A Human Error Analysis of Commercial Aviation Accidents Using the Human Factors Analysis and Classification System (HFACS)PDF

7/28/2019 A Human Error Analysis of Commercial Aviation Accidents Using the Human Factors Analysis and Classification Sys

1/20

DOT/FAA/AM-01/3

OfficeofAviationMedicineWashington,D.C.20591

A Human Error Analysis ofCommercial Aviation AccidentsUsing the Human Factors

Analysis and ClassificationSystem (HFACS)

DouglasA.WiegmannUniversityofIllinoisatUrbana-ChampaignInstituteofAviationSavoy,IL61874

ScottA.ShappellFAACivilAeromedicalInstituteP.O.Box25082OklahomaCity,OK73125

February2001

FinalReport

Thisdocumentisavailabletothepublic

throughtheNationalTechnicalInformation

Service,Springfield,Virginia 22161.

U.S. Departmentof Transpor tation

Federal Aviation

Administration


2/20

NONONONOTICETICETICETICENOTICE

ThisdocumentisdisseminatedunderthesponsorshipoftheU.S.DepartmentofTransportationintheinterestofinformationexchange.TheUnitedStatesGovernment

assumesnoliabilityforthecontentsthereof.


3/20

TechnicalReportDocumentationPage

1. Report No.

DOT/FAA/AM-01/3

2. Government Accession No. 3. Recipient'sCatalog No.

4. Title and Subtitle

A Human Error Analysis of Commercial Aviation AccidentsUsing the Human Factors Analysis and Classification System (HFACS)

5. Report Date

February 2001

6. Performing OrganizationCode

7. Author(s)

Wiegmann,D.A.1, andShappell, S.A.28. Performing OrganizationReportNo.

9. Performing OrganizationName and Address

1University of Illinois at Urbana-Champaign, Institute of Aviation,

Savoy, IL 618742FAA Civil Aeromedical Institute, P.O. Box 25082, Oklahoma City, OK 73125

10. Work Unit No. (TRAIS)

11. Contract orG rantNo.

99-G-00612. Sponsoring Agency name and Address

Office of Aviation Medicine

Federal Aviation Administration800 Independence Ave., S.W.Washington, DC 20591

13. Type of Report and Period Covered

14. Sponsoring Agency Code

15. Supplemental Notes

Work was accomplished under task # AAM-A-00-HRR-520.16. Abstract

The Human Factors Analysis and Classification System (HFACS) is a general human error frameworkoriginally developed and tested within the U.S. military as a tool for investigating and analyzing the humancauses of aviation accidents. Based upon Reasons (1990) model of latent and active failures, HFACSaddresses human error at all levels of the system, including the condition of aircrew and organizationalfactors. The purpose of the present study was to assess the utility of the HFACS framework as an erroranalysis and classification tool outside the military. Specifically, HFACS was applied to commercial aviationaccident records maintained by the National Transportation Safety Board (NTSB). Using accidents thatoccurred between January 1990 and December 1996, it was demonstrated that HFACS reliablyaccommodated all human causal factors associated with the commercial accidents examined. In addition, theclassification of data using HFACS highlighted several critical safety issues in need of intervention research.These results demonstrate that the HFACS framework can be a viable tool for use within the civil aviationarena.

17. KeyWordsAviation, Human Error, Accident Investigation,Database Analysis, Commercial Aviation

18. Distribution StatementDocument is available to the public through theNational Technical Information Service,Springfield, Virginia 22161

19. SecurityClassif. (of this report)

Unclassified20. SecurityClassif. (of thispage)

Unclassified21. No. of Pages

1722. Price

FormDOTF1700.7 (8-72) Reproductionofcompletedpageauthorized

i


4/20


5/20

ACKNOWLEDGMENTS

The authors thank Frank Cristina and AnthonyPape for their assistance in gathering,

organizingandanalyzingtheaccidentreportsusedinthisstudy.

iii


6/20


7/20

A HUMAN ERRORANALYSIS OF COMMERCIALAVIATION ACCIDENTS USING THE

HUMAN FACTORS ANALYSIS AND CLASSIFICATION SYSTEM (HFACS)

INTRODUCTION

Humans,bytheirverynature,makemistakes;there-fore,itshouldcomeasnosurprisethathumanerrorhasbeenimplicatedinavarietyofoccupationalaccidents,including70%to80%ofthoseincivilandmilitaryaviation (OHare,Wiggins,Batt,&Morrison,1994;Wiegmann and Shappell, 1999; Yacavone, 1993). Infact,whilethenumberofaviationaccidentsattributablesolelytomechanicalfailurehasdecreasedmarkedlyoverthepast40years,thoseattributableatleastinpartto

humanerrorhavedeclinedatamuchslowerrate(Shappell&Wiegmann,1996). Given such findings, itwouldappearthatinterventionsaimedatreducingtheoccur-renceorconsequencesofhumanerrorhavenotbeenaseffectiveasthosedirectedatmechanicalfailures.Clearly,ifaccidentsaretobereducedfurther,moreemphasismustbeplacedonthegenesisofhumanerrorasitrelatestoaccidentcausation.Theprevailingmeansofinvestigatinghumanerrorin

aviationaccidentsremainstheanalysisofaccidentandincidentdata.Unfortunately,mostaccidentreportingsystemsarenotdesignedaroundanytheoreticalframe-workofhumanerror. Indeed,mostaccidentreportingsystemsaredesignedandemployedbyengineersandfront-lineoperatorswithonlylimitedbackgroundsinhumanfactors.Asaresult,thesesystemshavebeenusefulforidentifyingengineeringandmechanicalfailuresbutare relatively ineffective and narrow in scope wherehumanerrorexists.Evenwhenhumanfactorsaread-dressed,thetermsandvariablesusedareoftenill-definedand archivaldatabasesarepoorlyorganized.The endresultsarepost-accidentdatabasesthattypicallyarenotconducivetoatraditionalhumanerroranalysis,making

the identification of intervention strategies onerous(Wiegmann&Shappell,1997).

