Tailoring ties:
Leadership in the spaces between –— When safety and ef fi ciency are essential
L. Vederhus
Leaders,decision-makersandpractitionersoftenfacechal- lengestotheirorganizationalvisions,decisions,andbeha- viors,duetotheuncertaintiesofoutcomesinanincreasingly complexandinterconnectedworld.Thenecessityforrobust andresilientorganizationsinrapidlychangingcontextshas alsoledtoashiftinleadershiptheorizing.Themorerecent complexity leadership theory (CLT) acknowledges the dynamicsoftherealitiesinwhichorganizationsandleaders operate, and emphasizes adaptive behavior of the whole systemfor copingin changingcircumstances.This genera- tivemodeldisplaysanadaptivespacethatiscoexistingwith, and balancing, exploratory-innovative/entrepreneurial initiatives and administrative-bureaucratic/operational needsoforganizationallife.Enablinglearningopportunities in the adaptive space is the most crucial feature of this model. Some critical issues have been raised regarding complexity leadership theory. One question has been whether CLTis complex enough, withthe main argument thatCLTfailstointegratetheleadersthemselveswithinthe organizational complexity where they are inevitably embedded. This can have the consequence of leaving us withakindoftruncatedversionofthereality,andtherefore notquitefathomingcomplexity.Understandingthatleader- ship isco-constructed withinorganizations’ evolvingstory enablesustovisualizeandapplyconceptsofself-organiza- tionandadaptationwithfarmoreconfidence.Furthermore, addressingleadershipasanemergingpropertyofcollective agencyindicates thattheoutcomesdepend onwholesys- tems.Thus,thetotalsystemisputtotheforeground,into thefigure,formingadynamicspaceonthebackgroundof constituent individuals and the connectedness between them. Behavioral characteristics of such large structured entities emerge by the way this dynamic organizational
spacemeets,alignswith,adaptstooractsonsurrounding challenges.
Thisapproachcontrastsprevious theoriesof leadership thatprimarilyhaveafocusonthepersonholdingtheposition definedasleader.Leadershiptheorieshavefavoredaposi- tion-centeredapproach for decades, withthe focal point being the leader, and the idea being that this position employsuniquepowersfordifferentwaysofleading,often in hierarchically organized structures. A whole-systems approachofferstheopportunityofusingnetworkmethods inguidingleadershipinthespacesbetween,integratingthe formalleaderpositioninsidethestructure,andfacilitating people’sperformancebybetterroutesofinformationflow andexchange,whereworkloadbecomesvisible,andpoten- tialthreatscanbemitigatedbyadjustmentandrearrange- mentsforproblem-solvingandlearning.However,different organizationtypesencountervariouskindsofchallengesin theircontexts.Thiscallsformorerefinedandcustom-made methodsinordertobeabletoguideadaptivebehavior.Of particular interest for thisarticle is organizationsdealing with risk and immediate consequences of behavior and performance. Due to the possible magnitudes of conse- quences,extendingintothephysicalenvironmentandpeo- pleexternaltotheorganization,safetybecomessignificant.
Further,likeformostorganizations,acompetitiveworklife alsoaccentuatescertaindemandsforefficiency.
Basedonthepresentedchallengesandthequestforways ofdealingwithcomplexity,thisarticleoutlinessometopics relatedtostructuralawarenessinorganizationsfacingboth complexityandrisk.Socialnetworkanalysis(SNA)equipsus withaconceptualframeworkforinterpretingpropertiesof structures. Assuming that awareness of such properties enablesorganizationstoanalyze,learn,andadaptintheir Availableonlineatwww.sciencedirect.com
ScienceDirect
j our na l ho me pa ge :w ww . e l se v i e r . com / l oca t e / or gdy n
https://doi.org/10.1016/j.orgdyn.2019.02.009
0090-2616/©2019TheAuthor.PublishedbyElsevierInc.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.
org/licenses/by/4.0/).
circumstances,we shouldtrytounderstand howdifferent typesoforganizations cantailorthetieswithintheir own structurestomeetsituatedchallenges,andstartadiscus- sion that allows for explorations of such ties in non-risk situations–—realorsimulated.Totuneintosomeresearch problems,forthistexttheycanbestatedasfollows:What metricsof network properties canreflect safety andeffi- ciency,and howcan network structuresbe optimized for safetyandefficiencypurposes?
LEADERSHIP IN DIFFERENT TYPES OF ORGANIZATIONS
Ina frameworkforexaminingleadership inextreme con- texts,adifferentiationoffourtypesoforganizationsbased oncontext and responsibilitieswasfounduseful. Trauma organizationsconductwork inforinstancehospitalemer- gencyroomsandambulances,wheresavinglivesisthecore occupation.Criticalactionorganizationsmayhaveamore activeroleincreatingtheextremityofthecontext,likein military combat units, SWAT, or fire, search and rescue teams.High-reliabilityorganizations(HRO)dependfunda- mentally on avoidance or prevention of extremity that could develop into crises and massive catastrophes, and inthiscategoryisbothnormalpolice,organizationsoper- atinginair,spaceandattheseas,andotherorganizations in riskconditions, like nuclear power plants. In the final category, all other organizations were labelled naïve, meaning they haveno initial role in an extreme context butmaybeexposedtosuchextremityforotherreasons,be ittornados,fires,attacks,andsoon.Complexitytheorycan contributeconceptuallyandanalytically(orpossiblyrather by synthesis than by analysis) as a useful approach for understandingandconductingleadershipinallthesetypes oforganizations.TheargumentforCLTbuildsonthepartof complexity theory that emphasizes open systems –— the complex adaptivesystems (CAS).CAS provides a broader accessfor advancements of leadership and several other organizationalandsocietalaspects.Sincebothdevicesand virtualspacesinterconnectuseverfaster,andthedynamics ofthefieldsthatemergecanappearasathreattocontrol- depending systems, the current position held by many is thatweneedadvancedmodelstohandlefuturechallenges.
Complex adaptive systems have featuresof connectivity, autonomy,emergence,non-equilibrium,non-linearity,self- organization, and co-evolution, which are concepts that enable us to understand the challenges better, but also opens ourunderstanding of coping by forming andtrans- forminglearningprocesses.AlthoughCAScanbevaluable forapplicationinorganizations,Iurgethatweshouldkeep in mind the importance of another part of complexity theory, complex physical systems (CPS). The rapidity in ourexploringand exploitingthe physicalworldaroundus alsoaddstothecomplexity.Evenifelementsstandfixedin CPS,asopposedtothoselearningandadaptingagentsina CAS, we still have to acquire a lot of knowledge about fundamentalbuildingblocks,andthelaws,states,powers and forces we interact with in CPS contexts. The future perspectives of physical systems capable of learning by artificialintelligencealsocalls ustoexercise cautionand bewareofagency,andconsiderintertwiningCASandCPS.
