweenexposureand riskandshowsanarrowfromRisktoTravelresistance.Perceived risk,which isweakly correlatedtoactualrisk, influencestravel behaviour(Vlakveldetal.,2008).Theperceptionthatacertaintype ofvehiclesuchasabicycleisunsafecanbeadeterrenttoitsuse (Heinenetal.,2010).Animportantconcepttoexplaintheinfluence ofexposureonriskisthesocalled‘non-linearityofrisk’.Itholds thatthenumberofcrashesatagivenroadsectionorintersection
increasesproportionallylesswiththeincreaseinthevolumeat thatfacility(atleastaboveacertainamountoftrafficthatresults ininteractionsbetweenroadusers).Therearepossible explana-tionsrelatedtoroaduserinteractionandinfrastructure,butmost theoreticalinvestigationintotherelationshipbetweenflow and safetyseemstolackdetail(Ardekanietal.,2000).Oneexplanation isthatthesecondandsubsequentvehiclesofaplatoonmayhave amuchlowerchanceofbeinginvolvedinaright-anglecollision atasignalizedintersectionthanthefirstvehicle(Ardekanietal., 2000).Otherresearchershavesuggestedthatimproved infrastruc-turemaybeoneoftheexplanationsforthenon-linearityofrisk, e.g.Jensen(1999)arguesthatcitiesaredesignedtomeetdifferent travelbehaviour.Similarly,attheindividuallevelithasbeenfound thatdriverstravellingmorekilometreshavelowercrashratesper kilometre.Anexplanationforthisisthatthesedriversaccumulate mostoftheirkilometresonfreewaysorotherdividedmultilane highwayswherecrashratesarelower(Janke,1991).
2.2.3. Crashrisktheories
Theinteractionbetweenvehicles,roadinfrastructureandroad usersplaysaroleincrashrisk,whichcanbeexplainedusing theo-riesfromphysicsandsocialsciences.Theinteractionbetweenroad usersandroadsisoftencalled‘humanfactors’,whilethe interac-tionbetweenroadusersandvehiclesislabelledas‘man–machine factors’(Birthetal.,2009).Fourtypesofsocalled‘functionaldriver behaviourmodels’havetheabilitytodescribehowtheroad envi-ronmentandvehiclescanbeadaptedtofitroadusers’capabilities inordertoreducecrashrisk:perceptionmodels,cognitive mod-els,workload models,and motivational models(Michon, 1985; Ranney,1994;WellerandSchlag,2007).Thefirstthreedescribe whatroadusersareabletohandle;motivationalmodelsexplain whatdriversaremotivatedtodo.Physicalfactorsbasedonphysics helptoexplaintheinteractionbetweenvehiclesandinfrastructure, e.g.frictionbetweentyresandtheroadsurfacetoenablesteering andbraking(Elvik,2006)andsuperelevationtonegotiateacurve (Aram,2010).
2.2.4. Injuryrisktheories
Theories from physics, such as Newtonian mechanics, and medicine(Sobhanietal.,2011)havebeenusedtoexplaininjury risk,i.e.theseverityof injuriesincurredin a crash.The energy damagemodel,oftenattributedtoGibson(1961),isbasedonthe suppositionthatdamage(injury)isaresultofanincidentenergy whoseintensityatthepointofcontactwiththerecipientexceeds thethresholdoftherecipient(Viner,1991;Toftetal.,2012).Crash energymaybereleasedwhenthereisafailureofhazardcontrol mechanismssuchasbarriers.Inroadtrafficitisthekineticenergy producedbythemovementofpeopleandvehiclesthatisapotential crashenergy.Massdifferencesarecrucialwhenmotorvehiclesand vulnerableroaduserscollide.Energymaybeexchangedbetween vehicles,roadusers,andinfrastructure,meaningthatitaffectsall threesafetypillars.Crashesmaybefatalwhenforcestransferred tovictimsexceedtheirbiomechanicaltolerance,whichdependson age,healthstatus,stature,andotherfactors(Corbenetal.,2004).
3. Cyclingsafetyrelatedtolanduseandinfrastructure
Todemonstratetheframework’susefulness,thissectionapplies ittotherelationshipbetweencyclingsafety,andlanduseand infra-structurecharacteristics.Theframeworkelementsmostrelevantto thisissueareLocationsofactivities(landuse),Travelresistance (net-workandroadcharacteristics),andInfrastructure(roaddesign).We havesearchedforscientificliteratureoncyclingandcyclingsafety, preferablyempiricallyvalidatedorotherwisetheoreticallyfeasible, thatissuitablefordescribingdifferentpartsofthemodel.
3.1. Travelbehaviourandexposure
Thissectiondescribescyclingtravelbehaviour(Section3.1.1.) andthedistributionoftrafficovertimeandspace(Section3.1.2).It referstobothmotoristsandcyclistsbecausemodalsplitand distri-butionovertimeandspacedeterminethedegreetowhichcyclists areexposedto(highspeed)motorists.
3.1.1. Cyclingtravelbehaviour(volumesandmodalsplit)
Thissectiondescribesstudiesthatrelate cyclingtoland use andinfrastructurecharacteristics.Morestudiesfocusedonmode choicethanoncycling frequency(Heinenet al.,2011).Because thedecisiontocycleandcyclingfrequencyarestrongly interre-lated,itwasdecidednottomakeanyfurtherdistinctionbetween theminthis section.Landuseandinfrastructurecharacteristics affectcycling distances.Thisis importantbecausethedisutility ofcyclingincreasesmorethanproportionallyforlongerdistances, whichmightbeexplainedbyphysiologicalfactorsandspeed(Van Weeetal.,2006).Heinenetal.(2010)concludefromtheirliterature reviewonbicyclecommutingthatdistanceisadauntingfactorfor cyclists.Landusecharacteristicswhichcontributetoshortertravel distances,suchasahigherpopulationdensity(e.g.acompactcity) andmixedland-use,havebeenfoundtoaffectcyclingpositively (Heinenetal.,2010).
Resistance is strongly linked to the physical and functional characteristicsofinfrastructurenetworks.Thefollowingeffectson bicycleuseforutilitarianpurposes(allpurposesapartfrom recre-ational/leisurepurposes)havebeenfound:
• Roadstructuredensity:AccordingtoSouthworth(2005),adenser roadstructureis moresuitablefor non-motorized transporta-tionbecausedistancesaregenerallysmaller.However,neither Moudon et al. (2005) nor Zacharias (2005) found significant empiricalevidencethatcanconfirmtheinfluenceofthedensity ofroadwaysandblocksizeoncycling.
• Bicyclepaths:WhileHeinenetal.(2010)havefoundseveral stud-ieswhichconcludethatmorebicyclepathsresultin ahigher shareofcycling(e.g.BarnesandThompson,2006),theyalsofound studiesinwhichnosignificanteffectwasfound(e.g.Moudon et al., 2005). Additional infrastructure might make little dif-ferenceincountrieswherecyclingfacilitiesaremorecommon (Heinenetal.,2010).
• Numberofstops:RietveldandDaniel(2004)havefoundthatthe numberofstopscyclistshavetomakeontheirroutesisa deter-renttocycling.
3.1.2. Distributionoftrafficovertimeandspace
Littleresearchhasbeendoneontheeffectofinfrastructureon thedistributionofcyclingtrafficany24hperiod,exceptperhaps thereluctanceofoldercycliststocycleindarkness,whichmaybe influencedbythevisualdesignofinfrastructureandthepresenceof streetlighting(SchepersandDenBrinker,2011).Itisobviousthat landuse(thedistributionofactivitylocationsoverspace)hasan effectonthedistributionoftraffic(includingcycling)overtimeand space.Forinstance,anentertainmentcentremayattractyoung vis-itorsatnight.Itslocationattheedgeoftownmayresultinlonger averagedistances betweenitand thelocationsofthedwellings ofyoungvisitors,resultinginalowershareofcyclingandlonger cyclingdistancesforthosewhodocycle.Highexposureto danger-oussituationssuchasdrivingatnightatweekendhasbeenfoundto beacauseofthehighcrashrateofyoungnovicedrivers(Vlakveld, 2005).Similarly,researchsuggeststhatyoungstersfrequentlycycle atnightandfrequentlyafterhavingconsumedalcohol(Reurings, 2010;SchepersandDenBrinker,2011).