The Accident Investigation ProcessTofurtherillustratethispoint,letusexaminethe

accident investigation and intervention process sepa-ratelyforthemechanicalandhumancomponentsofanaccident.Consider first the occurrence of anaircraftsystemormechanicalfailurethatresultsinanaccidentor

injury(Figure1).Asubsequentinvestigationtakesplacethatincludestheexaminationofobjectiveandquantifi-ableinformation,suchasthatderivedfromthewreckageandflightdatarecorder,aswellasthatfromtheapplica-tionofsophisticatedanalyticaltechniqueslikemetallur-gical tests and computer modeling. This kind ofinformation is then used to determine the probablemechanicalcause(s)oftheaccidentandtoidentifysafetyrecommendations.Uponcompletionoftheinvestigation,thisobjec-

tive information is typically entered into a highly-

structuredandwell-defined accident database.Thesedata can then be periodically analyzed to determinesystem-widesafetyissuesandprovidefeedbacktoinves-tigators,therebyimprovinginvestigativemethodsandtechniques.Inaddition,thedataareoftenusedtoguideorganizations(e.g.,theFederalAviationAdministration[FAA],NationalAeronauticsandSpaceAdministration[NASA],DepartmentofDefense[DoD],airplanemanu-facturers and airlines) in deciding which research orsafetyprograms to sponsor. As a result, theseneeds-based, data-driven programs, in turn, have typicallyproduced effective intervention strategies that eitherpreventmechanicalfailuresfromoccurringaltogether,ormitigatetheirconsequenceswhentheydohappen.Ineithercase,therehasbeenasubstantialreductionintherateofaccidentsduetomechanicalorsystemsfailures.In stark contrast, Figure 2 illustrates the current

human factors accident investigation andpreventionprocess.Thisexamplebeginswiththeoccurrenceofanaircrewerrorduringflightoperationsthatleadstoanaccidentorincident.Ahumanperformanceinvestiga-tionthenensuestodeterminethenatureandcausesofsucherrors.However,unlikethetangibleandquantifi-

ableevidencesurroundingmechanicalfailures,theevi-denceandcausesofhumanerroraregenerallyqualitativeandelusive.Furthermore,humanfactorsinvestigativeand analytical techniques are often less refined andsophisticatedthanthoseusedtoanalyzemechanicalandengineering concerns.As such, the determination ofhumanfactorscausaltotheaccidentisatenuouspracticeatbest;allofwhichmakestheinformationenteredintheaccidentdatabasesparseandill-defined.

1.


8/20

Feedback

MechanicalFailure

- Catastrophicfailuresareinfrequentevents

- Whenfailuresdooccur,theyareoftenlesssevereorhazardousduetoeffectiveinterventionprograms.

Data-DrivenResearch

ResearchSponsors- FAA,DoD,NASA,&airplane

manufacturers provideresearchfunding.

- Research programsareneeds-basedanddata-driven.Interventions

are

therefore

veryeffective.

AccidentInvestigation

- Highlysophisticatedtechniquesandprocedures

- Informationisobjectiveandquantifiable

- Effectiveatdeterminingwhythefailureoccurred

AccidentDatabase

- Designedaroundtraditionalcategories

- Variablesarewell-definedandcausallyrelated

- Organizationandstructurefacilitateaccessanduse

DatabaseAnalysis

- Traditionalanalysesareclearlyoutlinedandreadilyperformed.

- Frequentanalyseshelp identifycommonmechanicalandengineeringsafetyissues.

Mitigation

Preve

ntion

EffectiveIntervention

andPreventionPrograms

Figure 1. General process of investigating and preventing aviation accidents involving mechanical orsystems failures.

As a result,when traditional data analyses areper-formedtodeterminecommonhumanfactorsproblems

acrossaccidents,theinterpretationofthefindingsandthesubsequentidentificationofimportantsafetyissuesareof limitedpractical use. Tomakematters worse,resultsfromtheseanalysesprovidelimitedfeedbacktoinvestigatorsandareoflimitedusetoairlinesandgovern-mentagenciesindeterminingthetypesofresearchorsafety programs to sponsor. As such, many researchprogramstendtobeintuitively-,orfad-driven,ratherthan data-driven, and typically produce interventionstrategiesthatareonlymarginallyeffectiveatreducingtheoccurrenceandconsequenceofhumanerror.Theoverallrateofhuman-errorrelatedaccidents,therefore,

hasremainedrelativelyhighandconstantoverthelastseveralyears(Shappell&Wiegmann,1996).

Addressing the ProblemIftheFAAandtheaviationindustryaretoachieve

theirgoalofsignificantlyreducingtheaviationaccidentrateoverthenexttenyears,theprimarycausesofaviationaccidents(i.e.,humanfactors)mustbeaddressed(ICAO,1993). However, as illustrated in Figure 2, simply

increasingtheamountofmoneyandresourcesspentonhumanfactorsresearchisnotthesolution.Indeed,a

greatdealofresourcesandeffortsarecurrentlybeingexpended.Rather,thesolutionistoredirectsafetyeffortsso that they address important human factors issues.However,thisassumesthatweknowwhattheimportanthuman factors issues are. Therefore, before researcheffortscanbesystematicallyrefocused,acomprehensiveanalysisofexistingdatabasesneedstobeconductedtodeterminethosespecifichumanfactorsresponsibleforaviationaccidentsandincidents.Furthermore,iftheseeffortsaretobesustained,newinvestigativemethodsandtechniqueswillneedtobedevelopedsothatdatagath-eredduringhumanfactorsaccidentinvestigationscanbe

improvedandanalysisoftheunderlyingcausesofhumanerrorfacilitated.Toaccomplishthisimprovement,ageneralhuman

errorframeworkisneededaroundwhichnewinvestiga-tivemethodscanbedesignedandexistingpostaccidentdatabasesrestructured.Previousattemptstodothishavemetwithencouraging,yetlimited,success(OHare,etal.,1994;Wiegmann&Shappell,1997).Thisisprima-rily because performance failures are influenced bya

2.


9/20

Feedback

HumanError

- Errorsoccurfrequentlyandarethemajorcauseofaccidents.

- Fewsafetyprogramsareeffectiveatpreventingtheoccurrenceorconsequencesoftheseerrors.

ResearchSponsors- FAA,DoD,NASA,&Airlines

providefundingforsafetyresearchprograms.

- Lackofgooddataleadstoresearch programsbasedprimarily

on

interests

and

intuitions.Interventionsarethereforelesseffective.

Fad-DrivenResearch

Mitigation

Preve

ntion

IneffectiveInterventionandPreventionPrograms

AccidentInvestigation

- Lesssophisticatedtechniquesandprocedures

- Informationisqualitativeandillusive

- Focusonwhathappenedbutnotwhyithappened

AccidentDatabase

- Notdesignedaroundanyparticularhumanerrorframework

- Variablesoftenill-defined

- Organizationandstructuredifficulttounderstand

DatabaseAnalysis

- Traditionalhumanfactorsanalysesareonerousduetoill-definedvariablesanddatabasestructures.

- Fewanalyseshavebeenperformedtoidentifyunderlyinghumanfactorssafetyissues.