Ideas from actor—network theory, where living and non- livingelementsareintegratedby relevance,maythusbe beneficial adding to the social network theory approach fromwhichthistextelaborates.
This text is motivated by the particular challenges in leadership of HROstoday. New concepts andtechnologies that aregenerated for ourtime, likethe knowledge era, environmentalism, industry4.0,unmannedor autonomous vehicles andvessels,andsoon,challengeourrepresenta- tionsofrealityandworklife.HROsalreadyhavetomerge thepropertiesofCASwiththecontextofCPS,thusencom- passingmultiplelayersoffactorschallengingperformance, butalsoholdingmanyoptionaldynamicprocessesforcop- ing, some immediate, and some medium- to long-term.
Typical for HROs is theneed for safetyand efficiency for performanceoutcomes.Threatsmayarisefromthecontext thatsurroundstheiractivities.Thiscouldbeweather,waves, oceaniccurrents,wind,fog,icingetc.Pressuresandstres- sors from stakeholders, owners, customers, and others, couldfurtherintensifyextremityinthephysicalproximity ofrisks.Otherdemandingfactorsarepresentthroughtech- nology,bothwhenthetechnologyistoocomplicatedinits outfits, andin cases of failures or weaknesses. The most prevalentrootcausewhenaccidentsoccurishoweverstill ascribedtothehumanelement.Althoughtheseparationof humanandtechnologyinaccidentinvestigationmaynotbe veryhelpful,theinherentagencyandlearningcapabilityof human beings points to a responsibility we (yet) cannot assignto, orground in,technology,andthehuman ability of adapting to new contexts leaves us with the agentic powersofpreventionanyway.
The perspectiveshiftin leadership, fromthe onefocal point of power tocomplete networks exhibiting adaptive behavior,allowsustoreflectontheusefulnessofnetwork approaches for HROs.How can we assess safety and effi- ciencybymeansofadaptivenetworkproperties?
First, for the organizations, simply drawing diagrams could be of great value –— charting the relations of the people(or positions ofalternatingpeople, asis often the case in around-the-clockwork of HROs)by visualizing the linesconnectingthem.Theselines,orties,canbefurther explored, for instance looking into directionality,weight, type or quality. A joint session with the focus being the networkcanhelp establishsharedmentalmodelsofbase- linedandrestructurednetworks,andcanhavemajorimpact ontherepresentationandunderstandingoftheworkloadin thesystem,andpossibleinformationtransferdeficits.Mod- elingstructuresandenvisioninghowworkactuallyloadson personsbytheamountofconnections,thecoordinationand communication required, and the risk of shortcomings in informationtransfer,maysparkinitiativesofrearrangement andadjustmentscoherentwithadaptiveneedsofthecon- text.Becauseriskcanaccelerateintocrisisquitefastand unnoticed,itshouldbevitaltoestablishredundancieswithin the whole network, to form functionality for different phases.Exemplifyingwithdifferentscenariosandsimulating howthedistributionofinformationandtheassigneddeci- sionsaffectoutcomescanbeworthwhileinvestment.Basic featuresofourorganizationsmaybeinconcordancewithour convictions, but we could easily imagine some surprises here.Lackofinformation,orunclearcommunication,both impairingdecisionsandperformance,areoftenstatedasthe
most influential causes of incidents. Bringing the formal leaderintoanintegratedflexiblepositionmayalsoenhance performance by envisioning more rapid restructuring to elicitthefullcopingcompetenceofthewholesystem.This alsomeansthatsuchrelationalknowledgeandskillsbecome acomplexnetworkproperty(a“swarmintelligence”)rather thanameresummationof individuals’competencies.The typeoforganization,thepurposeoftheactivities,andthe kinds of challenges, willhowever influence explicationof successcriteria.
Second, for the scientific audience, analyzing network patternsand their outcome spaces have a clearscientific interestforunderstandingsafetyandefficiencyinorganiza- tions.Althoughsystematicsisavirtueofscience,studying complexadaptivesystemsmayrevealsomeinherentresis- tancetosystematics.Studyingtheemergenceofprocesses maybeamoreinterestingpurposethanconstructingtran- sitory typologies, taxonomies, or “truths”. Products like taxonomiesof patterns can have ashort shelf lifewith a dynamicworldview,andsharingknowledgeofprocessesmay beofhigherinterestthangeneralizabilityofpatternsthat could notbe achievable withinother particular organiza- tionalframes.Therearemajordifferencesinthepurposes andthescopeofactionfordifferentkindsoforganizations.
Usingnetworkmethods tounderstand processesandfunc- tionalityofstructuresforspecificorganizationsmaythere- fore have limitations for decontextualized application.
Knowinga rangeof suchstructuralfunctionality, anddys- functionalityoften pointed outin accidentinvestigations, canstillcontributetoactuallytransferringideasforadap- tivespacesbyexploringcontextualboundaries.
THE PROPERTIES OF NETWORKS
Social network analysis (SNA) and network metrics origi- natedinaconceptionofthe“socialorganism”,andthisis oneoftheareasofcontinuedeffortstomatchtheneedsfor understandingsocialandsocietaldevelopmentwithappro- priate methods. Although applied as sociometry and a method of investigating group dynamics, the measure- ments,ormetrics,nowreacheswideraudiences,andthe potentialof these methodsare far fromfully utilized. In socialnetworkanalysis,entitiesarerepresentedaspoints, andtherelationsbetweentheseentitiesarerepresentedas lines. The naming of points and lines appear in much variation,aspointsaredescribedasnodes,egoandalter, vertex (pl. vertices) and so on. Lines are referred to as connections, ties and edges, among others. The SNA- metricsareduplex,whichmeansyoucanhaveafocalpoint inthenetworkandarriveatseveralmeasuresofcentrality forthisvertex, orfind centralizationmeasures forwhole networks.Inasafetyperspective,theroleofthenodeasa personwillbetohavefullcompetencetodotheoperative workthispositiondemands,andtobeabletocommunicate with other persons within the operational environment according totheneeds of the situation bothpresent and infuture.Therepresentationofthesituationisreflectedin ideasofsituationalawareness,andrepresentationsofthe operative totality by the concepts of mental models and shared mental models. The impact of representing the operative network as a mental model, and the ways of
overcomingobstaclestosafetywithinthisnetwork,seems underestimated, however. Clarifying and conceptualizing the network properties and constituents of the ties that enablesafe and efficientwork is therefore based onthe advantagesofSNAinanalyzingcomplexnetworks.