Aconceptthathelpstodescribethedistributionoftrafficover spaceis‘streethierarchy’.Thisaffectsroutechoicebymanipulating
travel times,i.e. resistance(see formore information Hummel, 2001).ThisconceptbecameveryinfluentialafterBuchanan(1963) publishedTrafficinTowns.Inahierarchicalroadstructure,lower orderroads(accessroadsinwhatBuchanannamed ‘environmen-talareas’)serveaccesstraffic,whilehigherorderroadsservean efficient flow of through motor traffic (through roads such as motorways).Inbetweenaredistributororcollectorroadsto dis-tributetrafficfromthroughroadstoaccessroadsandviceversa. Amotorwaynetworkwherecyclistsarenotallowed,withgrade separatedintersections,reducescyclists’exposuretohighspeed motorists.Accessroadsaredesignedforlowspeedstokeepthrough motortrafficaway.Ahighshareofshortbicycletripsresultsina highnumberofkilometresbeingtravelledonaccessroadswhere exposure to(high speed) motorists is limited. Research shows thatthenumberofbicycle–motorvehiclecrashesisindeedhigh along distributorroads and low on access roads (Berends and Stipdonk,2009;Schepersetal.,2011).Evaluationstudieshavenot yetaddressedthecombinedeffectofageneralroadhierarchywith bicycle-specificmeasuressuchasbicyclebridgesandtunnelsto alleviatepotentialsafetyproblemsatdistributorroads. Depend-ingonhowtheroadnetworkfitstheneedsofdifferenttransport modes,aroadhierarchymayaffecttravel timesfor driversand cyclistsdifferently,therebyaffectingmodalchoice.Cyclistsmay benefitfromshortcutswhereroadsareclosedfordriversandfrom beingallowedtouseone-waystreetsinbothdirections,etc. Provid-ingmoredirectroutingforonemodeincontrasttotheothermay increasemodeshareforthefavouredmode(FrankandHawkins, 2008).
3.2. Therelationshipbetweenexposureandrisk
This section describes the relationship between exposure (resultingfromtravelbehaviour)andrisk:firstlytheeffectof expo-sureonrisk(Sections3.2.1and3.2.2)andsecondlytheeffectof (perceived)riskonexposure(Section3.2.3).
3.2.1. Theeffectofbicyclevolumesonroadsafety
The number of crashes at a given road section or inter-section increases proportionally less than the increase in the volumeatthatfacility;thesameappliestobicycle–motorvehicle crashes(BrüdeandLarsson,1993;Elvik,2009)andsingle-bicycle crashes(Schepers,2012).Cyclingsafetyresearchdescribesthe non-linearityofriskasthe‘safetyinnumbers’phenomenon(Jacobsen, 2003). Jacobsen’s (2003) explanation is that motorists modify theirbehaviourwhentheyexpectorexperiencepeoplewalking andbicycling.Theoriesregardingexpectancyintrafficwhichcan underpin this are described by researchers such as Houtenbos (2008)andTheeuwesandGodthelp(1995).Othershavesuggested thatimprovedinfrastructuremaybeoneoftheexplanationsfor thenon-linearityofrisk(BrüdeandLarsson,1993;Wegmanetal., 2012).Thenon-linearityofriskimpliesthatcyclistsaresaferwhere therearemorecyclists.Itisdifficulttodrawconclusionsabouthow roadsafetyingeneralwillbeaffectedbecausethenon-linearityof riskalsoappliestoothermodesoftransport.
3.2.2. Modalsplitandroadsafety
Cyclingisassociatedwithaconsiderablyhigherriskofinjury accidentsthan travel by car (Wegman et al., 2012). One could thereforeexpectthatamodalshiftfromcartobicyclewouldhave negativeeffectsonroadsafetyingeneral.However,thereare rea-sonswhytheeffectislimited.Themostimportantoneisthatafter shiftingfromcardrivingtocycling,individualsarelesshazardous toothervulnerableroadusers(includingcyclists)becauseofthe loweramountsofkineticenergyexpendedintheeventofacrash. Anumberofstudieshaveaccountedforthisfactor(seeforother explanationsSchepersandHeinen,2013).
UsingexistingAccidentPredictionModels(APMs)inwhicha nonlinearrelationshipbetweencrashesandvolumesisassumed, Elvik(2009)wasthefirsttoestimatetheroadsafetyeffectsofshifts fromcartobicycle(andwalking).Hisresultssuggestthatifthere areverylargetransfersoftripsfrommotorvehiclestowalkingor cycling,thetotalnumberofaccidentsmaybereduced.Hismethod wasrecentlyappliedtoDutchdatabySchepersandHeinen(2013). Theirresultssuggestthattransferringshorttripsmadebycarsto bicyclesdoesnotchangethenumberoffatalities,butincreasesthe numberofseriousroadinjuries.StipdonkandReurings(2012) fol-lowedadifferentapproachtodeterminetheeffectofanexchange overa shortperiodoftime,i.e.withoutadaptinginfrastructure. Insteadof(stochastic)APMs,theyappliedadeterministicmodel, assumingalinearrelationshipbetweenvolumesandroadcrashes. Thestudyresultssuggestthatamodalshiftfromcarstobicycles leadstoasmallincreaseinthenumberoffatalitiesandagreater increaseinthenumberofhospitalizedcasualties.Thelatterisdue tothehighnumbersofcyclistsinjuredinsingle-bicyclecrashes. BothStipdonkandReurings(2012)andSchepersandHeinen(2013) findthateffectsvaryacrosstheagegroups.Elderlydriversaresafer insideacarthanonabicycle.Fromaroadsafetyperspective,the car–bicycleshiftis,onbalance,advantageousforyoungdriversand disadvantageousforelderlydrivers.
Studiesonthehealth effectsof amodalshiftfromshortcar tripstocycling(e.g.DeHartogetal.,2010)havenotyetincluded thehealthburdenanticipatedfromanincreasednumberof single-bicycle crashvictims.This is becausestudies thatincorporated single-bicycle crashes in estimations of the road safety effects ofamodalshiftwerepublishedonlyrecently(i.e.Stipdonkand Reurings,2012;SchepersandHeinen,2013).However,giventhe largehealthbenefitsassociatedwithphysicalexercise(DeHartog etal.,2010;Ojaetal.,2011),itislikelythatthathealthbenefits willoutweighthehealthrisks,evenifthesenon-fatalcrasheswere included.
3.2.3. Theeffectofriskonbicycleuse
People,especiallynon-cyclists(HeinenandHandy,2012), gen-erallyperceivecyclingtobelesssafethanwalking,drivingacar orusingpublictransport.Thiswouldimplythatthisformoftravel resistanceishigherforcycling (Elvik andBjørnskau,2005).The riskofanaccidentis adeterrenttocycling(Parkinetal.,2007; Heinenetal.,2010).Researchindicatesthat cyclistsprefer ded-icatedbicycleinfrastructurebecausetheyperceiveittobesafer (Heinenetal.,2010).Forinstance,Gårderetal.(1998)foundan increased volume of cyclistsat road sectionsafter cycle tracks hadbeeninstalled.Vandenbulcke-Plasschaert(2011)suggeststhat actualand perceivedrisksof cyclingmay beoneof thefactors explainingthehighamountofcycling inFlandersinthe north-ernpartofBelgium,ascomparedtoWalloniain thesouth.The samereasoningmaybevalidinexplaining differencesbetween countries.RietveldandDaniel(2004)foundthatsafetyappearsto matterasacomponentingeneralizedcostsandthatitexplainspart ofthevariationintheamountofbicycleuseinDutchmunicipalities. PucherandBuehler(2008)suggestthatsafetymayaffectthe com-pilationofthepopulationofcyclistsbecausewomen,theelderly andparentsofyoungchildrenappeartobeespeciallysensitiveto perceivedroadsafety. Thismaybeanotherfactorthat explains differences insafety betweendifferentcountries, i.e. cyclistsin countrieswithhigheramountsofcyclingmaybemorecautious. Finally, the injuries incurredin crashesmay affectbicycle use. Ormeletal.(2008)foundmorethanone-thirdofallhospitalized single-bicyclecrashvictimscycledlessaftertheiraccident,because ofacombinationofphysicalproblemsandfearoftakinganother fall.