Figure 2. General process of investigating and preventing aviation accidents involving human error.

varietyofhumanfactorsthataretypicallynotaddressedbytraditionalerrorframeworks.Forinstance,withfew

exceptions(e.g.,Rasmussen,1982),humanerrortax-onomiesdonotconsiderthepotentialadversementalandphysiologicalconditionoftheindividual(e.g.,fa-tigue,illness,attitudes)whendescribingerrorsinthecockpit.Likewise,latenterrorscommittedbyofficialswithinthemanagementhierarchysuchaslinemanagersandsupervisorsareoftennotaddressed,eventhoughitiswell known that these factors directly influence theconditionanddecisionsofpilots(Reason,1990).There-fore,ifacomprehensiveanalysisofhumanerroristobeconducted, a taxonomy that takes into account themultiplecausesofhumanfailuremustbeoffered.

Recently,theHumanFactorsAnalysisandClassifica-tionSystem(HFACS)wasdevelopedtomeettheseneeds(Shappell&Wiegmann,1997a,2000a,andinpress).Thissystem,whichisbasedonReasons(1990)modeloflatentandactivefailures,wasoriginallydevelopedfortheU.S.NavyandMarineCorpsasanaccidentinvestigationanddataanalysistool.Sinceitsoriginaldevelopment,however,HFACShasbeenemployedbyothermilitary

organizations(e.g.,U.S.Army,AirForce,andCanadianDefenseForce) as anadjunct topreexistingaccidentinvestigationandanalysissystems.Todate,theHFACSframeworkhasbeenappliedtomorethan1,000militaryaviationaccidents,yieldingobjective,data-driveninter-ventionstrategieswhileenhancingboththequantityandqualityofhumanfactorsinformationgatheredduringaccidentinvestigations(Shappell&Wiegmann,inpress).OtherorganizationssuchastheFAAandNASAhave

exploredtheuseofHFACSasacomplementtopreexist-ingsystemswithincivilaviationinanattempttocapital-izeongainsrealizedbythemilitary(Ford,Jack,Crisp,&Sandusky,1999).Still,fewsystematiceffortshaveexam-inedwhetherHFACSisindeedaviabletoolwithinthe

civilaviationarena,eventhoughitcanbearguedthatthesimilaritiesbetweenmilitaryandcivilianaviationout-weightheirdifferences.Thepurposeofthepresentstudywas to empirically address this issue by applying theHFACSframework,asoriginallydesignedforthemili-tary,totheclassificationandanalysisofcivilaviationaccidentdata.Beforebeginning,however,abriefover-viewoftheHFACSsystemwillbepresentedforthose

3.


10/20

readerswhomaynotbefamiliarwiththeframework(fora detailed description of HFACS, see Shappell andWiegmann,2000aand2001).

HFACSDrawinguponReasons(1990)conceptoflatentand

activefailures,HFACSdescribeshumanerrorateachoffourlevelsoffailure:1)unsafeactsofoperators(e.g.,aircrew), 2) preconditions for unsafe acts, 3) unsafesupervision, and4)organizational influences.A briefdescriptionofeachcausalcategoryfollows(Figure3).

Unsafe Acts of OperatorsTheunsafeactsofoperators(aircrew)canbeloosely

classifiedintooneoftwocategories:errorsandviolations(Reason,1990).Whilebotharecommonwithinmostsettings,theydiffermarkedlywhentherulesandregula-tionofanorganizationareconsidered.Thatis,errorscanbedescribedasthoselegalactivitiesthatfailtoachieve

theirintendedoutcome,whileviolationsarecommonlydefinedasbehaviorthatrepresentsthewillfuldisregardfor the rules and regulations. It iswithin these twooverarchingcategoriesthatHFACSdescribesthreetypesoferrors(decision,skill-based,andperceptual)andtwotypesofviolations(routineandexceptional).

ErrorsOneofthemorecommonerrorforms,decision errors,

representsconscious,goal-intendedbehaviorthatpro-ceedsasdesigned;yet,theplanprovesinadequateorinappropriate for the situation.Often referred to ashonestmistakes,theseunsafeactstypicallymanifestaspoorlyexecutedprocedures,improperchoices,orsimplythemisinterpretationormisuseofrelevantinformation.Incontrasttodecisionerrors,theseconderrorform,

skill-based errors,occurswithlittleornoconsciousthought.Just as little thought goes into turning ones steeringwheel or shifting gears in an automobile, basic flight

PerceptualErrorsSkill-BasedErrors

UNSAFEACTS

Errors

DecisionErrors ExceptionalRoutine

Violations

InadequateSupervision

PlannedInappropriate

OperationsFailedtoCorrectProblem

SupervisoryViolations

UNSAFESUPERVISION

PRECONDITIONSFOR

UNSAFEACTS

SubstandardConditionsof

Operators

PRECONDITIONSFOR

UNSAFEACTS

AdversePhysiologicalStates

Physical/Mental

LimitationsAdverseMental

States PersonalReadinessCrewResourceMismanagement

SubstandardPracticesofOperators

ResourceManagement

OrganizationalClimate OrganizationalProcess

ORGANIZATIONALINFLUENCES

Figure 3. Overview of the Human Factors Analysis and Classification System (HFACS).

4.


11/20

skillssuchasstickandruddermovementsandvisualscanningoftenoccurwithoutthinking.Thedifficultywith these highly practiced and seemingly automaticbehaviors is that they are particularly susceptible toattentionand/ormemoryfailures.Asaresult,skill-basederrorssuchasthebreakdowninvisualscanpatterns,

inadvertentactivation/deactivationofswitches,forgot-tenintentions, andomitted items inchecklists oftenappear.Eventhemanner(orskill)withwhichonefliesanaircraft(aggressive,tentative,orcontrolled)canaffectsafety.While, decision and skill-based errors have domi-

natedmostaccidentdatabasesandtherefore,havebeenincludedinmosterrorframeworks,thethirdandfinalerrorform,perceptual errors,hasreceivedcomparativelylessattention.Nolessimportant,perceptualerrorsoccurwhensensoryinputisdegraded,orunusual,asisoftenthecasewhenflyingatnight,intheweather,orinothervisuallyimpoverishedenvironments.Facedwithactingonimperfectorlessinformation,aircrewruntheriskofmisjudgingdistances,altitude,anddecentrates,aswellasarespondingincorrectlytoavarietyofvisual/vestibu-larillusions.

ViolationsAlthoughtherearemanywaystodistinguishamong

typesofviolations,twodistinctformshavebeenidenti-fiedbasedontheiretiology.Thefirst,routine violations,tendtobehabitualbynatureandareoftenenabledbya

systemof supervision andmanagement that toleratessuchdeparturesfromtherules(Reason,1990).Oftenreferredtoasbendingtherules,theclassicexampleisthat of theindividual who driveshis/her automobileconsistently5-10mphfasterthanallowedbylaw.Whileclearlyagainstthelaw,thebehavioris,ineffect,sanc-tionedbylocalauthorities(police)whooftenwillnotenforcethelawuntilspeedsinexcessof10mphoverthepostedlimitareobserved.

