93
1 INTRODUCTION
Ships operate inahighly risky milieu; typically, the
people on board adapt a set routine of shift work
disruptedbyarrivalat,workingin,andsailingfrom
port. Recently, international maritime authorities
have performed significant contributions to improve
safetyatsea intheshippingtransportationindustry.
But,thereisnoremarkabledecreaseintota
lnumber
of the shipping accidents. From the economical
perspective,shipsareveryimportantcommoditiesas
they offer jobs for people and enhance the financial
activitiesbytransportinggoodsandpassengersfrom
one node to another. This is a presence which
involves living in the pla
ce of work for prolonged
periods,creatingaunparalleledformofworkinglife
which almost no doubt increases the risk of human
error.Marinetraffic, whichisa natural consequence
of growing world trade, not only increased the
number of ship accidents but also caused more
frequentin‐shipoccupationalaccidents (Özdemirand
Güneroğlu2015,Uğurlu2016).Thereisnodoubttha
t
maintainingtheseafarers’safetyisutterlyimportant.
According to the International Labour Organization
(ILO), seafarer is any person who works in any
positiononboardaseagoingshiporvesselengaged
incommercialmaritimenavigation,whetherpublicly
orprivatelyowned,othertha
nashipofwar.Beinga
seafarer includes professionally difficulties in
addition to the harsh environmental factors. Unlike
otherprofessions,theyspend24hoursadayatwork.
Therefore,whendangerousoccupationsarelisted,the
A MCDM Approach with Fuzzy AHP Method for
Occupational Accidents on Board
Ü.Özdemir
M
ersinUniversity,Mersin,Turkey
İ.Altinpinar&F.B.Demirel
KaradenizTechnicalUniversity,Trabzon,Turkey
ABSTRACT: Occupational accidents on board criteria determining is a challenging procedure in shipping
industry as the ideal safety ship management strategy depends on many factors involving in shipping
transportationTherearemanylegislations,agreementsandpracticestoobtainseriesofsecuritymeasuresin
ordertoensuresafetyandsecurityofseafarers.Ca
usesofon‐boardoccupationalaccidentsneedtobeevaluated
in a correct manner to regulate more functional practices and also to lower the on‐board accident rates.
However,causesofon‐boardaccidentscanbeextremely complex.Therefore,scientificmethodsshouldbeused
toevaluatethecausesandtodeterminethemea
surestobetaken.Theevaluationoftheparametersisofgreat
importanceforthefutureofthemaritimesectorandintermsofdevelopment.Inthisstudy,factorshavebeen
identifiedthatleadtoseafarers’occupationalaccidents onboardandwetriedtopresentalt
ernativesolutions
whichcanbeappliedonthisissue.Severityofthereasonsthatledtotheaccidentsandtheirrelationshipswith
eachotherareidentifiedtobeabletosortthroughthealternativesolutionswithamodelusingthefuzzyAHP
(AnalyticHierarchyProcess)methodapproach.Resultsofthestudyrevealedtha
tthemostimportantcriteria
fortheoccupationalaccidentsonboardcriteriaselectionarerespectively;humanfactors,lackofmanagement,
ship‐bornetroubles,cargotroublesandenvironmentalfactors
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 12
Number 1
March 2018
DOI:10.12716/1001.12.01.10
94
riskof accidentisratherhighforseafarers(Roberts,
2002). There are many different reasons for the
occupational accidents on board. Hanson’s study
carriedoutasurveythatshowedthatmortalinjuries
amongDanishseafarerswere11.5 timeshigherthan
averageratesamongtheDanishmaleworkersashore
between1986
and1993(Hansen,1996).Accidentson
ships may pose great risks for personnel and the
environment,accordingtothenatureoftheaccident
(Portela 2005, Uğurlu et al. 2016). Accident can be
described as an unexpected event that results in a
personal injury or loss of property, or both.