3.3. Crashrisk
Thissectiondescribeshowcyclingriskisaffectedby infrastruc-turecharacteristics.Theriskofcollisionsdependsonthenumber ofpotentialconflictpointsandhow wellroadusersareableto handleconflicts.Forinstance,aroundaboutreducesthenumber of potential conflictpoints compared to an intersection which hasfavourablesafetyeffectsingeneral(Elvik,2004),althoughthe effectsfoundforcyclistsarenot consistent(Brüdeand Larsson, 2000;Dijkstra,2004;Danielsetal.,2009).Sakshaugetal.(2010) havefoundahighernumberofconflictand interactiontypesat roundaboutswhere cyclistsaremixedwithothervehicles com-paredtoaroundaboutwithseparatecyclecrossings.Theriskof single-bicyclecrashesisinfluencedbyhowwellcyclistsare sup-portedwhenbalancingandsteeringtheirbicycles,andavoiding obstacles(Schepersand KleinWolt,2012).Theabovementioned issuesrefertotheframework’slinkofInfrastructuretoRoadusers (Section3.3.1)andtoVehicles(Section3.3.2).Humanfactors theo-riesorergonomicstheorieshelpexplainhowroadscanbedesigned tofitroadusers’needsandcapabilities(Birthetal.,2009).Theories fromphysicshelptodescribehowinfrastructurecanbedesigned tohelpcyclistssafelybalanceandsteertheirbicycles.
3.3.1. Humanfactors
The application of ergonomics theories for optimal cycling safetydependsonthecontext.Whileacompleteoverviewof appli-cationsis outside thescopeof this text,this papergives some examplestoshowthevalueofhumanfactortheoriesforcycling safety.
Theoriesonperceptionhelpunderstandtowhatextentroad usersareable toperceiveobjectsand where theroad isgoing. Forexample,ambient-focaldichotomyisapowerfultheorywhich describesvisionanddrivingintermsofthevisualsystemasbeing twoparallelstreamsofprocessing,labelledtheambientandfocal subsystems(LeibowitzandOwens, 1977;Schieberet al.,2008). Thepropositionisthatvisualprocessingproceedsalongtwo par-allelstreams,onededicatedtovisualorientationforthequestion “WhereamI?”(ambientvision)andtheothertoobject recogni-tionandidentificationforthequestion“Whatisit?”(focalvision) (LeibowitzandPost,1982;Previc,1998).Driversuseambientvision totrackandminimizeinstantaneouserrorsinlaneposition.They usefocalvisiontoanticipatehazardsandfuturealterationsinthe courseoftheroad(Donges,1978).SchepersandDenBrinker(2011) recentlyusedtheambient-focaldichotomyinastudywherethey showedthatthevisualdesignofbicyclefacilitiesplaysarolein single-bicyclecrashes.
Apowerfultheoryfromcognitivepsychologyis‘expectancy’ theory(TheeuwesandHagenzieker,1993;Houtenbos,2008). Con-cepts such as Self-Explaining Roads (Theeuwes and Godthelp, 1995), geometric consistency(Fitzpatrick et al., 1999), and the SustainableSafetyprincipleof predictabilityandrecognisability (WegmanandAarts,2006)allholdthatroadsshouldbedesigned inlinewithroadusers’expectationsandsuchthattheycreatethe rightexpectations.Anoften-citedviolationof expectationsthat resultsinerrorsoccursatpriorityintersectionswithtwo-way bicy-cletracks.Theriskofbicyclecrashesisfoundtobeelevatedbecause driversenteringfromtheminorroadhavedifficultiesindetecting cyclistsfromtheright(incaseofright-handdriving)(Räsänenand Summala,1998;Schepersetal.,2011).Summalaetal.(1996) stud-ieddrivers’scanningbehaviouratT-intersections.Driversturning rightfromtheminorroadscannedtherightlegoftheT-intersection lessfrequentlyandlaterthanthoseturningleft.Theirexplanationis thatdriversturningrightfocustheirattentiononcarsfromtheleft becausethosecomingfromtherightposenothreattothem.The visualscanningstrategyseemstoconcentrateonmorefrequent andmajorpotentialdangers(Summalaetal.,1996).
Workloadmodelsindicatethathumanshavealimited infor-mation processing capacity. Taking into account individuals’ capabilities,workloadcanbeeithertoolow(‘underload’)ortoo high(‘overload’)duetothetaskdemandsofdrivingorcyclingand doubletaskssuchasmobilephoneuse(DeWaard,1996;DeWaard etal.,2010).Aconceptlinkedtoroaddesign andrelatedtothe probabilitythatsomeroaduserswillbeoverloadedis‘complexity’. AccordingtoElvik(2006),the‘lawofcomplexity’holdsthatthe moreunitsofinformationaroadusermustattendto,thehigher becomestheaccidentrate.Thisespeciallyappliestosituations sub-jecttotimepressure.Forinstance,olderdriversandcyclistsare moreoften involvedin left-turningcrashesand situations with associatedtimepressureswheretrafficfromseveraldirectionshas tobescanned(Goldenbeld,1992;Davidse,2007).Fromthe per-spectiveofworkload,theoppositeof‘complexity’is‘monotony’ (Birthetal.,2009)or‘highwayhypnosis’–reducedalertnesson long,straightroads(SWOV,2012).
Motivational models describe how road users adapt their behaviour to the environment if the driving task is self-paced (Ranney,1994).Homeostasismodelsassumethatdriversare con-stantlyawareof,monitorandseektomaintainasetlevelorrange ofa variable,suchas risk(e.g.Wilde’sriskhomeostasis model; Wilde,1982)ortaskdifficulty(e.g.Fuller’stask-difficulty homeo-stasismodel;Fuller,2005).Asecondgroupoftheoriesclaimsthat variablessuchasaperceptionorfeelingofriskareonly experi-encedatcertaintimesduringdriving,i.e.whenacertainthreshold isexceeded(Lewis-Evansetal.,2011),forinstancetheZero-Risk theorydeveloped byNäätänenandSummala(1974).Aproblem withthesemotivationaltheoriesisthattheydo notdescribeto whatextentroadusersmayadapttheirbehaviourinresponseto certainmeasures.Bjørnskau(1994)proposedhypothesesdesigned toexplainroaduserbehaviouraladaptationtoroadsafety meas-ures.Forinstance,highlyvisiblechangestotheroadaremorelikely toleadtobehaviouraladaptationthanmeasuresthatroadusers donoteasilynotice.Visiblemeasuressuchasblue-painted bicy-clecrossingshavebeenshowntoresultinbehaviouraladaptation. Fewercycliststurnedtheirheadstoscanfortrafficorusedhand signalsafterthemeasurewasimplemented(Hunteretal.,2000). NotethatSection3.2.3describestheriskperceivedanditseffect ondecisionstakenbeforetrafficparticipation,whereasthissection describedtheperceptionofriskanditseffectonbehaviourduring trafficparticipation.
3.3.2. Physicalfactors
Inthecontextofthispaper,theterm‘physicalfactors’isused torefertotheinteractionbetweenvehiclesandroad infrastruc-ture.Anexampleisthefrictionbetweentiresandtheroadsurface neededforbraking.Nybergetal.(1996)haveshownthataslippery roadsurfacecontributestosingle-bicyclecrashes.Theytherefore adviseinvestmentinwintermaintenance.Theremainderofthis sectionfocusesonbicyclestabilitybecauseitmayplayan impor-tantroleinsingle-bicyclecrashes.
Acontrollingridercan balanceaforward-moving bicycleby turningthefrontwheelinthedirectionofanundesiredlean,i.e. steeringtotherightwhenfallingtotheright,andviceversa.This movestheground-contactpointsundertheriderandresultsina zig-zagmovement.Mostbicyclescanbalancethemselves (‘rider-less’)ifmovingaboveagivenspeed,becausetheyareabletosteer intotheleanautomatically.Godthelpand Wouters(1978)used anexperimenttoestimatethatundernormalcircumstancesand speeds,cyclistsrequireatrackwidthofabout1metreto accom-modatetheresultingzig-zagmovementandspaceforthebicycle. Theyrecommendaminimumwidthof2mforone-waybicycle trackstoenablecycliststoovertakesafety.