Exceptional violations,ontheotherhand,areisolateddeparturesfromauthority,neithertypicaloftheindi-vidual nor condoned bymanagement. For example,

whiledriving65ina55mphzonemightbecondonedbyauthorities,driving105mphina55mphzonecertainlywould not. It is important to note, that while mostexceptionalviolationsareappalling,theyarenotconsid-eredexceptionalbecauseoftheirextremenature.Rather,theyareregardedasexceptionalbecausetheyareneithertypicaloftheindividualnorcondonedbyauthority.

Preconditions for Unsafe ActsSimplyfocusingonunsafeacts,however,islikefocus-

ingonapatientssymptomswithoutunderstandingtheunderlyingdiseasestatethatcausedit.Assuch,investi-gatorsmustdigdeeperintothepreconditionsforunsafeacts.WithinHFACS,twomajorsubdivisionsarede-

scribed: substandard conditionsof operators and thesubstandardpracticestheycommit.

Substandard Conditions of the OperatorBeingpreparedmentally is critical innearly every

endeavor;perhapsitisevenmoresoinaviation.Withthisinmind,thefirstofthreecategories,adverse mentalstates,wascreatedtoaccountforthosementalconditionsthatadverselyaffectperformance.Principalamongtheseare the loss of situational awareness,mental fatigue,circadiandysrhythmia,andperniciousattitudessuchasoverconfidence,complacency,andmisplacedmotiva-tionthatnegativelyimpactdecisionsandcontributetounsafeacts.Equallyimportant,however,arethose adverse physi

ological statesthatprecludethesafeconductofflight.Particularlyimportanttoaviationareconditionssuchasspatialdisorientation,visualillusions,hypoxia,illness,intoxication,andawholehostofpharmacologicalandmedicalabnormalitiesknowntoaffectperformance.Forexample,itisnotsurprisingthat,whenaircrewsbecomespatiallydisorientedandfailtorelyonflightinstrumen-tation,accidentscan,andoftendo,occur.

Physical and/or mental limitationsoftheoperator,thethirdandfinalcategoryofsubstandardcondition,in-cludesthoseinstanceswhennecessarysensoryinforma-tion is either unavailable, or if available, individualssimplydonothavetheaptitude,skill,ortimetosafelydealwithit.Foraviation,theformeroftenincludesnotseeingotheraircraftorobstaclesduetothesizeand/orcontrastoftheobjectinthevisualfield.However,therearemany timeswhen a situationrequires such rapidmentalprocessingorreactiontimethatthetimeallottedtoremedytheproblemexceedshumanlimits(asisoftenthecaseduringnap-of-the-earthflight).Nevertheless,

evenwhenfavorablevisualcuesoranabundanceoftimeisavailable,thereareinstanceswhenanindividualsimplymaynotpossessthenecessaryaptitude,physicalability,orproficiencytooperatesafely.

5.


12/20

Substandard Practices of the OperatorOftentimes,thesubstandardpracticesofaircrewwill

leadtotheconditionsandunsafeactsdescribedabove.Forinstance,thefailuretoensurethatallmembersofthecrewareactinginacoordinatedmannercanleadtoconfusion(adversementalstate)andpoordecisionsin

thecockpit.Crew resource mismanagement,asitisre-ferredtohere,includesthefailuresofbothinter-andintra-cockpitcommunication,aswellascommunicationwithATCandothergroundpersonnel.Thiscategoryalsoincludesthoseinstanceswhencrewmembersdonotwork together asa team,orwhen individualsdirectlyresponsiblefortheconductofoperationsfailtocoordi-nateactivitiesbefore,during,andafteraflight.Equallyimportant,however,individualsmustensure

that they are adequately prepared for flight.Conse-quently,thecategoryofpersonal readinesswascreatedtoaccountforthoseinstanceswhenrulessuchasdisregard-ingcrewrestrequirements,violatingalcoholrestrictions,orself-medicating,arenotadheredto.However,evenbehaviorsthatdonotnecessarilyviolateexistingrulesorregulations(e.g.,runningtenmilesbeforepilotinganaircraft ornotobservinggooddietary practices)mayreducetheoperatingcapabilitiesoftheindividualandare,therefore,capturedhere.

Unsafe SupervisionClearly,aircrewsareresponsiblefortheiractionsand,

assuch,mustbeheldaccountable.However,inmany

instances, they are theunwitting inheritors of latentfailuresattributabletothosewhosupervisethem(Rea-son, 1990). To account for these latent failures, theoverarchingcategoryofunsafesupervisionwascreatedwithinwhich four categories (inadequate supervision,plannedinappropriateoperations,failedtocorrectknownproblems,andsupervisoryviolations)areincluded.The firstcategory, inadequate supervision,refersto

failureswithinthesupervisorychainofcommand,whichwasadirectresultofsomesupervisoryactionorinaction.Thatis,ataminimum,supervisorsmustprovidetheopportunityforindividualstosucceed.Itisexpected,

therefore,thatindividualswillreceiveadequatetraining,professionalguidance,oversight,andoperationalleader-ship,andthatallwillbemanagedappropriately.Whenthisisnotthecase,aircrewsareoftenisolated,astheriskassociatedwith day-to-day operations invariably willincrease.However,theriskassociatedwithsupervisoryfailures

cancomeinmanyforms.Occasionally,forexample,theoperationaltempoand/orscheduleisplannedsuchthat

individualsareputatunacceptableriskand,ultimately,performanceisadverselyaffected.Assuch,thecategoryofplanned inappropriate operationswas created to ac-countforallaspectsofimproperorinappropriatecrewschedulingandoperationalplanning,whichmayfocusonsuchissuesascrewpairing,crewrest,andmanaging

theriskassociatedwithspecificflights.Theremainingtwocategoriesofunsafesupervision,

thefailure to correct known problems and supervisoryviolations,aresimilar,yetconsideredseparatelywithinHFACS.Thefailuretocorrectknownproblemsreferstothose instanceswhendeficiencies among individuals,equipment, training,or other relatedsafetyareas areknowntothesupervisor,yetareallowedtocontinueuncorrected. For example, the failure to consistentlycorrect or discipline inappropriate behavior certainlyfostersan unsafe atmospherebut isnot consideredaviolationifnospecificrulesorregulationswerebroken.Supervisoryviolations,ontheotherhand,arereserved

forthoseinstanceswhenexistingrulesandregulationsarewillfullydisregardedbysupervisorswhenmanagingassets.For instance,permitting aircrew tooperate anaircraft without current qualifications or license is aflagrantviolationthatinvariablysetsthestageforthetragicsequenceofeventsthatpredictablyfollow.