Networkvisualizationcanbeseenasastaticformationof pointswithjust as static lines,and it couldtherefore be argued that SNA is inappropriate for understanding the complex dynamics of operative work. This is however a prematureattitude.Anysimulationofdynamicspresupposes visualization of theconstituent parts, and changesin the visualconstellationsshouldmirrorthedevelopmentsinreal networks. This calls for other dynamic metrics, but the conceptualizationsofnetworksmusthaveaninitialexplicit design.Anydescription canreturnareduction,sofurther elaborationandrefinementofconceptsarenecessary.Itis possibletocomparedifferentnetworkstructures,andcon- trast their structuralmeasures, either of node-metrics or network-metrics,thatenableustounderstandwhatdistin- guishesefficient andsafestructuralpropertiesfromother properties.
With complexity leadership and network methods as pointsofdeparture,hopingthereadernowclearlyvisualizes afeasibilitylinebetweenthosetwopoints,thisarticlewill nowturntoadiscussionofexistingmetricsforonespecific high-reliability organizational context, where safety and efficiency are essential. This can also help contribute to develop the metrics further, by exploring measurement needs in order to have process- and outcomes-oriented leadership guided by such methods. Fundamental to any copingstrategy isthe understanding ofthe organizational structure–—thatis,thenetwork–—anditspropertiesinany challengingsituation.“Tailoringthe ties”,whichprimarily meansenablingorganizationalconnectionsandcommunica- tionthatisadequatefordifferentphasesandpurposesofthe work, may have great impact on the flow of the work (efficiency) and the workloadbeing handled within limits (safety).Forleadership, westillneed positioneddecision- makingforsafetyandefficiencypurposes.Tocatalyzesuch leadershipinthespacesbetween,wealsoneedmoreela- borateresearchanddevelopmentofnetworkmethodology.
SAFETY AND EFFICIENCY IN HRO
Forhigh-reliabilityorganizationsandoperations,whererisks may emerge in complex ways, safety has been of major concern, also stimulating various approaches in scientific discourse.Authorities andgoverning bodiesinvolve inthe practicing industries by regulations and laws for perfor- manceandcontrol systems,serving protection ofpeople, material, organizations, environments and future aims.
However,in competitiveindustries,workersandorganiza- tions may be put under strain on safety, since the more efficientbiddermaybepreferredwhenthatiscost-effective forthebuyer.Afterall,moneytalksalsoinoperativebusi- nesses.For short-termactivities,efficiency measuresmay havedifferent constituentsthan for wider-reaching activ- ities,whereeffectivity(thetotalgoalachievement)canbe reachedinmanyways,moreorlessefficiently.Safetycanbe thwartedbythe questfor efficiency interms oftimeand costreduction,buttheparadoxliesinthesamemeasures
potentiallyenhancingsafetybymakingtheperformingwork- ersmorealertandaware.Itisassumedthatriskincreasesin routinesituationswhereyoulowerandlimitconcentration andawareness,whereasahighlyawareandfocusedperson maybelesspronetoharm.
Efficiencycanbeseenasaranking-andselectionfactorof organizations competing for assignments in for instance offshoreindustries,whichisthechosencaseforillustration inthistext.Duetothepracticeof tender-preparationsin such contractual industries, with a multitude of parties involvedinthefinaloperativeactivity,thebuyingoperator choosesamongbidders,orhireentitiesbasedonforinstance availability,orpreferencesforspecialequipmentorcondi- tions.Theconstellationofamultitudeoforganizationsmay thereforebeuniqueineveryoffshoreoperationand,adding tothecomplexityandriskcomprehension,itmustbenoted that new workers are recruited continuously, and people relocatewithinthesepracticesasinothers,soanynetwork establishedforperformanceofanoffshoreoperationmaybe noveltoallthoseinvolved.Thetiesmaythereforebenovel inany operation,if notcompletely then atleast tosome extent,wheretiesconnectpositionsmannedbynewcrew andcrew unknowntooneanother.Communicationandall kindsofinformationexchangearepresent –—or absent–—
betweenthepositions involved.These tiesarethefunda- mentalcondition for safeand efficient executionof high- reliabilityworkinintegratedoperations.Inmanyorganiza- tions, recruitment focuses muchon the capability of the person,andtoalesserextentonthetiesthatgoalongwith thepractices.Skillsincooperationandcommunicationare oftenexplicitlyexpectedasasafetyprerequisite,butrarely addressedforevaluationoftheoperativenetwork.
Dimensioning of manning for offshore operations is beyondthe scope of this text, but shouldbe mentioned.
Thenumber and diversity of positions havesome impor- tance for safety due to the structuring of work and the redundancythatmaybeneedediftherearehighlycritical consequenceswhenonepersonhasaworkloadthatexceeds possiblehandling,orifapersonoughttobefunctionallyset completelyoutoftheloopduringtheoperation.Anexam- plecan befrom aviationin the number ofpilots andco- pilots for long-distance flights. For example will severe illnessinoneairlinepiloteffectuateanotherpilottotake possessionofthetasks, andthepossibility,plausibility or probability of suchevents underlies theredundancy con- siderations.
Asocialnetworkanalysisapproachtoquestionsafetyand efficiencydoesnotimplythatanodemustbeapositionor person.Approachesmayconsidernodesaswholeorganiza- tions,asdepartmentswithinorganizations,orevenasthe operationsthemselves,whereorganizationscanbeseenas unitsdispatchedtosuchoperationsinaflow.Morecomplex approaches,forinstancebytwo-modenetworks,canbeof interest,wherepersonsanddiverseorganizationsinbipar- tite networks may be of importance to understand the empiricalties’ mobilityproperties,depending onthepur- poseandownershipoftheapproach.
In high-reliability work, there is quite an amount of imperative safetyregulations, but the continuous innova- tionsin thesefields of work,andextensionand spreadof operational areas around the globe, result in a delay for authorities to identify the risks arising, and to advance
preventive initiatives to mitigate such risks. A network approachmayassistinunderstandingsomeoftheriskfactors thatcansetin.
Itisvitaltoelaborateontheroleofsafetyandefficiency as awhole,tomake visiblethe balancing ofissuesfor all involved parties. Leadersand decision-makerscan contri- butetooptimizenetworkstrategiesinworkwhererelations andcommunication areessential.The simple elements of nodesandconnectionsamongtheseinnetworksareopento analysis through graph theory, algebraic approaches and spatialapproaches. Howtousethe metricsforsafetyand efficiencypurposesdependsontheoperationalizationofthe propertiesofthetiesinsuchrespects.