Mooreetal.(2009)foundself-stabilityatspeedsabove approx-imately 15km/h fora commonlyused Dutchcity bicycleand a
malerider.Stabilizingabicycleatlowspeedrequiresmoreactive steering.Severalfactors,includinggeometry,massdistributionand gyroscopic effect all contribute in varying degrees to this self-stability. Long-standing hypotheses and claims that any single effect,suchasgyroscopicortrail,issolelyresponsibleforthe stabi-lizingforcehavebeendiscredited(Kooijmanetal.,2011).Theroleof speedinstabilitysuggeststhatthedesignofbicyclefacilitiesshould enablecycliststomaintainaminimumspeed,e.g.sufficientlylarge curveradiiandnottoosteepaslope(seee.g.CROW,2007).
Thestabilityalsodependsonthefreedomofthefrontforkto swivel.Ifitislocked,suchaswhenthefrontwheelbecomesstuck inthetramrails,thebicyclecannotberidden.Adifferenceinheight betweentheroadsurfaceandshouldersurfacemakesitdifficultfor thecyclisttosteerbackafterridingofftheroad,andcanleadtofalls (SchepersandKleinWolt,2012).Finally,itisobviousthatroad sur-faceirregularitiessuchaspotholescontributetolossofcontroland therebysingle-bicyclecrashes(Nybergetal.,1996).Dutchdesign guidelinesadvisethatbicyclecrossingsintersecttramrails perpen-dicularly,thedifferencebetweentheleveloftheroadandshoulder surfacebeminimal,andtheroadsurfacebewellmaintained,etc. (CROW,2007;VanBoggelenetal.,2011).
3.4. Injuryrisk
Thissectionexplainsinjuryriskforcyclistsandhowitisaffected by infrastructure characteristics.The amountof kinetic energy producedisafunctionofthemassandvelocity(speed):1/2mv2
(m=mass;
v
=speed).Thelawofconservationofenergystatesthat thetotalamountofenergyinanisolatedsystemisconservedover time.Inroadcrashes,kineticenergyispartlyconservedandpartly convertedtoothertypesofenergysuchasdeformationenergyand heat.Partofthekineticenergyistransferredtothevictimsinvolved inthecrash.Crashesmaybefatalwhentheseforcesexceedthe vic-tims’biomechanicaltolerance(Corbenetal.,2004).Crumplezones, airbags,andcrashbarriersslowthestoppingprocessandspread thecrashenergyofthecrashoutovertime,reducingthepeakspike ofenergytothehumanbody.Similarly,airbagsonthewindscreen (Rodariusetal.,2008)orbicyclehelmets(Elvik,2011)mayprotect cyclistsintheeventofacrash.Thisprincipleiscalled‘physical for-givingness’inthecaseofroadsidefurniture(Wegmanetal.,2012). Theprinciplehastoourbestknowledgenotyetbeenappliedtothe designofforinstanceobstacleswithwhichcyclistsmaycollide.Whendifferentcategoriesofvehiclesorroaduserscrash,their compatibilityintermsofmassandspeedinfluencestheaccident outcome. Compatibility refers tothe differences between cate-goriesofroadusersintermsofthekineticenergyproducedbytheir movements.Thesmallerthesedifferences,themorecompatibleare roadusers.
Elvik(2010)calculated foreach transportmode, theratioof thenumberofcasualtiesamongthose inothermodesof trans-portdividedbythenumberofcasualtiesinthevehicletypeunder question.Forinstance,theratiofortransportmodexwouldbe0.5 if500roaduserswereinjuredinothermodesversus1000inmode x.Theratiorangedfrom0.03forpedestriansand0.05forcycliststo 0.27forcaroccupantsand3.46fortruckoccupants.Theproblem ofincompatibilitycontributedtothedevelopmentofthe Sustain-ableSafetyprincipleof‘homogeneity’.Thisstatesthatwhereroad usersorvehicleswithlargemassdifferencesusethesametraffic space,thespeedsshouldbesolowthatthemostvulnerableroad usersandtransportmodescomeoutofacrashwithoutanysevere injuries(Wegmanetal.,2012).
Theideathatthemostsevereinjuriescanbepreventedby keep-ingspeedsunderathresholdforcertaincombinationsofroadusers ledtotheconceptof‘safespeeds’.TingvallandHaworth(1999) consider30km/h a safespeedwhere vulnerableroad usersare mixedwithmotorizedvehicles.Severalstudieshaveconfirmedthat
thereisathresholdaround30km/h,abovewhichtheprobabilityof injuryandfatalityforpedestriansandcyclistscollidingwithmotor vehiclesstronglyincreases(Kimetal.,2007;Rosénetal.,2011). However,thisdoesnotapplytolorries,wherefarlowerspeedscan easilyendinafatalityifacyclistgoesunderthewheels(Schoon, 2006).
4. Exampleofapplicationoftheframework:converting one-wayintotwo-waybicycletracks
Themeasureofconvertingone-wayintotwo-waycycletracksin anurbanareaisdescribedtoexemplifytheframework’sapplication tocyclingsafety.Themeasureisconsideredbyamunicipalitythat wantstoincreasetheamountofcyclingandimprovecyclingsafety. Wefirstlyaddresspotentialeffectsonexposure.Cyclistshave rightof waywhile travellingalong distributors.Two-waycycle tracksmakearoutealongsuchroadsevenmoreattractive.Cyclists do not have tocross theroad to travel at the right-handside (in the case of right-hand driving). The reduced stopping fre-quencymay, toa small extent,increase the amountof cycling (Rietveldand Daniel,2004)and mayaffect routechoiceto the extent that theroute along thedistributor reducestravel time (GommersandBovy,1987).Morecycliststravellingalong distribu-torsinsteadofthroughtraffic-calmedareasaffectstheexposureof cycliststohigh-speedmotorists.Throughmotortrafficiskeptout oftraffic-calmedareas,resultinginareducedexposureonaccess roads.Mostbicycle–motorvehiclecrashesatdistributorroadsare collisionsbetweenthroughcyclistsand motoristsfromtheside road(Scheperset al.,2011).Secondly, weaddress the relation-ship betweenexposureand risk. Theslightly increasedamount ofcyclingandhigherproportionofcyclinginthemodalsplitcan beexpectedtohavehardlyanyeffectonthenumberoffatalities, althoughitwouldincreasethenumberofseriouslyinjured casu-alties(StipdonkandReurings,2012;SchepersandHeinen,2013). Lastly,wedescribepotentialeffectsonriskduetotwo-wayinstead ofone-waybicycletracks.Two-waycycletracksincreasecrashrisks becausecyclistsattheleftsideoftheroadcomefroman unex-pecteddirectionfordriversfromtheminorroad(Summalaetal., 1996;Schepersetal.,2011).
Theresultsoftheanalysissuggestthatconvertingone-way bicy-cletracksintotwo-way mayslightlyincrease bicycleusageand increasethenumberofbicycle–motorvehiclecrashes.Although theframework does notallow forquantitative assessment, the exampleshowsitsvalueforidentifyingpotentialeffects.Theeffect oncrashesduetoachangeddistributionofcyclistsovertheroad network is normallynot addressedin roadsafety research.For instance,thestudybySchepersetal.(2011)isoneofthefewwhich controlledforvolumesofbothmotoristsandcycliststodetermine theeffectoftwo-wayversusone-waybicycletracksatunsignalized priorityintersectionsontheriskofbicycle–motorvehiclecrashes. Tothebestofourknowledge,thereisnostudythatincludesthe potentialadverseeffects of morecyclistschoosingroutes along distributorroadswherecyclists areexposed toahigher riskof bicycle–motorvehiclecrashes.