Organizational InfluencesFallible decisions of upper-level management can

directlyaffectsupervisorypractices,aswellasthecondi-

tionsandactionsofoperators.Unfortunately,theseorganizational influencesoftengounnoticedorunreportedbyeventhebest-intentionedaccidentinvestigators.Traditionally,theselatentorganizationalfailuresgen-

erallyrevolvearoundthreeissues:1)resourcemanage-ment, 2) organizational climate, and 3) operationalprocesses.Thefirstcategory,resource management,refersto themanagement, allocation, and maintenance oforganizational resources, including human resourcemanagement (selection, training, staffing), monetarysafetybudgets,andequipmentdesign(ergonomicspeci-fications). In general, corporate decisions about how

suchresourcesshouldbemanagedcenteraroundtwodistinctobjectivesthegoalofsafetyandthegoalofon-time,cost-effectiveoperations.Intimesofprosperity,bothobjectivescanbeeasilybalancedandsatisfiedinfull.However,theremayalsobetimesoffiscalausteritythat demand some give and take between the two.Unfortunately,historytellsusthatsafetyisoftentheloserinsuchbattles,assafetyandtrainingareoftenthefirsttobecutinorganizationsexperiencingfinancialdifficulties.

6.


13/20

Organizational climatereferstoabroadclassoforga-nizationalvariablesthatinfluenceworkerperformanceandisdefinedasthesituationallybasedconsistenciesintheorganizationstreatmentofindividuals(Jones,1988).One telltale sign of an organizations climate is itsstructure,asreflectedinthechain-of-command,delega-

tion of authority and responsibility, communicationchannels,andformalaccountabilityforactions.Justlikein thecockpit, communication andcoordination arevitalwithinanorganization.However,anorganizationspoliciesandculturearealsogoodindicatorsofitscli-mate. Consequently, when policies are ill-defined,adversarial,orconflicting,orwhentheyaresupplantedbyunofficialrulesandvalues,confusionabounds,andsafetysufferswithinanorganization.Finally,operational processreferstoformalprocesses

(operationaltempo,timepressures,productionquotas,incentivesystems,schedules,etc.),procedures(perfor-mance standards, objectives, documentation, instruc-tionsaboutprocedures,etc.),andoversightwithintheorganization (organizational self-study, risk manage-ment,andtheestablishmentanduseofsafetyprograms).Poorupper-levelmanagementanddecisionsconcerningeach of these organizational factors can also have anegative,albeitindirect,effectonoperatorperformanceandsystemsafety.

SummaryThe HFACS framework bridges the gap between

theory and practice byproviding safetyprofessionalswithatheoreticallybasedtoolforidentifyingandclassi-fyingthehumancausesofaviationaccidents.Becausethesystemfocusesonbothlatentandactivefailuresandtheirinterrelationships,itfacilitatestheidentificationoftheunderlyingcausesofhumanerror.Todate,HFACShasbeenshowntobeusefulwithinthecontextofmilitaryaviation, as both a data analysis framework and anaccidentinvestigationtool.However,HFACShasyettobeappliedsystematicallytotheanalysisandinvestigationofcivilaviationaccidents.Thepurposeofthepresentresearchproject,therefore,wastoassesstheutilityofthe

HFACSframeworkasanerroranalysisandclassificationtoolwithincommercialaviation.Thespecificobjectivesofthisstudywerethree-fold.

ThefirstobjectivewastodeterminewhethertheHFACSframework,initscurrentform,wouldbecomprehensiveenoughtoaccommodatealloftheunderlyinghumancausal-factorsassociatedwithcommercialaviationacci-dents,ascontainedintheaccidentdatabasesmaintainedby the FAA and NTSB. In other words, could the

frameworkcapturealltherelevanthumanerrordataorwouldaportionofthedatabasebelostbecauseitwasunclassifiable?Thesecondobjectivewastodeterminewhether theprocess of reclassifying thehumancausalfactorsusingHFACSwasreliable.Thatis,woulddiffer-entusersofthesystemagreeonhowcausalfactorsshould

becodedusingtheframework?Finally,thethirdobjec-tivewastodeterminewhetherreclassifyingthedatausingHFACSyieldabenefitbeyondwhatisalreadyknownaboutcommercialaviationaccidentcausation.Specifi-cally,wouldHFACShighlightanyheretoforeunknownsafetyissuesinneedoffurtherinterventionresearch?

METHOD

DataA comprehensive review of all accidents involving

CodeofFederalAirRegulations(FAR)Parts121and135ScheduledAirCarriersbetweenJanuary1990andDecember1996wasconductedusingdatabaserecordsmaintainedbytheNTSBandtheFAA.Ofparticularinteresttothisstudywerethoseaccidentsattributable,atleastinpart,totheaircrew.Consequently,notincludedwereaccidentsduesolelytocatastrophicfailure,mainte-nanceerror,andunavoidableweatherconditionssuchasturbulence and wind shear. Furthermore, only thoseaccidentsinwhichtheinvestigationwascompleted,andthecauseoftheaccidentdetermined,wereincludedinthisanalysis.Onehundrednineteenaccidentsmetthese

criteria,including44accidentsinvolvingFARPart121operators and 75 accidents involving FAR Part 135operators.

HFACS ClassificationThe119aircrew-relatedaccidentsyielded319causal

factorsforfurtheranalyses.EachoftheseNTSBcausalfactorswassubsequentlycodedindependentlybybothanaviationpsychologistandacommercially-ratedpilotusingtheHFACSframework.OnlythosecausalfactorsidentifiedbytheNTSBwereanalyzed.Thatis,nonewcausalfactorswerecreatedduringtheerror-codingprocess.

RESULTS

HFACS ComprehensivenessAll319(100%)ofthehumancausalfactorsassociated

withaircrew-relatedaccidentswereaccommodatedus-ing theHFACSframework. Instances of all but twoHFACS categories (i.e., organizational climate andpersonal readiness)wereobservedasleastonceinthe

7.


14/20

accidentdatabase.Therefore,nonewHFACScategorieswereneededtocapturetheexistingcausalfactors,andnohumanfactorsdatapertainingtotheaircrewwereleftunclassifiedduringthecodingprocess.

HFACS ReliabilityDisagreementsamongraterswerenotedduringthe

codingprocessandultimatelyresolvedbydiscussion.Usingtherecordofagreementanddisagreementbe-tweentheraters,thereliabilityoftheHFACSsystemwasassessedbycalculatingCohenskappaanindexofagreementthathasbeencorrectedforchance.Theob-tainedkappavaluewas.71,whichgenerallyreflectsagoodlevelofagreementaccordingtocriteriadescribedbyFleiss(1981).