AN ILLUSTRATORY CASE: THE NETWORK ONBOARD IN ANCHOR HANDLING OFFSHORE OPERATIONS
The example chosen must be regarded as a small-scale network structure,allowing ustoexploresome basicfea- turesofnetwork properties,buildingthe groundfor more advancedapproachestonetworkmethods applicationsfor safetyandefficiencyinanyhigh-reliabilitywork.
Inoffshoreoperations,thephysicalsurroundingsarealso complex,withamultitudeoffactorsinfluencingthework.
The weather window is for instance an interpretation of probabilityofchangesinwindandwaves,wherethelimita- tionsforworkalsorestsonthecapacityofthevessel.Also, manyvesselscanbeinvolvedinthesameoperation,repre- senting separate organizations in tandem operations like anchorhandling.Tofurthersupplementtheunderstanding oforganizationalcomplexity,therearealsoconstellationsof manyorganizationsrepresentedonboardoneandthesame largevesselforadvancedsubseaoperations.Anchorhand- lingoperationsandrigmoveshavebeenanextensiveprac- ticeoffshoreformanyyears,whereexplorationofresources likeoilandgasbeneathseabeddependsonmovingdrilling rigsintodifferentoceanicareas.Suchoperationshavebeen associatedwithparticularlydemandingwork,wheremargins aresmallforsevereoutcomes.Afterthelossofthevessel BourbonDolphinduringanchorhandlingin2007,theOfficial Norwegian Report summarized recommendations to strengthenbarriersforsafetyinfutureoperations.Amongst thesearethehumanelementsaspectsoftraininginsimu- lators, certification of crew, familiarization in the vessel environment, start-up-meetings and communication, and extended responsibility between actors within the total operationduetothemreciprocallyactinginsuchoperations.
Collaborating,cooperating andcommunicatingindemand- ingworkenvironmentscandrainmentalenergy,butcanalso allowforremovaloftheworkloadbymoreadequateinfor- mationflowandresourceallocationstoensurealltasksare conductedinaccordancetoplans.
Even though organizationaldimensioning isoutside the scopeofthisarticle,thesizeofanetworkiscrucial.Onecan choosetostartanalyzingthenetworkwithineachorganiza- tionorapplythecompleteoperationalstructureforanaly- sis, trying to identify components, clusters and cliques withinthoseoperativestructures.Theauthorityindifferent stages of operations alters, however, so in order to be applicable,theanalysiscouldbeorganizedforeachleader
positiontobeabletoidentifyherorhisnetofliabilitiesand responsibilities as well as the complexity of the decision- making network. To exemplify, oil-drilling rigs may be in chargeofthesequenceoftheunmooringfromseabedpre- vioustorig-moves,whereforinstancefourdifferentvessels arepresenttobecoordinatedindoingthiswork.Anengi- neering-or consultancy companymaybeincharge of the scope to be followed for the operation. Each vessel is responsible for decisionsregardingthe break-downof the scope into manageable tasks and resource allocation onboard. Failures,shortages or deficits in anypart of the integrated operation may branch off quite fast, as the connectionssupporttheflowandsequenceinallotherparts.
Animportant proceduralaspecttodayis therequirement, responsibility,and authority ofevery person regardlessof positiontousetheword“stop”ifneededduringanyopera- tion(whetherthisappliesalsooutsidetheNorwegian con- tinentalshelfisnotknownbytheauthor).Therecanbesome resistance todoso, sincetheconsequences may beeven morehazardous forthe entire network ifyou impedethe ongoingevents,and,besides,canhavefar-fetchingconse- quencesforreputationsofpersonsororganizations.
Workingaroundtheclockmeansthatpersons thathold thepositionsalteraccordingtotheirshiftroutines.Ina24-h period,somehavesixhoursonandsixhoursoffcontinually, (andskewitsixhoursforthenextturnoffshoretopartthe burdenofthenightshift),othershave8hshiftsonandoff, and others have even longer periods. This work schedule standsforfourweekson,andthentheyhavefourweeksoff, or,forsomeorganizations,twoweeksonandfourweeksoff.
The focus onteam and team qualitiesfor safetyin high- reliabilityworkcanthusbeseentobecultivatedwithineach operation, but encounter resistance between operations.
However,networksoffriendshipsandotherlinkagesappear betweencrewalsowhenonshore.Furthermore,specifically preparedteamtrainingcoursesareconductedinsimulators for offshore crew, recognizing the role of the teams in enhancingsafety. Thecontentsof theseare,amongother things,sometheoriesonpersonality,team,risk,andsafety regulations, some practical exercises on safejob analysis andexecutionoftasks,anddebriefingandevaluationofthe simulatedtasks.Thefrequencyofsuchteamtrainingcourses maynotkeepupwiththeneedsevolvingfromrecruitment andreplacementpracticeofpeopleindifferentpositions, but team training courses for complex operations can be expensive, and may thus not turn out favorable among a multitudeofothersafetyinterventionswithinorganizations.
Theimportanceofthetiesis,however,unquestionable.
Thetiesconstituteteamqualities,andtheoperationalnet- workcanbeassessedifsafety-andefficiencymeasuresare clearlyexpressed.
Thenetworkformationthat“provideanaturalcounter- balanceinsocialexchange”issaidtobethetriad,butwe shallalsoconsiderdyads.Someconstellationsinoperative workaremadeupof combinationsofdepartmental dyads andtriads, suchas onboardtheseanchor handlingvessels withamanningofthreeonthebridge,amanningofthreeon deck,andlikewisetwoorthreeintheenginerooms,aswell as twoor threeinthe service functions.This is ofcourse almostdoubledastotal,asnewshiftstakeover.Forspecial operationstherearealsootherexpertisepersonnelserving thecranes,ROV-equipment,andsoon.Thebuyerisoften
represented onboard, as well as other accompanied resourcesfordifferentpartsofthescopeorservices.
Avisualrepresentationoftheinvolvedpositionsandtheir roles,aswellasthenamesofallpeopleattendingwillenable a map of the ties that eventually underlie the operative phases.Anynetworkchangesmustconsequentlyberepre- sentedinthemindsofinvolvedresources.
Thoseinvolved,though,maymultiplyrealizethenetwork structurethatcan beformed.They mayrepresent people accordingtotheformalpositionsordepartmentalaffiliation, to operational sequences’ connectivity, to statuses, their lengthofservice,personalpreferences,andsoon.Theexis- tenceofidiosyncrasiesofnetworksmaybecounteredposi- tively for involved crew by emphasizing thecomplexity in more formallyapparentnetworksbymeansofaggregatednetworks andfurtherbymultiplexity.Aggregatedtiesexistwithinone networkwheresome tieshave several qualities.Multiplex networksarecomprisedoftieswithpluralrealizationswhere thesamepeopleattendtwoormorenetworks.Theconceptof richtiescarriesthesepropertiesofpeopleknowingeachother byamultitudeofinformationnetworks,enablingunderstand- ingof botheach other and thetie, to thelevelof astrengththat mayaidbothsafetyandefficiencypurposes.