We see
that ship related occupational accidents can be
inspected under three main headings. These can be
listed as auxiliary causes, sudden causes and other
causes (Hansen 1996, Hansen et al. 2002). Certain
operationsonboardmaycauseextremelydangerous
working conditions. Inadequate safety of ship
operations are not only inclined
by the material
precautionsbutalsohumanfactorssuchasseafarers’
behaviors,habits,lack ofattentionsandoccupational
educations (Roberts 2000, Martins and Maturana
2010,Uğurluetal.2016).Mostofmarineaccidentsare
caused by some form of human error as well as
incidents(Havold2000,Rothblum2000,
Toffoliet.al.
2005, Hetherington et al. 2006, Grech et. al. 2008,
Talley2009,ÖzdemirandGüneroğlu2015).Previous
researchesdemonstratethatforeachseriousaccident
inthemaritime industry,orinany other area,there
are a larger number of incidents, a big number of
near‐missesandhuge
numberofsafety‐criticalevents
and unsafe acts (Grech et. al. 2008). Example of
workers’ unsafe acts include the determination to
proceedwithworkinunsafeconditions,disregarding
standardsafetyproceduressuchasnotwearingsafety
equipment, working while intoxicated, occupational
illiteracy,workingwithinsufficientsleepandfatigue
(Abdelhamidand
Everett2000).Personalinjuriesare
muchworsewhentheyareonboard,duetothefact
that seafarer’s health care opportunities are poorer
thanthoseashore.Acriticalconflictinthetreatment
of seafarer at sea is that medical care on‐board is
appliedbyamedicalhealthofficerwhois
notmedical
professional(Oldenburgetal.2010).
Maritimeaccidentsusuallyoccurduetofailureof
adecisionascombinationofcoincidentalincidentsor
processes, as a general rule by negligence of one or
more independent components that are required to
action accurately for the successful finalizations of
decision flow (Özdemir
and Güneroğlu 2015). Main
reasons of ship related accidents can be listed as;
human factors (psychological, physical, human
relations, team work, communication);
machinery/equipment factors (incorrect machinery
andequipmentlayout,absentordefectiveprotectors,
inadequate standardization, inadequate control and
maintenance, inadequate engineering services);
environmental factors (inadequate knowledge,
improper working methods, improper
working
environment) and management factors (inadequate
management organization, incomplete rules and
regulations, inadequate security management plan,
educationalinadequacies,inadequatehealthcontrols,
employment of incompetent personnel, etc.) (O’Neil
2003, Portela 2005, Hetherington and Flin 2006,
ÖzdemirandGüneroğlu2015,Uğurlu2016)
Inthisstudy,weuseda quantitativeapplication
to
examine the reasons of work related accidents in
ships and tried to discover alternative solutions for
the matter. Work related accidents on ships have
several interconnected reasons. For the solutions of
such problems in which various factors and criteria
must be analysed and evaluated, using a “fuzzy
multi‐criteriadecision
‐makingmodel”(FMCDM)can
beapositiveapproach.Inthisstudy,wedetermined
the criteria that cause work related accidents and
offeredanappropriate methodology for thesolution
of the problem by using a fuzzy multi‐criteria
decision‐making model. Fuzzy AHP (Analytic
HierarchyProcess)isusedtodetermineand
rankthe
accident‐related criteria according to their
importance.
2 METHODOLOGY
The problem of the causes of occupational accident
selection criteria on ship is considered as a multi‐
dimensionalcomplexissuethatcanberesolvedbya
MCDM approach. In such a complex problem, the
availabilityofmanychoicesandtheir
relativeimpact
onthefinalsolutionarealwaysriskyastheymaybe
misleadingthedecisionmakerifthereisnotareliable
tool in hand (Özdemir and Güneroğlu 2015, Uğurlu
2015,Güneroğluet.al 2016).In thisstudy,causes of
occupational accident selection criteria’ weighting
was
implemented using Fuzzy AHP following
Buckley (1985) by pairwise comparisons of the
experts’ scores were applied for ranking and
evaluatingthecriteria.