HFACS AnalysesUnsafe ActsTable1presentspercentagesofFARParts121and

135aircrew-relatedaccidentsassociatedwitheachoftheHFACScategories.Anexaminationofthetablerevealsthat at the unsafe acts level, skill-based errors were

associatedwiththelargestpercentageofaccidents.Ap-proximately60%ofallaircrew-relatedaccidentswereassociatedwithatleastoneskill-basederror.Thisper-centagewasrelativelysimilarforFARPart121carriers(63.6%)andFARPart135carriers(58.7%).Figure4,panel A, illustrates that the proportion of accidents

associatedwithskill-basederrorshasremainedrelativelyunchangedovertheseven-yearperiodexaminedinthestudy.Notably,however,thelowestproportionofacci-dentsassociatedwithskill-basederrorswasobservedinthelasttwoyearsofthestudy(1995and1996).Amongtheremainingcategoriesofunsafeacts,acci-

dentsassociatedwithdecisionerrorsconstitutedthenexthighestproportion(i.e.,roughly29%oftheaccidentsexamined,Table1).Again,thispercentagewasroughlyequalacrossbothFARPart121(25.0%)andPart135(30.7%)accidents.Withtheexceptionof1994,inwhichthe percentage of aircrew-related accidents associatedwithdecisionerrorsreachedahighof60%,thepropor-tion of accidents associated with decision errors re-mainedrelativelyconstantacrosstheyearsofthestudy(Figure4,panelB).

Table1.PercentageofAccidentsAssociatedwitheachHFACScategory.HFACSCategory FARPart121 FARPart135 TotalOrganizationalInfluencesResource

Management

4.5

(2)

1.3

(1)

2.5

(3)

OrganizationalClimate 0.0(0) 0.0(0) 0.0(0)OrganizationalProcess 15.9(7) 4.0(3) 8.4(10)

UnsafeSupervisionInadequateSupervision 2.3(1) 6.7(5) 5.0(6)PlannedInappropriateOperations 0.0(0) 1.3(1) 0.8(1)FailedtoCorrectKnownProblem 0.0(0) 2.7(2) 1.7(2)SupervisoryViolations 0.0(0) 2.7(2) 1.7(2)

PreconditionsofUnsafeActsAdverseMentalStates 13.6(6) 13.3(10) 13.4(16)AdversePhysiologicalSates 4.5(2) 0.0(0) 1.7(2)Physical/mentalLimitations 2.3(1) 16.0(12) 10.9(13)Crew-resource

Mismanagement

40.9

(18)

22.7

(17)

29.4

(35)

PersonalReadiness 0.0(0) 0.0(0) 0.0(0)

UnsafeActsSkill-basedErrors 63.6(28) 58.7(44) 60.5(72)DecisionErrors 25.0(11) 30.7(23) 28.6(34)PerceptualErrors 20.5(9) 10.7(8) 14.3(17)Violations 25.0(11) 28.0(21) 26.9(32)

Note:NumbersintablearepercentagesofaccidentsthatinvolvedatleastoneinstanceofanHFACScategory.Numbersinparenthesesindicateaccidentfrequencies.Becausemorethanonecausalfactorisgenerallyassociatedwitheachaccident,thepercentagesinthetablewillnotequal100%.

8


15/20

0

10

20

30

40

50

60

70

80

90 91 92 93 94 95 96

Percentage

Year

A.

0

10

20

30

40

50

60

70

80

90 91 92 93 94 95 96

Percentage

Year

D.

0

10

20

30

40

50

60

70

80

90 91 92 93 94 95 96

Percentage

Year

C.

0

10

20

30

40

50

60

70

80

90 91 92 93 94 95 96

Percentage

Year

B.

Figure4.Percentage of aircrew related accidents associated with skill-based errors(Panel A), decision errors (Panel B), violations (Panel C) and CRM failures (Panel D)across calendar years. Lines represent seven year averages.

Similar to accidents associated with decision errors,those attributable at least in part to violations of rules andregulations were associated with 26.9% of the accidents

examined. Again, no appreciable difference was evidentwhen comparing the relative percentages across FARParts 121 (25.0%) and 135 (28.0%). However, anexamination of Figure 4, panel C, reveals that the relativeproportion of accidents associated with violations in-creased appreciably from a low of 6% in 1990 to a highof 46% in 1996.

Finally, the proportion of accidents associated withperceptual errors was relatively low. In fact, only 17 of the119 accidents (14.3%) involved some form of perceptualerror. While it appeared that the relative proportion ofPart 121 accidents associated with perceptual errors was

higher than Part 135 accidents, the low number ofoccurrences precluded any meaningful comparisons acrosseither the type of operation or calendar year.

Preconditions for Unsafe ActsWithin the preconditions level, CRM failures were

associated with the largest percentage of accidents. Ap-proximately 29% of all aircrew-related accidents wereassociated with at least one CRM failure. A relatively

larger percentage of FAR Part 121 aircrew-accidentsinvolved CRM failures (40.9%) than did FAR Part 135aircrew-related accidents (22.7%). However, the per-

centage of accidents associated with CRM failures re-mained relatively constant over the seven-year period forboth FAR Part 121 and 135 carriers (Figure 4, panel d).

The next largest percentage of accidents was associ-ated with adverse mental states (13.4%), followed byphysical/mental limitations (10.9%) and adverse physi-ological states (1.7%). There were no accidents associ-ated with personal readiness issues. The percentage ofaccidents associated with physical/mental limitation washigher for FAR Part 135 carriers (16%) compared withFAR Part 121 carriers (2.3%), but accidents associatedwith adverse mental or adverse physiological states were

relatively equal across carriers. Again, however, the lownumber of occurrences in each of these accident cat-egories precluded any meaningful comparisons acrosscalendar year.

Supervisory and Organizational FactorsVery few of the NTSB reports that implicated the

aircrew as contributing to an accident also cited someform of supervisory or organizational failure (see Table

9.


16/20

1).Indeed, only 16% ofallaircrew-relatedaccidentsinvolvedsomeformofeithersupervisoryororganiza-tionalinvolvement.Overall,however,alargerpropor-tionofaircrew-relatedaccidentsinvolvingFARPart135carriersinvolvedsupervisoryfailures(9.3%)thandidthoseaccidentsinvolvingFARPart121carriers(2.3%).

Incontrast,alargerproportionofaircrew-relatedacci-dentsinvolvingFARPart121carriersinvolvedorganiza-tionalfactors(20.5%)thandidthoseaccidentsinvolvingFARPart135carriers(4.0%).