Inthisway,ideasthatarealreadypresentinteamthe- oriesforsafetypurposes,likethoseoforganizationalsafety cultures,experiencediversity,commonoperationalservice, durationofexperience,languagecomprehensionanddiffi- culties,culturaldiversity,tonamesome, canbemodeled intothestructureofties.
PROPERTIES OF SAFETY TAILORED TIES
Tiesincomplexnetworkscanbeanalyzedbythedirectionor bi-directionalityofthe lines, andareciprocal tiemaynot necessarily bea bettertie for allpurposes. Beingable to receiveinformationmaybeasimportantasbeingabletogive information,torequest,orcommand,duringanoperational phase,andtodosofromorontothelocation(s)whereeither theinformation or theact imposedby it hassignificance.
Closed-loopcommunicationhashoweverbecomecustomin HROs. Duetotheenormousamountofinformation, any worker capableofestablishingapictureofallothers’tiesaswellas own,maybeabletoadjustquicklyiftherearebreakdownsof anysortthatmayinfluencetheflow,orincasesofemerging hithertounfamiliarsituationsthatdoesn’tcomplywithincor- poratedpathsofinformation.Agreed-uponpathsofinforma- tionflowoftenfollowthehierarchicalstructure,andthismay wellhaveanefficiencyadvantage(tobediscussedinthenext section),butmightobstructsafetybecauseofthedistance fromsenderstoreceivers.
Thetieshavemetricsofin-degreeandout-degreeinSNA, observedby direction. Theseare basic measures, butthe metricsiscomplicatedbyideasofstrength,weight,inten- sity,distance, path alternativesandcircularity,as well as theimportanceofthesignofsuchquantifications.Ideally, one could argue that a neutral tie is a better starting segment than a valued tie of either positive or negative value.Inreallife,“likesanddislikes”amongpeoplemakea markonrelations,andtheseneedneitherbereciprocal.The negativesignofarelationallinealsoputsstrainontheother relations,well-knownintheSNAliteratureoftriads.
Aleaderwhounderstandstheimportanceofrelationsis oftenabletobalancethetriadsbyestablishingtriadsthat seemtohaverichandpositivesignedties.Thismakesthe distancebetweennodesminimalindepartmentswhoworkin same shifts, and deep knowledge of each other relieve human resources to keep full attention to the important signalsintheenvironment,alsoessentialforsafety.Itmay alsobeaddedthatsuchshort-distance-andrichtiesenables efficiency,asbehaviorcan bespeededupwhen youdon’t havetostopfor demandingarticulationsandexplanations towardsyoursignificantlytiednodes.
Nodearrangementisthereforeanadjustmentbasedon real-timechallengestosafebehavior,tooptimizethepos- sibletiesof exchange.A newworkerortrainee might for instancebemeteitherasaninterruptivenode,weakening safetyinanestablishedtriad,orsheorhecouldbeinvolved inatriadforlearningbyshadowingeitherofthenodes,orby successivelyfollowingeachnode.
The mostimportantsafetyego-metricsmaywellbethe richnessoftiesintriads,basedonanalysisofeachegointhe organization. However, safetydepends on thewhole net- work,andcentralityoftheconnectiveleader-rolemustbe emphasized. Each department in anchor-handling opera- tions may have a leader-role to bridge the information betweenvariousdepartments.Betweenthese,thereisalso aconnectionbytheleaderofthewholeoperation,orbythe representing node of this position. The ability to give immediatemessagesalsoacrossbridgesareofutmostimpor- tanceforsafetypurposes,andthecentralityofthisposition aswell as thecentralizationmeasuresof thenetwork are essential. Centralization is overall-measures of relations.
Quantification oftherelationsisalsoimportant,sincecal- culationsbuildonassumptionsofscalesofequalintervals, andthedegreemeasurecannotreflectanysafetyqualities ofanetworkinitself.Rather,thefields,orspacesbetween, that reflect the forces for optimal information flow with regard to safety and efficiency point to closeness- and betweennessmetrics,aswellasdensitymetrics.
Thegeodesic,shortest-path,measureisthereforealsoa necessaryanalysisinthiscomplexity.Tobeabletoarriveat informationfordecisionstobemade,thereisanadvantage ofshortpathsforinformationexchange,andatightstruc- turewillreflectthisproperty.Decisionmakinghaslocations withinnetworks,andfordecisionstobebasedonavailable information, the paths and flow of information must be explicit.Naturalisticdecisionmakingisdifferentfromtradi- tionalconceptsofdecisionmaking,beingcharacterizedby doubtsanduncertaintyinrealandcomplexsituations.Three differentuncertaintysituationshavebeensuggested:inade- quateunderstanding(owingtoequivocalinformation,owing tonovelty,orowingtoinstability),incompleteinformation (completelackofinformation,partiallylackinginformation, or unreliable information), and undifferentiated alterna- tives(equallyattractiveoutcomes,conflictamongalterna- tives,orincompatibleroledemands).Thestructurecannot initselfcompensatefor theuncertainties,but itcancon- tributetoinformationexchangewhere more information, morepreciseinformation,fullerdescriptionsormorereli- ablesourcesareneededfordecisionstobemade.
Another aspect in dealing with safety purposes is the redundancies systemized into structures. Equipment and technologyhave awiderangeofredundancysystems,and
althoughhumanresourcescanbesubstitutedbytechnology to some extent, there are tasks that depend on human behaviorandadjustmentabilities.Buildingredundancyinto networkstructuresmeansthatanodemayhavetotakeover othernodes’workiftheyforanyreasonbecome“un-tied”
within the network. Replacing entities automatically and immediatelyis namedwarm redundancy.Coldredundancy meansthatsomesignalhastobegivenbeforeanothernode is substituting.These re-arrangements may carryrisk and delayinthesystem,dependingonthepathtobetravelled for the information to be processed and a signal to be decidedonandpassedalongapathfurtheron.Theterms of active and passive redundancy arealso important in a networkapproach,astheactiveimpliesthatthesubstitute isalreadyactivewithinthenetwork,whereasthepassivehas tobe activatedinto thenetwork as anew personon the position.
So,redundancybyalternativepathsandreplacingnodes isalsocentralinpreventingissuestoescalateintocrises.