Thestepsoftheappliedtechniqueandrelatedcase
studyarepresentedinfollowingsub‐sections.
2.1 FuzzyAHP
The first step of implemented methodology is
enduredonBuckley’s
AHP technique. Actaully, this
techniqueisamagnifiedversionoftheoriginalAHP
technique by Saaty (2006) and instead of classic
rationalnumbers,itusesfuzzycomparisonratios.The
notable virtue of the Buckley’s Fuzzy AHP is the
expansion of the statements to Fuzzy environment
withrelativelyeasyeffortsas
wellastheguaranteeof
one absolute value. Complex, laborious and error‐
prone computational requirements are the main
disadvantagesofthesametechnique.TheFuzzyAHP
technique by Buckley (1985) can be summarized as
followed(Buckley,1985; Kafalı etal.,2014;Özdemir
andGüneroğlu,2017);
The first step of the
technique contains defining
themaincriteriathatpotentiallyaffectstheproblem
under investigationbyexperts and decision makers.
Afterwards,inordertoconverttheexpertevaluations
tothefuzzynumbers,apredefinedlinguisticscaleis
used. Then, expert evaluations were received as
pairwise comparisons that converted to the fuzzy
numberswith
inmatrixformasshowninEq.1,
95
k
à =
12
1
21
12
1
1
..
.
..
.
..
.
1
m
mm
Ã
Ã
Ã
m
ÃÃ
(1)
whereʺ
k
à ʺ is the response matrix by each expert.
Linguistic scale and corresponding fuzzy numbers
usedintheBuckleyʹstechniqueweregiveninTable1
(XuandYager2008,Kafalıetal.2014).
Then all data received as result of expertsʹ
evaluations is compiled by using weighted mean
formulaasgiven
inEq.2,
12
12
12
k
x
yxy kxy
xy
k
Z
AZA ZA
Ã
ZZ Z
(2)
InEq.2“
mn
à ”,isthejoinedcomparisonvalueof
thecritieriaʺxʺandʺyʺ;“
k
Z ”istheweightedvalueof
ʺk.ʺexpert; “
k
xy
A ” is the comparison value ofʺk.ʺ
expert evaluations corresponding toʺxʺ andʺyʺ
criteria. The decision matrix formed by weighted
meansofallexpertsʹscorescanbeshowninfollowing
matrixform(Eq.3)
ˇ
A=
12 1
21
2
12
1
1
1
m
m
mm
ÃÃ
Ã
Ã
ÃÃ
(3)
Afterobtainingthedecisionmatrix,theweightof
eachcriterioncanbecalculatedintwosteps.Thefirst
stepiscomputingthegeometricmeanofeachrowin
decisionmatrix,asshowninEq.4
ˇ
i
b =
12
ii in
aa a
1/n
(4)
where“n”standsfortotalnumberofcriteria,“
in
a ”is
thefuzzycomparisonvaluebetweentwocriteria“i.”
and“n.”and“
ˇ
i
b
”isthefuzzygeotmetricmeanof
the all compared criteria. As a second step, a fuzzy
weight of each criterion is calculated by applying
Eq.5.
i
w =
i
b ⊗
12 n
bb r
‐1
(5)
where,“
i
w ”isthefuzzyweightofcriterion“i.”.
The remaining part of the Buckley’s FuzzyAHP
technique requires conversion of the fuzzy numbers
to corresponding absolute values and calculation of
relativeweightsamongallcriteria.
where,“
ˇ
B ” is referred to triangular fuzzy number,
defuzzificationof“
ˇ
B ”canbeappliedusingEq.6.
ˇ
B
=
123
++
3
bbb
(6)
For better evaluation of the obtained values,
normalizationisappliedonthemaincriteriaasitisin
Eq.7,
R
i
w
N
=
1
N
i
n
N
i
i
w
w
(7)
where“
R
i
w
N”isreferredto normalized weight of
each criteria and “n” is the total number of the
criteria.Similarly,normalizationofsub‐criteriaisalso
performedforeachelementsofthematrix.