5. Discussion
This paper presented a conceptual road safety framework comprisingfactorsdeterminingexposuretorisk(resultingfrom travelbehaviour),risk(injuryandcrashrisk),andtherelationship betweenthesetwo.Modelsfortravelbehaviourandroadsafetyare notnew,buttothebestofourknowledge,nowhereinthescientific literatureisthereaframeworkcomprisingbothandthe relation-shipbetweenthem.Theframeworkcanbeusedtoacquireinsight intothepotentialeffectsofmeasures.Anexampleofsuchaneffect
(identifiedinSection4)isachanged distributionoftraffic over theroadnetworkresultingfromconversionof one-waybicycle trackstotwo-way. We believethatthis effect,which is impor-tantforexposureofcycliststohigh-speedmotorists,hasnotyet beenaddressedinanystudyonroadfactorsandbicyclecrashes. Theframeworkasdescribedinthispapercanhelptoidentifysuch potentialeffects.
5.1. Researchrecommendationsregardingtheframework
Futureresearchmayrequiretheframeworktobebroadened oradaptedforapplicationtosubjectsotherthancycling.Afirst additionalapplicationcouldbethepost-crashphaseinwhich,for instance,theemergencymedicalsystemisrelevant.For applica-tiontopedestriansafety,whereitwouldbeusefultobeableto describefallsinwhichnovehiclesareinvolved,itmaybe help-fultoreplace‘roadusers,vehicles,andinfrastructure’byHaddon’s (1980)‘host(thepedestrian),agent(kineticenergyandgravity), andenvironment(theroadandsocialenvironment)’.Secondly,the modelisconceptualanddoesnotallowforquantitativeassessment oftheeffectson(injury)crashnumbers.Empiricalstudieswould beneededforthis.However,itwouldbepossibletotestthemodel againstthefindingsofabroaderliteraturesearchoncyclingsafety. Werecommendaliteraturereviewtoexploretheeffectsofabroad rangeofmeasuresonbicycleusageandcyclingsafetyincludedby governmentsinbicycleplans.Theoutcomesmayrevealelements orlinksmissinginthemodel.
In the interests of maintaining ease of communication, we adviseagainstexpandingthenumberofelementsinthe frame-worktoomuch.Insteadwerecommendtoseekmodelsthatare morespecifictoexplainrelationshipsinthemodelinmoredetail. Thismightbemorecomplicatedthanitseems.Forinstance,inthe caseof thelinkbetweenroadusers and infrastructure,no the-oryofdriverbehaviourhasyetmanagedtoachievewidespread acceptance(Lewis-Evansetal.,2011),humanfactorstheoriesare developedmainlyfromtheperspectiveoftheindividualroaduser, whichhampersrelatingthemtodesignfeatures.Whiledesign prin-ciplessuchas‘geometricconsistency’canbeaccommodated in humanfactorstheories,mostarenotembeddedinmodelswhich couldshowtheirrelationshiptootherconcepts.Areviewof mod-elsforrelationsinourframeworkcouldbeofhelpforresearchers designingnewstudies.
5.2. Researchgapsregardingcyclingsafety,landuseand infrastructure
Thesubjectofcyclingsafety,landuseandinfrastructurewas usedinthispapertoexemplifyapplicationoftheframework.The literaturesurveyconductedforthispurposehasshownanumber ofresearchgapsofwhichwegivesomeexamplesintheareasof travelbehaviour,therelationshipbetweenexposureandrisk,and risk.Wefirstlyaddresstravelbehaviour.
5.2.1. Travelbehaviour
Heinenet al.(2010)indicate thatlargemodechoicestudies oftenlackfactorsrelatingspecificallytobicycleuse.Manybicycle researchstudiesexamineonlyalimitednumberoffactors,which makeitdifficulttodeterminetheirrelativeimportance.Littleorno researchfocusedontheeffectofcertainaspectsofinfrastructure, suchasbicycletracks,thepresenceoftrafficlights,andpavement quality(Heinenet al.,2010).Atthenetwork level, there isthe questionof whetherproviding more direct routing for cyclists ratherthanformotoristswouldincreasetheshareofcyclinginthe modalsplit(FrankandHawkins,2008).Animportantsafety consid-erationishowaroadhierarchywithbicycle-specificmeasuressuch asbicyclebridgesandtunnelsmaydecreasecyclistexposureto
high-speedmotoristsatdistributorsandtherebyimprovecycling safety.Tothebestofourknowledgethishasnotyetbeenstudied. 5.2.2. Therelationshipbetweenexposure(resultingfromtravel behaviour)andrisk
ResearchonthisissueresultedinAccidentPredictionModels thathelpestimatetheeffectofamodalshiftonroadsafety(e.g. Elvik,2009;SchepersandHeinen,2013).Researchershave sug-gestedseveralpossibleexplanationsforthenon-linearrelationship betweentrafficvolumesandriskbuttheinternalvalidityofcurrent studieswasnotstrongenoughforinferringcausality(Bhatiaand Wier,2011).Abetterunderstandingisimportantforpolicymakers, e.g.policycouldfocusoninfrastructureifthenon-linearityresults fromimprovedinfrastructure(Bhatiaand Wier,2011;Wegman etal.,2012).Asimilarlackofinternalvalidityseemstoapplyto researchontheeffectofperceivedriskonbicycleuse.Notonlyis reducedperceivedriskassociatedwithincreasedbicycleusage,it alsotendstoalterthecompilationofthecyclistpopulationbecause somegroups,suchaswomenandtheelderly,appeartobe espe-ciallysensitivetoroadsafety(PucherandBuehler,2008;Heinen etal.,2010).Theseparticulargroupsofcyclistsmayhavea differ-entriskprofilewhichcouldchangecyclistcrashrates.Thefactof thesegroupstakingupcyclingmaychangehowpeopleperceive theriskofcycling,andsoon.Explorationofthecausalmechanisms thatmightexplaintherelationshipbetweenriskandbicycleusage ischallengeforfuture research,andcouldhelpsubstantiatethe linksinbetweentheupperandlowerpartofourmodel.
5.2.3. Risk
Reynoldsetal.(2009)concludefromtheirreviewstudythat thereisgoodresearchontheeffectoftheprovisionofbicycle facil-itiesonroadsectionsandroundaboutsoncyclingsafety,butless onsignalizedandunsignalizedintersections.Furthermore,many researchersfocusonthepresenceofafacility,butpaylittle atten-tiontohowitisdesigned(Reynoldsetal.,2009),indicatingthat manystudiesarenotyetwellunderpinnedbyhumanfactors the-ories(Schepersetal.,2011).Whilemuchresearchhasfocusedon theriskofbicycle–motorvehiclecrashes,onlyafewstudiesfocused onsingle-bicyclecrashes(SchepersandKleinWolt,2012).Asmall numberofstudiesindicatedthataspectssuchaswinter mainte-nanceofbicyclefacilities,roadsurfacequalityandthevisualdesign ofinfrastructureplayaroleinsingle-bicyclecrashes(Nybergetal., 1996;SchepersandDenBrinker,2011).However,evaluation stud-iesontheeffectsofinfrastructureonsingle-bicyclecrashesarenot yetavailable.Theseresearchgapsshowthatourtheoretical under-standingofhowinfrastructureisrelatedtoroadusersandvehicles canbeimproved.
Finally,itisworthmentioningthatevenwhencyclingsafety researchisavailable,oneneedstotakeintoaccountthat,because oftheconsiderabledifferencesbetweencountriesincyclingand cyclingsafetyconditions,theoutcomescannoteasilybe general-ized(LooandTsui,2010;Wegmanetal.,2012).Ourunderstanding canbefurtherimprovedbyconductingsimilartypeofstudiesin differentcountries.
5.3. Relevancefortransportpolicy
Fromapolicyperspective,abroadframeworkcomprisingboth travel behaviour and risk factors is important because govern-mentsoftenhavegoalsforboth.Forinstance,theygenerallyaim toimproveroadsafetyandincreasetheamountofcyclingatthe sametime.Wegmanetal.(2012)thereforeposedthequestionof howtoconsiderincreasedcyclingandbettercyclingsafety simul-taneously.Effectsof changedamounts of cycling(and walking) suchasthoseresultingfromchangesincyclists’traveltime,road users’perceptionofsafety,andhealthareoftenignoredincurrent
cost–benefitanalysesofmeasuresdesignedtoimprovesafetyor mobilityforpedestriansandcyclists(Elvik,2000;VanWeeetal., inpress).