Two final comments on safety: workload has a major impactontheabilitytotransferinformation,andinstress- fulsituations,ithasbeenarguedthatanyverbaleffortmay leadtooverloadandrisk.Thismustbementionedbecause relieving the load of nodes by minimizing the need for information exchange enhances safety by balancing the workloadfortheinvolved parties.Thequality ofthe ties andtheimplicitknowledgethatconstitutesthesecanthus beastrongsafetyelement.Theotherpointtomakehereis that the ideaof trust may leadto misconceptions of the situation. Trust can reflect a valuable reliance between people,andcharacterizethetiesinanetwork.Ahighlevel oftrustmaynotnecessarily beagoodmeasure ofsafety, though.Trustcanalsobecompletelyirrationalandbased on other aspects of a tie than the organization network requirements. People may be considered trust-worthy in paralleland private networks, and still bethe nodethat breaks the chain orflow in a more demanding workload situation.
EFFICIENCY CONSIDERATIONS
Using theavailable resourcesin theoptimalwaytoreach goalsisonewayofdescribingefficiency.Anorganizationcan beeffectiveinthemeaningthattheyreachtheirgoals,but thismaybeachievedinaninefficientmanner.Tocompete for contracts in the high-reliability industries, safety and efficiency are interweavingnow. Highefficiency hasbeen seen as a potential risk-contributor, by rushing progress ratherthanmakesureandcertainthatallinvolvedentities and processes keep the pace. Overview of the totality is necessary tobe able toidentify discrepancies, risks, and issuesthatmaypave thewayfor alurkingcrisis. Manage- mentisthusnecessaryinatleastthreedifferentphasesthat canbesegregatedforoperationalpurposes.Asmentioned earlier,someagreed-uponpathsmaybelessefficientwhen deviationsoccur,andthismayoftenbethecase.Adaptingto therealityofnetworkchangesmeansthatfurthermodifica- tionsofthenetworkorthepathsforinformationflowmust bepreparedtobeinstantlyeffectuated.
First,riskmanagementisthebaselineconditionforany operation in HROs. This may call for a specific network
structure, optimal for safety and efficiency purposes, as discussedearlierbydistance(geodesics)andcentralization withhighin-degreevaluesforleaderstoidentifyanypoten- tialissuesofconcern.Preventingharmcanbebuiltintoall practicesinthisphase.
Next,ifanykind ofarising issue hastobemanagedto preventafurthercrisis, adecisionmaybemadetoswitch thenetwork structure intoaredundancy structureuntil a safestatehasbeenachievedandtheissuecanbehandled.
Thenatureofthisproblembringscentralitymeasuresinto call,tofocusresourceswhereneeded.
Third, a crisis phase may evolve –— either suddenly or sneakingly–—andcrisismanagementmaycallforstillother structures, depending on the number of abled nodes to engage and the dispatching procedure of the remaining resources in such a situation. The priorities must always involvearealisticmental modelof thenodesandties,to leadaphaselikethisefficiently.Theroleofaleaderposition istoenablethereductionofthealreadyapparentdisaster throughsuchresourcedispatching.
Tosumupsomeoftheideas,shortpathsenablefastand precisecommunication,servingbothsafetyandefficiency.
Density may lead to information overload, so structuring withclustersoftriadsandbridgingnodesfromeachtriadto centralnodesbetweentriads,seemstheoreticallyadvanta- geous.However,thispresumesthattiesarecultivatedinto positivelysignedandrichedgeswithinthetriads,andinto positiveandrichbridgesbetweenthedifferenttriadsaswell as betweentriads andcentraldecisionmakersor leaders.
Thebridgingnodesmusthavespecificpropertiesascatalysts intheflow,enablingtheconnectionstodeliverinformation without beingasked, whenthisisnecessary.Keypositions maythusbeidentifiednotnecessarilybytheirabilitytogive and receive information, but by the way theyoperate to enableotherstodoso–—apropertythatmaybecontraryto ideasfromsomeotherSNA-studies,butquiteinaccordance with complexity leadership theory. Also, initiating nodes maybedispersed accordingtothespecificity ofthe tasks inintegratedoffshoreoperations,andcentralitymeasures forbothsafetyandefficiencypurposesmaythereforeben- eficiallybeappliedbesidesforinstanceanobject-oriented flow-chartapproach,wherethescopeisbrokendowninto subsetsofthetaskstobedone.
Socialnetworkanalysishasanadvantageinoperationa- lizing ideas fromteam theoryin high-reliability work and complex networks. Leaders and decision makers should stress the node arrangement and tie quality measures beyondthequalitativeapproachthatisprincipaltoday.
ILLUSTRATION: AN ANCHOR-HANDLING VESSEL CONFIGURATION
The structure of positions (Fig. 1) in the example of an anchorhandlingoperationwilltypicallybefourdepartments oftriads.Thecaptain,chiefofficer,andsecondofficerareon thevessel’sbridge.Thechiefengineerandtwomoreengi- neersareorganizedforthemachine/engineroomandelec- trical equipment. A catering officer and two additional servicecrewconstitutesathirddepartment,andaforeman withtwoadditionalableseamenareatthe vessel’sdeck.
Thechiefofficerisofteninchargeofmaneuveringthevessel
andcommunicatingwiththedeckandenginepersonnel,as well as withcateringif necessary during operation (since these often have some medication access). The captain keepsmonitoringallactivities,butmaymostlybeinteract- ingthroughtheothertwobridgecrewaslongasthissuffices (however,allworkpermitsaresignedbythebridgeperson- nel,and indemandingoperations thecaptainalsodothis during operation, and she or he also keeps all contact onshore throughout operations). The second officer may be responsible for the main tools, the winches, during operation,constantlywatchingbothdeckactivities,bridge instrumentsandcamera-monitors,tounderstandtheforces actinguponthe vesselwhenconnectedboth toseabed or anchoringarrangements,andtoarigthatisalsoboundto othervessels.
Illustrating the communication or connection workload basedonthedegreecentralityofthenodes,Fig.2showsthe roleofthechiefofficer(thelargestnode),butthephysical andmonitoring workloadisnothighlightedinthisillustra- tion.Attentionisalsoaffectedbytheneedtoarticulateand make aconversation. This particular structure mayguard againstsuchsafety-stressors for crewwithheavy physical workloadorlittleexperience.
The averageshortest-path-length, thegeodesic, ofthe vesselnetworkis2.45.
The clustering is of medium size, due to the triadic structuringintofourdifferentdepartments,withameasure of.47(Table1).