Finally, relative fuzzy and absolute weights of
main and sub‐criteria are computed by multiplying
main and related each sub
‐criteria in the matrix as
showninEq.8andEq.9.
R
i
w
SN
=
w
N
⊗ )
i
w
SN
(8)
where,“
R
i
w
SN”isthefuzzyrelativeweightof“i.”
sub‐criterion, “
w
N” is the fuzzy weight of the
related main criterion and “(
i
w)
SN” is the fuzzy
weightofthesamesub‐criterion.
R
i
w
SN
=
R
w
N
X
R
i
w
SN
(9)
In Eq.9, “
R
i
w
SN” is the normalized relative
absoluteweightofthe“i.”sub‐criterion,“
R
w N”is
the normalized absolute weight of the related main
criterionand“
R
i
w SN” is the normalized absolute
weightofthesamesub‐criterion.
3 CASESTUDY
With an aim of obtaining data sets by establishing
criteriawithintheframeworkofresearchmodel;ship
crews,companyofficials,academicians,analystsand
casualties/casualties’relativeswereinterviewed.Also,
safety reports were evaluated and criteria were
established by
considering the agreements and
conventionspublishedformaritimesafety.Asaresult
oftheevaluations made,itwas decided that 5 main
criteriashowninTable1,wouldbestudied.
Maritime management is comprised of 9 people
(Oceangoing master/3, ship casualties/3,
academician/3) and these people have experiences
about ship’s crew.
Questionnaire forms of compiled
criteria shown in Table 1 were applied to the
participants with an aim of obtaining opinions of
decisionmakers.Pairedcomparisonmatrixesofeach
expertrelatedtoallthecriteriawereobtainedinthe
formofverbalstatementsasaresultoftheevaluation
of all the
criteria. Due to the fact that all the
questionnaire data, which were collected from the
experts, were in verbal forms, they need to be
96
convertedtotriangularfuzzynumbersinaccordance
with fuzzy number equivalents of linguistic scale,
which was specified before (Xu and Yager, 2008;
Özdemir and Güneroğlu, 2017; Özdemir and Çetin,
2017). The values in Table 2 were used in these
conversions.
Table1.Criteriadeterminedforthestudy.
_______________________________________________
# Criteria
_______________________________________________
1 Environmentalfactors(Seacondition,weathercondition
etc.)‐C1
2 Lackofmanagement(Lackofshiprules,failuretotake
measures,lackofmanagement,lackofcommunication
etc.)–C2
3 ShipborneTroubles(Shipage,shipcondition,condition
ofequipment,equipmentinadequacy,poorlightingetc.)
‐C3
4 Human
Factor(Lackoftraining,unawareness,
carelessness,occupationalwillies,fatigue,dangerous
movementsetc.)–C4
5 CargoTroubles(inappropriateloading,dangerous
cargoesetc.)‐C5
_______________________________________________
Incorporatedfuzzydecisionmatrixes of the data,
whichwereobtainedasaresultofpa iredcomparison
of main criteria by using formula 10 and 11, were
calculatedasinTable3.
12
1/
N
ij ij ij ij
CNcc c
(10)
12
1/
N
ij ij ij ij
DNdd d
(11)
Table2. Linguistic terms used for Buckley’s Fuzzy AHP
(Kafalıetal.2014,ÖzdemirandGüneroğlu2016,Özdemir
andGüneroğlu2017,ÖzdemirandÇetin2017)
_______________________________________________
LinguisticTermsFuzzyNumber
_______________________________________________
Slightlymoreimportant(Row)(1,3,5)
Stronglymoreimportant(Row) (3,5,7)
Highlymoreimportant(Row)(5,7,9)
Absolutelymoreimportant(Row) (7,7,9)
Equallyimportant(1,1,3)
Absolutelymoreimportant(Column) (0.111,0.111,0.143)
Highlymoreimportant(Column) (0.111,0.143,0.200)
Stronglymoreimportant(Column)
(0.143,0.200,0.333)
Slightlymoreimportant(Column)(0.200,0.333,1.000)
_______________________________________________
The step after calculating incorporated fuzzy
decision matrixes calculations was the calculation of
criterionweightsaccordingtoBuckleyapproachand
it is carried out in the second step. As a first step,
geometricaverageofeachlineofincorporatedfuzzy
decision matrixes is calculated. This process is
expressedinformula4.