DISCUSSION

HFACS ComprehensivenessTheHFACSframeworkwasfoundtoaccommodate

all319causalfactorsassociatedwiththe119accidentsinvolvingFARParts 121 and 135 scheduled carriersacross the seven-year period examined. This findingsuggeststhattheerrorcategorieswithinHFACS,origi-nallydevelopedforuseinthemilitary,areapplicablewithincommercial aviation aswell. Still, someof theerror-factorswithintheHFACSframeworkwereneverobservedinthiscommercialaviationaccidentdatabase.Forexample,noinstancesofsuchfactorsasorganiza-tionalclimateorpersonalreadinesswereobserved.Infact,veryfewinstancesofsupervisoryfactorswereevi-dentatallinthedata.Oneexplanationforthescarcityofsuchfactorscould

bethat,contrarytoReasonsmodeloflatentandactive

failuresuponwhichHFACSisbased,suchsupervisoryandorganizationalfactorssimplydonotplayaslargeofaroleintheetiologyofcommercialaviationaccidentsasonceexpected.Consequently,theHFACSframeworkmayneedtobepareddownorsimplifiedforusewithcommercialaviation.Anotherexplanation,however,isthatthesefactorsdocontributetomostaccidents,yettheyarerarelyidentifiedusingexistingaccidentinvesti-gationprocesses.Nevertheless,theresultsofthisstudyindicatethattheHFACSframeworkwasabletocaptureallexistingcausalfactorsandnonewerror-categoriesoraircrewcause-factorswereneededtoanalyzethecom-

mercialaccidentdata.

HFACS ReliabilityTheHFACSsystemwasfoundtoproduceanaccept-

able level of agreement among the investigatorswhoparticipatedinthisstudy.Furthermore,evenafterthislevelofagreementbetweeninvestigatorswascorrectedforchance,theobtainedreliabilityindexwasconsideredgoodbyconventionalstandards.Still,thisreliability

indexwassomewhatlowerthanthoseobservedinstudiesusingmilitary aviation accidentswhich, in some in-stances, have resulted in nearly complete agreementamonginvestigators(Shappell&Wiegmann,1997b).Onepossibleexplanationforthisdiscrepancyisthe

differenceinboththetypeandamountofinformation

availabletoinvestigatorsacrossthesestudies.Unlikethepresentstudy,previousanalystsusingHFACStoanalyzemilitaryaccidentdataoftenhadaccesstoprivilegedandhighlydetailedinformationabouttheaccidents,whichpresumablyallowedforabetterunderstandingoftheunderlyingcausalfactorsand,hence,producedhigherlevels of reliabilities. Another possibility is that thedefinitions and examples currently used to describeHFACSaretoocloselytiedtomilitaryaviationandarethereforesomewhatambiguoustothosewithinacom-mercial setting. Indeed,the reliabilityoftheHFACSframeworkhasbeenshowntoimprovewithinthecom-mercialaviationdomainwheneffortsaretakentopro-videexamplesandcheckliststhataremorecompatiblewith civil aviation accidents (Wiegmann, Shappell,Cristina&Pape,2000).

HFACS AnalysisGiven the large number of accident causal factors

contained in the NTSB database, each accident ap-peared,atleastonthesurface,toberelativelyunique.Assuch, commonalties or trends in specific error formsacrossaccidentswerenotreadilyevidentinthedata.Still,

therecodingofthedatausingHFACSdidallowforsimilarerror-formsandcausalfactorsacrossaccidentstobeidentifiedandthemajorhumancausesofaccidentstobediscovered.Specifically, theHFACSanalysis revealed that the

highest percentage of all aircrew-related accidents asassociatedwithskill-basederrors.Furthermore,thispro-portionwaslowestduringthelasttwoyearsofthisstudy,suggestingthataccidentsassociatedwithskill-baseder-rorsmaybeonthedecline.Tosome,thefindingthatskill-basederrorswerefrequentlyobservedamongthecommercialaviationaccidentsexaminedisnotsurpris-

inggiventhedynamicnatureandcomplexityofpilotingcommercial aircraft, particularlyin theincreasinglycongestedU.S.airspace.Thequestionremains,how-ever,astothedrivingforcebehindthepossiblereduc-tion in such errors. Explanations could includeimprovedaircrewtrainingpracticesorperhapsbetterselectionprocedures.Anotherpossibilitymightbetherecenttransitionwithintheregionalcommuterindus-tryfromturboproptojetaircraft.Suchaircraftare

10.


17/20


18/20

flyingaircraftoperatingunderFARPart135areyoungerandmuch less experienced. Furthermore, suchpilotsoftenflylesssophisticatedandreliableaircraftintoareasthatarelesslikelytobecontrolledbyATC.Asaresult,theymayfrequentlyfindthemselvesinsituationsthatexceedtheirtrainingorabilities.Suchaconclusionis

supportedbythefindingspresentedhere,sincealargerpercentageofFARPart135aircrew-relatedaccidentswereassociatedwiththephysical/mentallimitationsofthepilot.However,asmallerpercentageFARPart135aircrewaccidentswereassociatedwithCRMfailures,possiblybecausesomeFARPart135aircraftaresingle-piloted, which simply reduces the opportunity forCRMfailures.ThesedifferencesbetweenFARParts121and135

schedulecarriersmaybelessevidentinfutureaviationaccidentdatasincethefederalregulationswerechangedin1997.SuchchangesrequireFARPart135carriersoperatingaircraftthatcarrytenormorepassengerstonowoperateundermorestringentFARPart121rules.Thus,thehistoricaldistinctioninthedatabasebetweenFARPart135and121operatorshasbecomesomewhatblurredintheyearsextendingbeyondthecurrentanaly-sis.Therefore,futurehuman-erroranalysesandcom-parisons across these different types of commercialoperationswillthereforeneedtoconsiderthesechanges.

SUMMARY AND CONCLUSIONS

This investigation demonstrates that the HFACSframework,originallydevelopedforandproveninthemilitary,canbeusedtoreliablyidentifytheunderlyinghuman factors problems associated with commercialaviationaccidents.Furthermore,theresultsofthisstudyhighlight critical areas of human factors in need offurthersafetyresearchandprovidethefoundationuponwhich to build a larger civil aviation safety program.Ultimately,dataanalysessuchasthatpresentedherewillprovidevaluableinsightaimedatthereductionofavia-tionaccidentsthroughdata-driveninvestmentstrategiesandobjectiveevaluationofinterventionprograms.The

HFACSframeworkmayalsoproveusefulasatoolforguidingfutureaccidentinvestigationsinthefieldanddeveloping better accident databases, both of whichwould improve theoverall qualityandaccessibility ofhumanfactorsaccidentdata.