The structure of a smallnetwork likethis is based on substantial and fundamental characteristics of the task- specificity of the positions, and the need for redundancy andbackupineachtriad.Itispossiblethattherealnetwork structureonboardvesselscouldbeorganizedotherwise,and also,thattheyfunctionotherwisein otherphasesofwork thanwhatisidealwhentheyenterthe500-msafetyzone encirclingthedrillingrig.Examplesofotherstructuresmay includestargraphs,ringgraphs,fullgraphsandtreegraphs Figure1 Arrangementsofacommunicationstructureonboard avesselduringademandingoperationofanchorhandlingofan oildrillingrig.Nodes 1—3areenginecrew,nodes10—12 are cateringcrew,andnodes7—9aredeckcrew.The(hierarchically more central) nodes 4—6 are bridge personnel onboard the vessel
(Fig.3).Thecentralizationandclustering(transitivity)mea- suresfor these12-node examplescanbecomparedtothe vesselmeasuresinTable2.Varyingbetween0and1,wesee thatthedegreecentralizationonthevessel,asbasedonthe variabilityofallactorindices,isnotveryhigh.Furthermore, theclusteringismediuminsize,aswouldbeexpecteddueto thequadrupletriadicstructure.
Other centralization measures are also of interest. In additiontothemeasure ofdegree (.23),representingthe normalizedamountofadjacentalteriforallverticesinthe network,thebetweenness(.65),closeness(.26)andeigen- vector(.62)measures express,respectively,theextent to whichanyandallactorsisbridginganytwootheractorsin thenetwork,howcloseallactorsareonaveragebasedon paths,and how connected any and all are tohighly-con- nectednodes.
UsingWatts—Strogatzrandomgraphgenerationmodeling enablessomecomparisonofthesemeasurestoothermea- suresofgraphshavingsmall-worldproperties.
Examplesoftworandomlygeneratednetworksaregiven inFig.4,andthetotalof100randomlygeneratedgraphsin Fig.5.
Probability modeling (Fig. 6) shows that the assumed efficiencymeasurebyusinggeodesicinthevesselnetwork isnotverylow(2.45)comparedtothecompletelyrandomly generatedgraphs(measuresforall100inAppendixA).The
interaction arrangementintotriadswithacentraldepart- mentcanhardlybemorecogent,butinreality,theedges will have different interaction qualities, and quantifying thesequalitiesintodirectedandvalued(maybealsosigned;
thatis, openinguptonegativelyvaluedrelations)graphs, may add important information tothe complexity of the operativeinteractionsandthustounderstandingofinduced riskandsafetyinterventions.
Thiswouldalsoentailalteredclusteringofthecrew.The clusteringmeasuresforall100randomlyconstructedgraphs (tobefoundinAppendixB)havearangefrom.07to.60,as comparedtotheachieved.47.Itisdifficulttopredictthe effectofvaluedanddirectedconnectionsontheclustering, aswecouldexpectquitesomevariationsonboarddifferent vessels.Thesafetyculturewouldprobablymatteralot,and safety-orientationmaycounteractpolarizationanddisorga- nization, and isolateswould quite likely quitthe employ- mentsituationandseekotherworkandvessels.
DISCUSSION AND CONCLUDING REMARKS
Thistexthaveoutlinedsomebasicissueswhenapproaching complexitywithanetworkmodel,assumingthattheleader- shipisembeddedwithinthestructuresandco-constructed through entire performances. The organization type may
Table1 Sumupoflocalandglobalcentralitymeasuresofallnodesinavesselnetworkinoperativephases
Egono. 4 7 3,6,10 5 8,9 1,2,11,12
Globalcentrality(shortestdistancetoallotheralteri) 17 22 23 26 31 32
Localcentrality(absolutenumberofadjacentalteri) 5 4 3 2 2 2
Localcentralityasarelativemeasureofmaximumpossibleconnections .45 .36 .27 .18 .18 .18
Figure3 Fourstructuresof12nodesandtheircentralization measures.Topleft: stargraph,topright:ringgraph,bottom left:fullgraph,bottomright:treegraph
Figure2 Thedeckforeman(7),andthechiefofficer(4)are centralduringtheactualanchorhandlingphaseofanoperation.
As long as everything follows the scope without errors, the operativeworkloadofthecaptain(5)canberelievedforother leadershiptasks,includingtheconnectiontoothervesselsand rig,aswellasonshorecommunications.Thisisrarelythecase, duetotheuncertaintiesofbothnaturalforces,complexitiesin equipmentandunforeseeneventsthatemerge
definecontextual necessities for analysis, andthe chosen focusonspecificdemandsforsafetyandefficiencyincom- plexanddemandingenvironmentscanjustserveasastart- ing point of more advanced and empirically grounded analyses of networks in high-reliability work. Safety and efficiencymaybeincompatiblepropertiesfor someindus- tries,butthecompetitioninthemarketputspressureonthe organizationstodemonstrateboth.Theteaminteractionhas aparamountfunctionforhighlyspecializedworkincomplex environments,andindicatorsoffunctionalitymayfacilitate both training and other safety interventions like barrier management. Further work should expand the measures- portfolio and constructa range of data-collection instru- mentsforaselectionofinteractionalqualitiesthataffects thesafetyandefficiencyinhigh-reliabilitywork.Thefocus onstructuresratherthanprocessesmaylimitthescopeof leadershipagency,soknowingthatstructureformationisa dynamic process with adaptive aims to meet challenging circumstances and risks should enable organizations with moreopportunitiesofexplorationsforutilizingthepotential ofthenetworkproperties.Furthermore,eventhistorymod- eling could also be advantageous when we establish an understandingofcomplexadaptivesystemsandthepossi- bilitiesandlimitationsofsocialnetworkmetricsinassessing ourorganizations.
Leadershipmaycertainlybeunderstoodasan emerging propertyoftheorganization,butcannotbeunderstoodas Table2 Overallcentralizationmeasuresforsomenetworks
ofthesamenodesizeandtheillustratoryvessel-network Measure Degreecentralization Clustering Graphtype
Star .83 .00
Ring .00 .00
Full .00 1.00
Tree .11 .00
Vessel .23 .47
Figure 4 Two 12-node networks based on Watts—Strogatz graphgeneration
Figure5 Thetotalof100generated12-nodenetworks(Watts—Strogatzgraphgenerationmodel)
separatefromdecision-making.Anyact,anyperformance,is amanifestdecision–—whetherintendedandconsciousornot –—and anetwork approachallowsustoanalyze theways structuresaffectssuchdecisions.Italsoallowsustoestab- lish networks that enable learning opportunities in the adaptivespace.Suchlearningopportunitiesmayneedsimu- lations in high-reliability organizations, where real-life
explorations can have hazardous consequences.Forestall- ing,anticipationandimaginationmaybenecessarytoequip anorganizationwithadaptivecapacityincomplexenviron- ments, and testing the robustness of the organizational structure by pushing the demands and pressures may be beneficial, butshouldalwaysgoalongwithethicalconsid- erationsandpotentialsforerroneouslearning.