Inthesecondstep,geometric
average of matrix is calculated and then its fuzzy
weightvalueiscalculatedwiththehelpofformula5.
Thisprocesswasappliedforallmainandsub‐criteria.
Weighted fuzzy decision matrix, which was
calculatedformaincriteria,wasshowninTable4.
According
to Buckley approach, the next step is
conversion of fuzzy values into absolute values.
According to this, defuzzification and normalization
processesarecarriedout.Formula6wasusedforthis
calculation. Formula 7 was benefited in order to
evaluateabsoluteweightsinabetterway.Theresults,
in which calculated defuzzification
process was
includedforthecriteria,wereshowninTable4and
normalizationresultswereshowninTable5.
Table3.Aggregatedfuzzydecisionmatrixforthecriteria
_______________________________________________
C1 C2 C3 C4 C5
_______________________________________________
C11.000 0.125 0.324 0.621 1.458
1.000 2.354 1.000 0.388 2.642
1.000 3.652 2.024 0.547 2.033
C20.256 1.000 0.387 0.254 2.010
0.745 1.000 2.457 0.541 2.354
2.354 1.000 1.874 2.456 2.247
C30.385 0.845 1.000 0.451 1.687
1.845 1.354 1.000 0.365 1.347
2.214 2.410 1.000 1.354 4.024
C41.651 2.874 0.343 1.000 2.410
1.033 3.120 1.687 1.000 3.025
0.333 1.018 2.374 1.000 4.024
C50.897 0.985 0.852 1.241 1.000
0.624 0.458 0.349 0.314 1.000
1.025 1.303 1.541 0.652 1.000
_______________________________________________
Table4. Weighted fuzzy decision matrix for criteria (a),
defuzzifiedcriteriaweights(b).
_______________________________________________
(a) WeightedFuzzyDecisionMatrix
_______________________________________________
C1 0.034 0.398 0.403
C2 0.204 0.452 0.182
C3 0.024 0.065 0.035
C4 0.287 0.078 0.802
C5 0.136 0.304 0.246
_______________________________________________
(b) Defuzzifiedcriteriaweights
_______________________________________________
C1 0.026 0.228 0.542
C2 0.125 0.309 0.021
C3 0.203 0.065 0.520
C4 0.028 0.122 0.033
C5 0.077 0.201 0.217
_______________________________________________
Table5.Normalizedcriteriaweights
_______________________________________________
Criteria Normalization(Crisp) PercentageValue(%)
_______________________________________________
C1(5)0.124812,48
C2(2)0.243824,38
C3(3)0.187818,78
C4(1)0.259825,98
C5(4)0.183818,38
_______________________________________________
Total1.000100
_______________________________________________
4 RESULTS
Accordingtotheresults ofthestudy,thereasonsthat
occupational accident on board are specified as
followsinorderofpriorities:humanfactors(C4),lack
ofmanagement(C2),shipborneTroubles(C3),Cargo
Troubles (C5), and Environmental factors (C1).