InfluentialroadsafetyvisionssuchasVisionZero(Tingvalland Haworth,1999),SustainableSafety(WegmanandAarts,2006),and SafeSystems(ACT,2008)arecriticaltoroadsafetyimprovement. However,theyarenotdevelopedtoprovideinsightintotheeffects ofpoliciesontravelbehaviour.Aframeworksuchastheone out-linedinthispaperisastartingpointforacquiringinsightintoboth travelbehaviourandsafetyeffects,andexploringwhethersome measuresmaybeconflicting.
References
ACT,2008. NationalRoadSafetyActionPlan2009and2010.AustralianTransport Council,Canberra.
Ajzen,I.,1985.Fromintentionstoactions:atheoryofplannedbehavior.In:Kuhl,J., Beckmann,J.(Eds.),ActionControl:FromCognitiontoBehavior.Springer-Verlag, NewYork.
Aram,A.,2010. Effectivesafetyfactorsonhorizontalcurvesoftwo-lanehighways. JournalofAppliedSciences10,2814–2822.
Ardekani,S.,Hauer,E.,Jamei,B.,2000.Trafficimpactmodels.In:Gartner,N.,Messer, C.J.,Rathi,A.K.(Eds.),TrafficFlowTheory:AStateoftheArtReport.Federal HighwayAdministration,Washinton.
Asmussen,E.,Kranenburg,A.,1982.AnAnalysisoftheTrafficPhenomenon.Institute forRoadSafetyResearch,Leidschendam.
Barnes,G.R.,Thompson,K.B.,2006. Alongitudinalanalysisoftheeffectof bicy-clefacilitiesoncommutemodeshare.In:TransportationResearchBoard85th AnnualMeeting,TransportationResearchBoard,Washington,DC.
Berends,E.M.,Stipdonk,H.L.,2009. Deveiligheidvanvoetgangersenfietsersop 30km/uur-erftoegangswegen(PedestrianandCyclistSafetyon30km/hAccess Roads).InstituteforRoadSafetyResearch,Leidschendam.
Bhatia,R.,Wier,M.,2011.“Safetyinnumbers”re-examined:canwemakevalidor practicalinferencesfromavailableevidence?AccidentAnalysisandPrevention 43(1),235–240.
Birth,S.,Pflaumbaum,M.,Potzel,D.,Sieber,G.,2009.HumanFactorsGuidelinefor SaferRoadInfrastructure.WorldRoadAssociation,Paris.
Bjørnskau,T.,1994. Hypothesesonriskcompensation.In:Proceedingsofthe Con-ferenceRoadSafetyinEurope,StrategicHighwayResearchProgram(SHRP), Lille,France,pp.81–98.
Brüde,U.,Larsson,J.,1993. Modelsforpredictingaccidentsatjunctionswhere pedestriansandcyclistsareinvolved.Howwelldotheyfit?AccidentAnalysis andPrevention25(5),499–509.
Brüde,U.,Larsson,J.,2000. Whatroundaboutdesignprovidesthehighestpossible safety?NordicRoadandTransportResearch12(2),17–21.
Buchanan,C.,1963.TrafficinTowns.HerMajesty’sStationeryOffice,London. Chorus,C.G.,Arentze,T.A.,Timmermans,H.J.P.,2008.Arandomregret-minimization
modeloftravelchoice.TransportationResearchPartB:Methodological42(1), 1–18.
Corben,B.,Senserrick,T.,Cameron,M.,Rechnitzer,G.,2004. Developmentofthe VisionaryResearchModel.MonashUniversityResearchCentre,Clayton. CROW,2007. DesignManualforBicycleTraffic.CROW,Ede.
Daniels,S.,Brijs,T.,Nuyts,E.,Wets,G.,2009. Injurycrasheswithbicyclistsat roundabouts:influenceofsomelocationcharacteristicsandthedesignofcycle facilities.JournalofSafetyResearch40(2),141–148.
Davidse,R.,2007.AssistingtheOlderDriver:IntersectionDesignandIn-carDevices toImprovetheSafetyoftheOlderDriver.InstituteforRoadSafetyResearch, Leidschendam.
DeHartog,J.J.,Boogaard,H.,Nijland,H.,Hoek,G.,2010. Dothehealth bene-fitsofcyclingoutweightherisks?EnvironmentalHealthPerspective118(8), 1109–1116.
DeWaard,D.,1996. TheMeasurementofDrivers’MentalWorkload.Groningen University,Groningen.
DeWaard,D.,Schepers,J.P.,Ormel,W.,Brookhuis,K.A.,2010. Mobilephoneuse whilecycling:incidenceandeffectsonbehaviourandsafety.Ergonomics53, 30–42.
Dijkstra,A.,2004. Rotondesmetvrijliggendefietspadenookveiligvoorfietsers? (AreRoundaboutswithSeparateCycleTracksalsoSafeforCyclists?).Institute forRoadSafetyResearch,Leidschendam.
Donges,E.,1978.Atwo-levelmodelofdriversteeringbehavior.HumanFactors20 (6),691–707.
Elvik,R.,2000. Whicharetherelevantcostsandbenefitsofroadsafetymeasures designedforpedestriansandcyclists?AccidentAnalysisandPrevention32(1), 37–45.
Elvik, R., 2004. To what extent can theory account for the findings of roadsafety evaluation studies? Accident Analysis andPrevention 36 (5), 841–849.
Elvik,R.,2006.Lawsofaccidentcausation.AccidentAnalysisandPrevention38(4), 742–747.
Elvik, R., 2009. The non-linearity of risk and the promotion of environ-mentally sustainable transport. Accident Analysis and Prevention 41 (4), 849–855.
Elvik,R.,2010. Whysomeroadsafetyproblemsaremoredifficulttosolvethan others.AccidentAnalysisandPrevention42(4),1089–1096.
Elvik,R.,2011. Publicationbiasandtime-trendbiasinmeta-analysisofbicycle helmetefficacy:are-analysisofattewell,glaseandMcFadden,2001.Accident AnalysisandPrevention43(3),1245–1251.
Elvik,R.,Bjørnskau,T.,2005. Howaccuratelydoesthepublicperceivedifferences intransportrisks?Anexploratoryanalysisofscalesrepresentingperceivedrisk. AccidentAnalysisandPrevention37(6),1005–1011.
Elvik,R.,Vaa,T.,2009.TheHandbookofRoadSafetyMeasures.Elsevier,Oxford. EuroNCAP, 2012.European NewCar AssessmentProgramme,Retrieved from:
http://www.euroncap.com/(28.10.12).
EuroRap, 2011. European Road Assessment Programme, Retrieved from: http://www.eurorap.org/(28.10.12).
Fitzpatrick,K.,Wooldridge,M.D.,Tsimhoni,O.,Collins,J.M.,Green,P.,Bauer,K., Parma,K.D.,Harwood,R.K.,Krammes,D.W.,Poggioli,R.B.,1999. Alternative DesignConsistencyRatingMethodsforTwo-LaneRuralHighways; fhwa-rd-99-172.FHWA,Washington,DC.
Ford,A.,2009. ModelingtheEnvironment.IslandPress,Washington,DC. Frank,L.,Hawkins,C.,2008. AssessingTravelandEnvironmentalImpactsof
Con-trastingLevelsofVehicularandPedestrianConnectivity:AssessingAspectsof theFusedGrid.CanadaMortgageandHousingCorporation,Ottawa. Fuller,R.,2005. Towardsageneraltheoryofdriverbehaviour.AccidentAnalysis
andPrevention37(3),461–472.
Gårder,P.,Leden,L.,Pulkkinen,U.,1998.Measuringthesafetyeffectofraisedbicycle crossingsusinganewresearchmethodology.TransportationResearchRecord 1636,64–70.
Gibson,J.J.,1961.Thecontributionofexperimentalpsychologytotheformulationof theproblemofsafety.In:Jacobs,H.H.(Ed.),BehaviouralApproachestoAccident Research.AssociationfortheAidofCrippledChildren,NewYork.