Figure6 Topleft: averagegeodesics for Watts—Strogatzrandomly structured graphs with12 nodes for varyingprobabilities (between0and.99),topright:clustering(transitivity)forthesamerandomgraphs.Bottom:normalizedvaluesofclustering(black) andgeodesics(red)(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthis article)
SELECTED BIBLIOGRAPHY
Thereareamountsofresourcesincomplexitytheory,leader- ship theory, organization theory, social network analysis methods,andcontributionsregardingsafetyandefficiency thatcouldbeofinteresttothereader.Also,therearealots of empiricalresearchcontributingtodevelopments onall thesetopics.
Forcomplexityleadership theory,thearticlefromUhl- Bien,M.&Arena,M.(2017),Complexityleadership:Enabling peopleandorganizationsforadaptability,inOrganizational Dynamics,Volume46,Issue1,Pages9—20.ISSNhttps://doi.
org/10.1016/j.orgdyn.2016.12.001;andfromUhl-Bien,M.;
Marion, R. & McKelvey, B. (2007), Complexity Leadership Theory: Shifting leadership fromtheindustrial age tothe knowledge era, in The Leadership Quarterly, Volume 18, Issue 4, Pages 298—318. ISSN 1048-9843, https://doi.org/
10.1016/j.leaqua.2007.04.002haveinterestingpoints.
Someimportant questionstotheir approachhavebeen raised by Tourish, D. (2018), in the article Is complexity leadership theory complex enough? A critical appraisal, some modifications and suggestions for further research, in Organization Studies. ISSN 0170-8406, https://doi.org/
10.1177%2F0170840618789207.
Also,forelaboratingonleadershipemergingindifferent kinds of organizations, the articles by Hannah, S.T.; Uhl-
Bien,M.;Avolio,B.J.&Cavarretta,F.L.(2009).Aframework forexaminingleadership inextremecontexts, inTheLea- dershipQuarterly,Volume20,Issue6,Pages897—919.ISSN 1048-9843; and Hannah, S. T. & Lester, P. B. (2009). A multilevelapproachtobuildingandleadinglearningorga- nizations, in The Leadership Quarterly, 20, 34—48, have interestingperspectives.
Forsocialnetworkanalysis,thereareusefulresourcesin thebooksbyJohnScott(astartingpointcanbeWhatissocial networkanalysis,BloomsburyAcademic,2012),aswellasin the advanced book by Stanley Wasserman and Katherine Faust (Social network analysis. Methods and applications, fromCambridgeUniversityPress,2009).
Naturalisticdecisionmakingreferredtointhetextcanbe found in the article by Lipshitz, R. & Strauss, O. (1997), Copingwithuncertainty:Anaturalisticdecision-makingana- lysis,in Organizational Behaviorand HumanDecision Pro- cesses,69,149—163.
Also,fortheinterestedreader,thereportfollowingthe loss of Bourbon Dolphin can be found here: NOU 2008:
8. The Loss of the “Bourbon Dolphin” on 12 April 2007.ReportfromaCommissionappointedbyRoyalDecree of 27 April 2007. Oslo, Ministry of Justice and Public Security.
APPENDIX A
Tableofrandomlyconstructedgraphs’geodesics [1] 2.439394 2.439394 2.439394 2.393939 2.515152 2.5606062.3030302.272727
[9] 2.181818 2.363636 2.393939 2.151515 2.378788 2.5151522.4696972.045455
[17] 2.424242 2.212121 2.000000 2.000000 1.763636 2.2727272.0757582.151515
[25] 2.378788 1.863636 2.242424 2.651515 2.318182 2.2878791.9393942.045455
[33] 2.000000 2.621212 2.333333 1.854545 1.717949 2.0454552.1969702.060606
[41] 2.090909 2.515152 2.606061 1.954545 2.287879 2.0909092.2272732.200000
[49] 2.030303 1.854545 2.000000 2.181818 1.800000 2.1636362.1818182.242424
[57] 2.333333 2.757576 2.030303 2.257576 2.015152 2.3030302.4848482.106061
[65] 2.515152 1.954545 1.945455 2.075758 2.030303 2.1515152.0000002.227273
[73] 1.981818 2.018182 2.272727 2.075758 2.545455 2.3181822.0545451.954545
[81] 2.054545 2.363636 2.563636 2.136364 2.242424 2.5151522.0444442.212121
[89] 2.109091 2.090909 2.000000 2.378788 2.227273 2.2727272.0151522.106061
[97]2.0606062.4393941.9130432.106061
APPENDIX B
Tableofrandomlyconstructedclusteringmeasures [1] 0.53571429 0.53571429 0.53571429 0.55932203 0.519230770.54000000
[7] 0.51724138 0.49090909 0.48979592 0.48979592 0.470588240.40384615
[13] 0.33333333 0.52941176 0.55102041 0.40384615 0.510638300.48214286
[19] 0.33962264 0.38181818 0.46153846 0.37500000 0.319148940.34090909
[25] 0.39130435 0.30508475 0.48000000 0.40540541 0.428571430.19148936
[31] 0.27272727 0.28571429 0.29411765 0.41666667 0.418604650.34426230
[37] 0.41666667 0.23076923 0.29411765 0.20000000 0.136363640.36585366
[43] 0.09677419 0.37500000 0.25531915 0.36000000 0.193548390.34285714
[49] 0.32812500 0.42105263 0.12765957 0.19354839 0.250000000.43750000
[55] 0.25000000 0.28571429 0.25000000 0.20689655 0.191489360.28571429
[61] 0.18000000 0.37500000 0.20000000 0.23076923 0.363636360.27777778
[67] 0.33333333 0.23529412 0.13043478 0.16666667 0.230769230.29268293
[73] 0.07692308 0.20930233 0.23076923 0.19565217 0.310344830.17142857
[79] 0.30769231 0.32142857 0.17647059 0.37500000 0.125000000.16216216
[85] 0.38181818 0.26470588 0.11111111 0.26086957 0.243243240.21428571
[91] 0.27272727 0.34883721 0.25000000 0.18750000 0.069767440.14634146
[97]0.068181820.100000000.307692310.21951220
L.VederhusisanAssociateProfessoratUiTTheArcticUniversityofNorway,Tromsø. Herprofessional and researchinterestsarewide-ranging,originatinginpsychology,education,sociology,biology,andinformation science, and into complexity theory and systems sciences. She has also been teaching quantitative and qualitativeresearchmethodsandtheoryofscienceformanyyears.Recentresearchfocusesonleadership, safety,communication,learning,teaching,trainingandassessmentinorganizations(NorwegianUniversityof ScienceandTechnology,Norway;UniversityofStavanger,Norway;UiTTheArcticUniversityofNorway,Norway.
Email:lillian.vederhus@uit.no).