According to the results of the study, criteria of
humanfactorssuchaslackoftraining,unawareness,
carelessness,occupational willies,fatigue,dangerous
movements etc., which is ranked in the first place,
thatmainreasons of occupationalaccidenton board
were experienced. When the criteria are examined
generally,weseethatC2andC4aredirectlyhuman‐
induced while C3
and C5 are indirectly effected by
humanerror.C1standsoutasthedistinctivecriteria
thatdifferentiatesthemaritimeoccupationfromother
occupational fields. Safety culture affects behavioral
pattern of people against dangerous situations (Cox
97
andFlin1998,Cooper2000,Dursun2013).Reimanve
Oedewald(2002)collectedthecriteriaof‘goodsafety
culture’ in the literature under the main headings
such as: security policies; apparent sagacity of
management for security; democratic applications
and competencies; positive values with security
tendencies; open definitions of responsibilities and
necessities;
security priority operations; balance of
security and production; competent workers and
education; high motivation and work satisfaction;
mutualtrustandfairapproachbetweenmanagement
andworkers;updateofquality,rulesandregulations;
regular machinery maintenance; proper and regular
reportingofeveryincidentandinterpretation;healthy
informationflowfromdifferentmanagerial
levelsand
positions;adequatefundsandconstantdevelopment;
properdesignand business relationswithauthority.
Perception of safety culture of individuals is an
importantfactorinthe accidents that originate from
such human errors. Enhancing the perception of
safetyculturewillcontributetopreventaccidentsthat
results from human induced errors.
Several studies
andaccidentsreportshavewarnedofthedifficulties
encounteredbycrewswhoareconstantlyworkingon
shipsofdifferentsizes,withdifferentequipment,and
carrying different cargoes. Mariners often do not
understandhowtheautomationworksorunderwhat
set of operating conditions it was designed to work
effectively.
Dealing with this research enables the following
contributions to the shipping accident analysis and
prevention literaturethatimprovingthestructure of
the existing MCDM model and extending the
application of combine MCDM to occupational
accidentonboard.Whenthe previous studies about
privateering are considered, it is thought that this
study will significantly contribute to the literature
and future studies due to the lacking number of
quantitative studies about work accidents on ship,
whichareacceptedasamajorproblemformaritime
industry.
REFERENCES
Abdelhamid,T. S.,& Everett,J. G.(2000). Identifyingroot
causesofconstructionaccidents.Journalofconstruction
engineeringandmanagement,126(1),52‐60.
Accident preventionon boardship at seaand inport. An
ILO code of practice. Geneva, International Labour
Office,2ndedition,1996ISBN92‐2‐109450‐2.
Buckley,
J.J. (1985), “Fuzzy Hierarchical Analysis”, Fuzzy
SetsandSystems,Vol.17,pp.233‐247.
Cooper Ph. D, M. D. (2000). Towards a model of safety
culture.Safetyscience,36(2),111‐136.
Cox,S.,& Flin,R.(1998).Safetyculture:philosopherʹsstone
ormanofstraw?.Work&stress,12(3),
189‐201.
Dursun, S. (2013).İşgüvenliği kültürünün çalı şanların
güvenli davranışları üzerine etkisi. Sosyal Güvenlik
Dergisi(SGD),3(2).
Grech,M.,Horberry,T.,&Koester,T.(2008).Humanfactors
inthemaritimedomain.CRCPress.
Hansen, H. L. 1996. “Surveillance of Deaths on Board
Danish Merchant Ships
1986‐93.” Occupational &
EnvironmentalMedicine53:269–275.
Hansen, H. L., D. Nielsen, and M. Frydenberg. 2002.
“Occupational Accidents aboard Merchant Ships.”
Occupational&EnvironmentalMedicine59:85–91.
Havold, J. I. (2000). Culture in maritime safety. Maritime
Policy&Management,27(1),79‐88.
Hetherington,C.,R.Flin,andK.Mearns.2006.““Safetyİn
Shipping: The Human Element.” Journal of Safety
Research37(4):401–411.doi:10.1016/j.jsr.2006.04.007.
Kafalı, M., Özkök, M. and Çebi, S. (2014), “Evaluation of
Pipe Cutting Technologies In Shipbuilding”,
Brodogradnja,Vol.65,No.2,pp.33‐48.
Martins, M. R., and M.C.Maturana.2010. “Human Error
Contributionİn Collision and
Grounding of Oil
Tankers.” Risk Analysis 30 (4): 674–698. doi:
10.1111/risk.2010.30.issue‐4.