Godthelp,H.,Wouters,P.I.J.,1978. Koershoudenvoorfietsersenbromfietsers (TrackingansSteeringControlinCyclistsandMopedists).Verkeerskunde11, 537–543.
Goldenbeld,C.,1992. Ongevallenvanouderefietsersin1991(CrasheswithOlder Cyclists).InstituteforRoadSafetyResearch,Leidschendam.
Gommers,M.J.P.F.,Bovy,P.H.L.,1987.Routekeuzegedragennetwerkgebruik(Route ChoiceandNetworkUse).DelftUniversityofTechnology,Delft.
HaddonJr.,W.,1980.Advancesintheepidemiologyofinjuriesasabasisforpublic policy.PublicHealthReports95(5),411–421.
Hägerstrand,T.,1970. Whataboutpeopleinregionalscience?PapersinRegional Science24,6–21.
Hakkert,A.S.,Braimaister,L.,2002. TheUsesofExposureandRiskinRoadSafety Studies.InstituteforRoadSafetyResearch,Leidschendam.
Hatakka,M.,Keskinen,E.,Gregersen,N.,Glad,A.,1999. Theoriesandaimsof edu-cationalandtrainingmeasures.In:Siegrist,S.(Ed.),DriverTraining,Testingand Licencing.SwissCounsilforAccidentPrevention,Bern.
Heemskerk,M.,Wilson,K.,Pavao-Zuckerman,M.,2003.Conceptualmodelsastools forcommunicationacrossdisciplines.ConservationEcology7(3),8. Heinen,E.,Handy,S.,2012.Similaritiesinattitudesandnormsandtheeffecton
bicy-clecommuting:evidencefromthebicyclecitiesdavisanddelft.International JournalofSustainableTransportation6(5),257–281.
Heinen,E.,Maat,K.,VanWee,B.,2011. Day-to-daychoicetocommuteornotby bicycle.TransportationResearchRecord2230,9–18.
Heinen,E.,VanWee,B.,Maat,K.,2010. Commutingbybicycle:anoverviewofthe literature.TransportReviews30(1),59–96.
Heinrich, H.W.,1931. Industrial AccidentPrevention: AScientific Approach. McGraw-Hill,NewYork.
Houtenbos,M.,2008. ExpectingtheUnexpected:AStudyofInteractiveDriving BehaviouratIntersections.InstituteforRoadSafetyResearch,Leidschendam. Hummel,T.,2001.RouteManagementinSaferTransportationNetworkPlanning. Hunter,W.W.,Harkey,D.L.,Stewart,J.R., Birk,M.L.,2000. Evaluationofblue
bike-lanetreatmentinPortland,Oregon.TransportationResearchRecord1705, 107–115.
Jacobsen,P.L.,2003.Safetyinnumbers:morewalkersandbicyclists,saferwalking andbicycling.InjuryPrevention9,205–209.
Janke,M.K.,1991.Accidents,mileage,andtheexaggerationofrisk.AccidentAnalysis andPrevention23,183–188.
Jensen,S.U.,1999. PedestriansafetyinDenmark.TransportationResearchRecord 1674,61–69.
Kahneman,D.,Tversky,A.,1979.Prospecttheory:ananalysisofdecisionunderrisk. Econometrica47(2),263–291.
Kim,J.K.,Kim,S.,Ulfarsson,G.F.,Porrello,L.A.,2007. Bicyclistinjuryseverities inbicycle–motorvehicleaccidents.AccidentAnalysisandPrevention39(2), 238–251.
Kooijman,J.D.,Meijaard,J.P.,Papadopoulos,J.M.,Ruina,A.,Schwab,A.L.,2011. Abicyclecanbeself-stablewithoutgyroscopicorcastereffects.Science332 (6027),339–342.
Leibowitz,H.W.,Owens,D.A.,1977.Nighttimedrivingaccidentsandselectivevisual degradation.Science197(4302),422–423.
Leibowitz,H.W.,Post,R.B.,1982. Thetwomodesofprocessingconceptandsome implications.In:Beck,J.(Ed.),OrganizationandRepresentationinPerception. LawrenceErlbaum,Hillsdale,pp.343–363.
Leveson,N.,2004. Anewaccidentmodelforengineeringsafersystems.Safety Science42(4),237–270.
Lewis-Evans,B.,DeWaard,D.,Brookhuis,K.A.,2011. Speedmaintenanceunder cognitiveload–implicationsfortheoriesofdriverbehaviour.AccidentAnalysis andPrevention43(4),1497–1507.
Lie,A.,Tingvall,C.,2002. Howdoeuroncapresultscorrelatewithreal-lifeinjury risks?Apairedcomparisonstudyofcar-to-carcrashes.TrafficInjuryPrevention 3(4),288–293.
Loo,B.P.Y.,Tsui,K.L.,2010. Bicyclecrashcasualtiesinahighlymotorizedcity. AccidentAnalysisandPrevention42(6),1902–1907.
McFadden,D.,1974.Conditionallogitanalysisofqualitativechoice-behaviour.In: Zarembka,P.(Ed.),FrontiersinEconometrics.AcademicPress,NewYork,NY, pp.105–142.
Michon,J.A.,1985. Acriticalviewofdriverbehaviormodels.Whatdoweknow, whatshouldwedo?In:Evans,L.,Schwing,R.C.(Eds.),HumanBehaviorand TrafficSafety.PlenumPress,NewYork.
Mokhtarian,P.L.,Chen,C.,2004. Ttbornotttb,thatisthequestion:areviewand analysisoftheempiricalliteratureontraveltime(andmoney)budgets. Trans-portationResearchPartA:PolicyandPractice38(9–10),643–675.
Moore,J.K.,Hubbard,M.,Kooijman,J.D.G.,Schwab,A.L.,2009. Amethodfor esti-matingphysicalpropertiesofacombinedbicycleandrider.In:Proceedingsof theASMEConference,SanDiego,CA.
Moudon,A.V.,Lee,C.,Cheadle,A.D.,Collier,C.W.,Johnson,D.,Schmid,T.L.,Weather, R.D.,2005.Cyclingandthebuiltenvironment,ausperspective.Transportation ResearchPartD:TransportandEnvironment10(3),245–261.
Näätänen,R., Summala,H., 1974. Amodel forthe roleof motivational fac-torsindrivers’decision-making.AccidentAnalysisandPrevention6(3–4), 243–261.
Nyberg,P.,Bjornstig,U.,Bygren,L.O.,1996. Roadcharacteristicsandbicycle acci-dents.ScandinavianJournalofSocialMedicine24(4),293–301.
Oja,P.,Titze,S.,Bauman,A.,DeGeus,B.,Krenn,P.,Reger-Nash,B.,Kohlberger,T., 2011. Healthbenefitsofcycling:asystematicreview.ScandinavianJournalof Medicine&ScienceinSports21(4),496–509.
Ormel,W.,KleinWolt,K.,DenHertog,P.,2008. Enkelvoudigefietsongevallen (Single-BicycleCrashes).StichtingConsumentenVeiligheid,Amsterdam. Othman,S.,Thomson,R.,Lanner,G.,2009.Identifyingcriticalroadgeometry
param-etersaffectingcrashrateandcrashtype.AnnalsofAdvancesinAutomotive Medicine53,155–165.
Parkin,J.,Wardman,M.,Page,M.,2007.Modelsofperceivedcyclingriskandroute acceptability.AccidentAnalysisandPrevention39(2),364–371.
Previc,F.H.,1998.Theneuropsychologyof3-dspace.PsychologicalBulletin124(2), 123–164.
Pucher, J., Buehler, R., 2008. Making cycling irresistible: lessons from the Netherlands, Denmark and Germany. Transport Reviews 28 (4), 495–528.
Ranney,T.A.,1994.Modelsofdrivingbehavior:areviewoftheirevolution.Accident AnalysisandPrevention26(6),733–750.
Räsänen,M.,Summala,H.,1998.Attentionandexpectationproblemsinbicycle–car collisions: an in-depth study. Accident Analysis and Prevention 30 (5), 657–666.
