管理學(xué)課件第六章制定決策管理者工作的本質(zhì)

管理學(xué)課件第六章制定決策管理者工作的本質(zhì)DecisionMakingDecision(定義)Makingachoicefromtwoormorealternatives.TheDecision-MakingProcess(決策過程)Identifyingaproblemanddecisioncriteriaandallocatingweightstothecriteria.Developing,analyzing,andselectinganalternativethatcanresolvetheproblem.Implementingtheselectedalternative.Evaluatingthedecision’seffectiveness.2Exhibit6–1

TheDecision-MakingProcess3Step1:IdentifyingtheProblemProblemAdiscrepancybetweenanexistinganddesiredstateofaffairs.CharacteristicsofProblemsAproblembecomesaproblemwhenamanagerbecomesawareofit.Thereispressuretosolvetheproblem.Themanagermusthavetheauthority,information,orresourcesneededtosolvetheproblem.4Step2:IdentifyingDecisionCriteriaDecisioncriteriaarefactorsthatareimportant(relevant)toresolvingtheproblem.Coststhatwillbeincurred(investmentsrequired)Riskslikelytobeencountered(chanceoffailure)Outcomesthataredesired(growthofthefirm)Step3:AllocatingWeightstotheCriteriaDecisioncriteriaarenotofequalimportance:Assigningaweighttoeachitemplacestheitemsinthecorrectpriorityorderoftheirimportanceinthedecisionmakingprocess.5Exhibit6–2 CriteriaandWeightsforComputerReplacementDecisionCriterion WeightMemoryandStorage 10Batterylife 8CarryingWeight 6Warranty 4DisplayQuality 36Step4:DevelopingAlternativesIdentifyingviablealternativesAlternativesarelisted(withoutevaluation)thatcanresolvetheproblem.Step5:AnalyzingAlternativesAppraisingeachalternative’sstrengthsandweaknessesAnalternative’sappraisalisbasedonitsabilitytoresolvetheissuesidentifiedinsteps2and3.7Exhibit6–3 AssessedValuesofLaptopComputersUsingDecisionCriteria8Step6:SelectinganAlternativeChoosingthebestalternativeThealternativewiththehighesttotalweightischosen.Step7:ImplementingtheAlternativePuttingthechosenalternativeintoaction.Conveyingthedecisiontoandgainingcommitmentfromthosewhowillcarryoutthedecision.9Exhibit6–4 EvaluationofLaptopAlternativesAgainstWeightedCriteria10Step8:EvaluatingtheDecision’sEffectivenessThesoundnessofthedecisionisjudgedbyitsoutcomes.Howeffectivelywastheproblemresolvedbyoutcomesresultingfromthechosenalternatives?Iftheproblemwasnotresolved,whatwentwrong?11Exhibit6–5 DecisionsintheManagementFunctions12MakingDecisionsRationality（理性）Managersmakeconsistent,value-maximizingchoiceswithspecifiedconstraints.Assumptionsarethatdecisionmakers:Areperfectlyrational,fullyobjective,andlogical.Havecarefullydefinedtheproblemandidentifiedallviablealternatives.HaveaclearandspecificgoalWillselectthealternativethatmaximizesoutcomesintheorganization’sinterestsratherthanintheirpersonalinterests.13Exhibit6–6 AssumptionsofRationality14MakingDecisions(cont’d)BoundedRationality（有限理性）Managersmakedecisionsrationally,butarelimited(bounded)bytheirabilitytoprocessinformation.Assumptionsarethatdecisionmakers:WillnotseekoutorhaveknowledgeofallalternativesWillsatisfice—choosethefirstalternativeencounteredthatsatisfactorilysolvestheproblem—ratherthanmaximizetheoutcomeoftheirdecisionbyconsideringallalternativesandchoosingthebest.InfluenceondecisionmakingEscalationofcommitment