O’Neil, W. A. 2003. “The Human Elementİn Shipping.”
Journal of Maritime Affairs 2 (2): 95–97.
doi:10.1007/BF03195037.
Oldenburg,M.,Baur,X.,&Schlaich,C.(2010).Occupational
risks and challenges of seafaring. Journal of
occupationalhealth,52(5),249‐256.
Özdemir,
Ü. and Çetin, M.S. (2017). Investigation of An
QuantitativeResearchontheCausesofSeaPiracy.12th
International Conference on Marine Navigation and
Safety of Sea Transportation, Gdynia, Poland, in
progress.
Özdemir, Ü. and Güneroğlu, A. (2016), “Cargo Type
Selection Procedure Using Fuzzy AHP and Fuzzy
TOPSIS Techniques: “The
Case of Dry Bulk Cargo
Ships”,InternationalJournalofShippingandTransport
Logistics(inprogress).
Özdemir, Ü. and Güneroğlu, A. (2017). Quantitative
Analysis of the World Sea Piracy by Fuzzy AHP and
Fuzzy TOPSIS Methodologies. International Journal of
TransportEconomics,inprogress.
Özdemir,Üand Güneroğlu,A. (2015).Strategic
Approach
Model For Investigating The Cause Of Maritime
Accidents. Scientific Journal on Traffic and
TransportationResearch,27(2),113‐123.
Portela,R.2005.“MaritimeCasualtiesAnalysisasaToolto
Improve Research about Human Factors on Maritime
Environment.”JournalofMaritimeResearch2:3–18.
Reiman, T., & Oedewald, P. (2002). The
assessment of
organizational culture: A methodological study. VTT
ResearchNotes2140.Espoo,Finland.
Roberts, S. 2000. “Occupational Mortality among British
Merchant Seafarers (1986‐1995).” Maritime Policy &
Management 27 (3): 253–265.
doi:10.1080/030888300411095.
Roberts, S. E. (2002). Hazardous occupations in Great
Britain.TheLancet,360(9332),543‐544.
Rothblum,A.M.(2000,October).
Humanerrorandmarine
safety. In National Safety Council Congress and Expo,
Orlando,FL.
Talley,W.K.2009.“Determinantsofthe ProbabilityofShip
Injuries.”TheAsianJournalofShippingandLogistics25
(2):171–188.doi:10.1016/S2092‐5212(09)80001‐1.
Toffoli,A.,Lefevre,J.M.,Bitner‐Gregersen,E.,Monbaliu,J.,
2005. Towards the
identi‐ fication of warning criteria:
analysis of a ship accident database. Applied Ocean
Research27(6),281–291.
UğurluÖ.2016.ʺAcasestudyrelatedtotheimprovement
of working and rest hours of oil tanker deck officersʺ,
MARITIMEPOLICY&MANAGEMENT,vol.43,pp.524‐
539.
Uğurlu Ö., Köse E.,
Yildirim U., Yüksekyildiz E., 2015.
ʺMarineAccidentAnalysisforCollisionandGrounding
InOilTankerUsingFtaMethodʺ,MARITIMEPOLICY
&MANAGEMENT,no.2,pp.163‐185.
Uğurlu Ö., Kum S., Aydoğdu Y.V., 2016.ʺAnalysis of
OccupationalAccidentsEncounteredbyDeckCadetsin
Maritime Transportationʺ, Maritime Policy &
Management,vol.7,pp.1‐19.
Uğurlu Ö.,2015. ʺApplication of Fuzzy Extended AHP
MethodologyforSelectionofIdealShipforOceangoing
Watchkeeping Officersʺ, International Journal of
IndustrialErgonomics,vol.47,pp.132‐140.
98
Xu, Z. and Yager, R. R. (2008), “Dynamic Intuitionistic
Fuzzy Multi‐Attribute Decision Making”, International
Journal of Approximate Reasoning, Vol. 48, No. 1, pp.
246–262.