Reurings,M.C.B.,2010. Hoegevaarlijkisfietseninhetdonker?(HowDangerousis CyclinginDarness?).InstituteforRoadSafetyResearch,Leidschendam. Reynolds,C.C.,Harris,M.A.,Teschke,K.,Cripton,P.A.,Winters,M.,2009.Theimpact
oftransportationinfrastructureonbicyclinginjuriesandcrashes:areviewof theliterature.EnvironmentalHealth8,47.
Rietveld,P.,Daniel,V.,2004. Determinantsofbicycleuse:domunicipal poli-cies matter? Transportation Research PartA: Policy and Practice 38 (7), 531–550.
Rodarius,C.,Mordaka,J.,Versmissen,T.,2008.BicycleSafetyinBicycletoCar Acci-dents.TNOScienceandIndustry,Delft.
Rosén,E.,Stigson,H.,Sander,U.,2011.Literaturereviewofpedestrianfatalityriskas afunctionofcarimpactspeed.AccidentAnalysisandPrevention43(1),25–33. Rumar,K.,1999. TransportSafetyVisions,TargetsandStrategies:Beyond2000.
EuropeanTransportSafetyCouncil,Brussels.
Sakshaug, L., Laureshyn, A., Svensson, Å., Hydén, C., 2010. Cyclists in roundabouts—differentdesignsolutions.AccidentAnalysisandPrevention42 (4),1338–1351.
Schepers,J.P.,2012.Doesmorecyclingalsoreducetheriskofsingle-bicyclecrashes? InjuryPrevention18(4),240–245.
Schepers,J.P.,DenBrinker,B.P.L.M.,2011. Whatdocyclistsneedtoseetoavoid single-bicyclecrashes?Ergonomics54(4),315–327.
Schepers,J.P.,Heinen, E.,2013. Howdoes amodalshiftfromshortcartrips to cycling affect road safety? Accident Analysis and Prevention 50 (1), 1118–1127.
Schepers,J.P.,KleinWolt,K.,2012. Single-bicyclecrashtypesandcharacteristics. CyclingResearchInternational2,119–135.
Schepers, J.P., Kroeze, P.A., Sweers, W., Wust, J.C., 2011. Road factors and bicycle–motorvehiclecrashesatunsignalizedpriorityintersections.Accident AnalysisandPrevention43(3),853–861.
Schieber,F.,Schlorholtz,B.,Mccall,R.,2008.Visualrequirementsofvehicular guid-ance.In:Castro,C.(Ed.),HumanFactorsofVisualandCognitivePerformancein Driving.CRCPress,BocaRaton.
Schoon,C.C.,2006.Deproblematiekvanrechtsafstaandevrachtauto’s(TheProblem ofLorriesTurningRight).InstituteforRoadSafetyResearch,Leidschendam. Sobhani,A.,Young,W.,Logan,D.,Bahrololoom,S.,2011. Akineticenergymodel
oftwo-vehiclecrashinjuryseverity.AccidentAnalysisandPrevention43(3), 741–754.
Southworth,M.,2005. Designingthewalkablecity.JournalofUrbanPlanningand Development131(4),246–257.
Stipdonk,H.,Reurings,M.,2012.Theeffectonroadsafetyofamodalshiftfromcar tobicycle.TrafficInjuryPrevention13(4),412–421.
Summala,H.,Pasanen,E.,Räsänen,M.,Sievänen,J.,1996. Bicycleaccidentsand drivers’visualsearchatleftandrightturns.AccidentAnalysisandPrevention 28(2),147–153.
SWOV,2012. FactsheetAttentionProblemsBehindtheWheel.InstituteforRoad SafetyResearch,Leidschendam.
Theeuwes,J.,Godthelp,H.,1995. Self-explainingroads.SafetyScience19(2–3), 217–225.
Theeuwes,J.,Hagenzieker,M.P.,1993. Visualsearchoftrafficscenes:ontheeffect oflocationexpectations.In:Gale,A.G.,Freeman,M.H.,Haslegrave,C.M.,Smith, P.,Taylor,S.P.(Eds.),VisioninVehicleIV.ElsevierScience,Amsterdam. Tingvall,C.,Haworth,N.,1999. Visionzero–anethicalapproachtosafetyand
mobility.In:6thITEInternationalConferenceRoadSafety&TrafficEnforcement: Beyond2000,Melbourne.
Toft,Y.,Dell,G.,Klockner,K.K.,Hutton,A.,2012.ModelsofCausation:Safety.Safety InstituteofAustralia,Tullamarine,Victoria.
VanBoggelen,O.,Schepers,J.P.,Kroeze,P.A.,VanDerVoet,M.,2011. Fietsberaad-publicatie19:Samenwerkenaaneenveiligefietsomgeving(Collaboratingto ImproveCyclingSafety).Fietsberaad,Utrecht.
VandeKaa,E.J.,2010.Prospecttheoryandchoicebehaviourstrategies:reviewand synthesisofconceptsfromsocialandtransportsciences.EuropeanJournalof TransportandInfrastructureResearch10(4),299–329.
VanWee,B.,2009. Verkeerentransport.In:VanWee,B.,AnneAnnema,J.(Eds.), Verkeerenvervoerinhoofdlijnen(OutliningTrafficandTransport).Coutinho, Bussum.
VanWee,B.,Maat,K.,2003. Land-useandtransport:areviewanddiscussionof dutchresearch.EuropeanJournalofTransportandInfrastructureResearch3 (2),199–218.
VanWee,B.,Rietveld,P.,Meurs,H.,2006. Isaveragedailytraveltimeexpenditure constant?Insearchofexplanationsforanincreaseinaveragetraveltime.Journal ofTransportGeography14(2),109–122.
VanWee,G.P.,Hagenzieker,M.P.,Wijnen,W.Whichindicatorstoincludeinthe exanteevaluationsof thesafetyeffects ofpolicyoptions?Gapsin eval-uationanda discussionbasedonan ethicalperspective. TransportPolicy, inpress.
Vandenbulcke-Plasschaert,G.,2011. SpatialAnalysisofBicycleUseandAccident RisksforCyclists.UniversitécatholiquedeLouvain,Louvain.
Viner,D.,1991.AccidentAnalysisandRiskControl.DerekVinerLtd.,Melbourne. Vlakveld,W.P.,2005.Jongebeginnendeautomobilisten,hunongevalsrisicoen
maa-tregelenomditterugtedringen(YoungNoviceDrivers,theirCrashRiskand PreventiveMeasures).InstituteforRoadSafetyResearch,Leidschendam. Vlakveld,W.P.,Goldenbeld,C.,Twisk,D.A.M.,2008. Belevingvan
verkeersonvei-ligheid(PerceptionofLackofTrafficSafety).InstituteforRoadSafetyResearch, Leidschendam.
Vlakveld,W.P.,Louwerse,W.J.R.,2011. Derelatietussenroadprotectionscores (rps)enhetslachtofferrisicoopwegvakkenvanprovincialewegeninutrecht (TheRelationBetweenRoadProtectionScores(rps)andtheCasualtyRateat RoadSectionsintheProvinceofUtrecht).InstituteforRoadSafetyResearch, Leidschendam.
Vlek,C.,Jager,W.,Steg,L.,1997.Modellenenstrategieënvoorgedragsverandering terverminderingvancollectieverisico’s(ModelsandStrategiesforBehavioural ChangeinOrdertoReduceCollectiveRisks).Nederlandstijdschriftvoorde Psy-chologie52,174–191.
Wegman,F.C.M.,Aarts,L.T.,2006. AdvancingSustainableSafety:NationalRoad SafetyOutlookfor2005–2020.InstituteforRoadSafetyResearch,Leidschendam. Wegman,F.C.M.,Zhang,F.,Dijkstra,A.,2012. Howtomakemorecyclinggoodfor
roadsafety?AccidentAnalysisandPrevention44(1),19–29.
Weller,G.,Schlag,B.,2007. DriverBehaviourModel:Ripcord-InserestDeliverable d8.TechnischeUniversitätDresden,Dresden.
Wilde,G.J.S.,1982. Thetheoryofriskhomeostasis:implicationsforsafetyand health.RiskAnalysis2(4),209–225.
Zacharias,J.,2005.Non-motorizedtransportationinfourshanghaidistricts. Inter-nationalPlanningStudies10(3–4),323–340.