Still,theHFACSframeworkisnottheonlypossiblesystemuponwhichsuchprogramsmightbedeveloped.Indeed,thereoftenappearstobeasmanyhumanerrorframeworksas there are those interestedin thetopic(Senders&Moray,1991).Indeed,astheneedforbetterappliedhumanerroranalysismethodshasbecomemore

apparent,anincreasingnumberofresearchershavepro-posedothercomprehensiveframeworkssimilartoHFACS(e.g.,OHare,inpress).Nevertheless,HFACSis,todate,theonlysystemthathasbeendevelopedtomeetaspecificsetofdesigncriteria,includingcomprehensiveness,reli-ability,diagnosticity,andusability,allofwhichhavecontributedtotheframeworksvalidityasanaccidentanalysistool(Shappell&Wiegmann,inpress).Further-more,HFACShasbeenshowntohaveutilityasanerror-analysistoolinotheraviation-relateddomainssuchasATC(HFACS-ATC;Pounds,Scarborough,&Shappell,2000)andaviationmaintenance(HFACS-ME;Schmidt,Schmorrow,&Hardee,1998),andiscurrentlybeingevaluatedwithinothercomplexsystemssuchasmedi-cine(currentlyreferredtoasHFACS-MD).Finally,itisimportanttorememberthatneitherHFACSnoranyothererror-analysistoolcanfixtheproblemsoncetheyhavebeenidentified.Suchfixescanonlybederivedbythoseorganizations,practitionersandhu-manfactorsprofessionalswhoarededicatedtoim-provingaviationsafety.

REFERENCES

Bird,F.(1974).Management guide to loss control.At-lanta,GA:InstitutePress.

Fleiss,J.(1981).Statistical Methods for Rates and Proportions. NewYork:JohnWiley.

Ford,C.,Jack,T.,Crisp,V.&Sandusky,R.(1999).Aviation accident causal analysis.Advances inAviation Safety Conference Proceedings,(P-343).Warrendale, PA: Society of Automotive Engi-neersInc.

Geller,E.(March,2000).Behavioralsafetyanalysis:A

necessary precursor to corrective action. Professional Safety,29-32.

International Civil Aviation Organization (1993).Investigation of human factors in accidents andincidents(HumanFactorsDigest#7),Montreal:Canada.

12.


19/20

Jones,A.(1988).Climateandmeasurementofconsen-sus:Adiscussionoforganizationalclimate.InS.Cole,R.Demaree&W.Curtis,(Eds.),Applications of Interactionist Psychology: Essays in Honor ofSaul B. Sells (pp. 283-290). Hillsdale, NJ:Earlbaum.

NationalTransportationSafetyBoard(2000).Aviationaccident statistics. [On-l in e] . Ava il ab le :www.ntsb.gov/aviation/Stats.htm

OHare,D.(inpress).TheWheelofMisfortune.Ergonomics.

OHare,D.,Wiggins,M.,Batt,R.,andMorrison,D.(1994).Cognitivefailureanalysisforaircraftacci-dentinvestigation.Ergonomics,37,1855-69.

Pounds,J.,Scarborough,A.,&Shappell,S.(2000).AhumanfactorsanalysisofAirTrafficControlop-

erational errors (Abstract).Aviation, Space andEnvironmental Medicine,71,pp.329

Rasmussen,J.(1982).Humanerrors:Ataxonomyfordescribinghumanmalfunction in industrial in-stallations.Journal of Occupational Accidents, 4,pp.311-33.

Reason,J.(1990).Human error.NewYork:CambridgeUniversityPress.

Schmidt,J.,Schmorrow,D.,&Hardee,M.(1998).ApreliminaryhumanfactorsanalysisofNavalAvia-tionmaintenancerelatedmishaps.Proceedings of

the 1998 Airframe/Engine Maintenance and RepairConference (P329),LongBeach,CA.

Senders,J.,&Moray,N.(1991).Human error: Cause,prediction and reduction.Hillsdale,NJ:Earlbaum.

Shappell, S.,&Wiegmann, D. (1996).U. S. NavalAviationmishaps 1977-92:Differences betweensingle-anddual-pilotedaircraft. Aviation, Space,and Environmental Medicine,67,65-9.

Shappell,S.&WiegmannD.(1997a).Ahumanerrorapproachtoaccidentinvestigation:Thetaxonomyofunsafeoperations.The International Journal of

Aviation Psychology,7,pp.269-91.

Shappell, S.&Wiegmann,D. (1997b).A reliabilityanalysisoftheTaxonomyofUnsafeOperations(Abstract). Aviation, Space, and EnvironmentalMedicine,69,pp.620.

Shappell,S.&Wiegmann,D.(2000a). TheHumanFactors Analysis and Classification System(HFACS).(ReportNumberDOT/FAA/AM-00/7).WashingtonDC:FederalAviationAdministration.

Shappell,S.&Wiegmann,D.(2000b).IsproficiencyerodingamongU.S.Navalaircrews?Aquantita-tive analysis using theHumanFactors AnalysisandClassificationSystem(HFACS).Proceedingsof the 44th meeting of the Human Factors and Ergonomics Society.

Shappell,S.&Wiegmann,D.(2001).ApplyingReason:TheHumanFactorsAnalysisandClassificationSystem(HFACS).Human Factors and Aerospace

Safety,1,59-86.

Shappell,S.,Wiegmann,D.,Fraser,J.,Gregory,G.,Kinsey,P.,&Squier,H(1999).Beyondmishaprates: A human factors analysis of U.S. Navy/MarineCorpsTACAIRandrotarywingmishapsusing HFACS (Abstract).Aviation, Space, andEnvironmental Medicine,70,pp.416-7.

Wiegmann,D.&Shappell,S.(1997).Humanfactorsanalysisofpost-accidentdata:Applyingtheoreti-caltaxonomiesofhumanerror . The InternationalJournal of Aviation Psychology,7,pp.67-81.

Wiegmann,D.&Shappell,S.(1999).HumanerrorandcrewresourcemanagementfailuresinNavalavia-tionmishaps:AreviewofU.S.NavalSafetyCenterdata,1990-96.Aviation, Space, and EnvironmentalMedicine,70,pp.1147-51.

Wiegmann,D., Shappell, S.,Cristina,F.&Pape,A.(2000).Ahumanfactorsanalysisofaviationacci-dentdata:AnempiricalevaluationoftheHFACSframework(Abstract).Aviation, Space and Environmental Medicine,71,pp.328.

Yacavone, D. W. (1993). Mishap trends and cause

factorsinNavalaviation:AreviewofNavalSafetyCenterdata,1986-90.Aviation, Space and Environmental Medicine,64,392-5.

13.


20/20

Documents

A Human Error Analysis of Commercial Aviation Accidents Using the Human Factors Analysis and Classification System (HFACS)PDF