(承諾升級):anincreasedcommitmenttoapreviousdecisiondespiteevidencethatitmayhavebeenwrong.15TheRoleofIntuition（直覺的作用）Intuitivedecisionmaking（直覺決策）Makingdecisionsonthebasisofexperience,feelings,andaccumulatedjudgment.16Exhibit6–7 WhatisIntuition?Source:BasedonL.A.BurkeandM.K.Miller,“TakingtheMysteryOutofIntuitiveDecisionMaking,”AcademyofManagementExecutive,October1999,pp.91–99.17TypesofProblemsandDecisionsStructuredProblems（結(jié)構(gòu)良好的問題）Involvegoalsthatclear.Arefamiliar(haveoccurredbefore).Areeasilyandcompletelydefined—informationabouttheproblemisavailableandcomplete.ProgrammedDecision（程序化決策）Arepetitivedecisionthatcanbehandledbyaroutineapproach.18TypesofProgrammedDecisionsPolicyAgeneralguidelineformakingadecisionaboutastructuredproblem.ProcedureAseriesofinterrelatedstepsthatamanagercanusetorespond(applyingapolicy)toastructuredproblem.RuleAnexplicitstatementthatlimitswhatamanageroremployeecanorcannotdo.19Policy,Procedure,andRuleExlesPolicyAcceptallcustomer-returnedmerchandise.ProcedureFollowallstepsforcompletingmerchandisereturndocumentation.RulesManagersmustapproveallrefundsover$50.00.Nocreditpurchasesarerefundedforcash.20ProblemsandDecisions(cont’d)UnstructuredProblems（結(jié)構(gòu)不良的問題）Problemsthatareneworunusualandforwhichinformationisambiguousorincomplete.Problemsthatwillrequirecustom-madesolutions.NonprogrammedDecisions（非程序化決策）Decisionsthatareuniqueandnonrecurring.Decisionsthatgenerateuniqueresponses.21問題類型、決策類型和組織層次程序化決策非程序化決策組織層次頂層底層結(jié)構(gòu)良好的結(jié)構(gòu)不良的問題類型?PrenticeHall,20026-2222Exhibit6–8 ProgrammedversusNonprogrammedDecisions231、個體決策2、群體決策相對于個人決策，群體決策有一些優(yōu)點：(1)能更大范圍地匯總信息；(2)能擬訂更多的備選方案；(3)能得到更多的認(rèn)同；(4)能更好地溝通；(5)能作出更好的決策等。但群體決策也有一些缺點，如花費較多的時間（群體決策的效率較低）、產(chǎn)生“從眾現(xiàn)象”，以及責(zé)任不明等。適用范圍對于復(fù)雜、重要和需有關(guān)人員廣泛接受的決策問題，組織最好要采取群體的方式來制定決策。按決策者分類

在實踐中，群體決策往往轉(zhuǎn)換為一系列個體決策來求解。24Decision-MakingConditionsCertainty（確定性）Asituationinwhichamanagercanmakeanaccuratedecisionbecausetheoutcomeofeveryalternativechoiceisknown.Risk（風(fēng)險性）Asituationinwhichthemanagerisabletoestimatethelikelihood(probability)ofoutcomesthatresultfromthechoiceofparticularalternatives.Uncertainty（不確定性）25確定型、風(fēng)險型、非確定型決策確定型決策（Certainty）備選方案只存在一種自然狀態(tài)的決策。風(fēng)險型決策（Risk

）備選方案存在兩種或兩種以上自然狀態(tài)，每種自然狀態(tài)發(fā)生的概率可以估計的決策。非確定型決策（Uncertainty

）

備選方案存在兩種或兩種以上自然狀態(tài)，每種自然狀態(tài)發(fā)生的概率無法估計的決策。26確定性決策方法－－盈虧平衡分析盈虧平衡點：企業(yè)經(jīng)營活動處于不盈也不虧狀態(tài)的保本點。利潤=總收入–總成本其中：總收入=銷售價格(P)*產(chǎn)銷量(Q)總成本與產(chǎn)銷量有關(guān)總成本：包括固定成本（在一定期間內(nèi)，當(dāng)企業(yè)產(chǎn)銷量變化時其總額保持不變的成本，如機器的折舊等）和變動成本

（指隨產(chǎn)銷量的增加而同步增加的費用或成本。如直接人工費、原材料消耗等費用）27產(chǎn)銷量固定成本

變動成本總成本成本028例題的圖解

1045000產(chǎn)銷量（千臺）金額（萬元）13銷售收入總成本固定成本29總成本與單位產(chǎn)品成本利潤＝產(chǎn)銷量*單價–產(chǎn)銷量*單位變動成本–固定成本＝QP–QCV–F

＝Q(P–CV)–F式中P–CV為單位（產(chǎn)品）的貢獻毛收益，即產(chǎn)品銷售單價超過單位變動成本的部分

由QP–QCV–F＝0推導(dǎo)可得（此時盈虧相平）：

Q＝盈虧平衡點產(chǎn)銷量＝

30

[分析]

是否同意接受該外商的訂貸，要看降低了售價后是否還能給企業(yè)帶來利潤。

[可能的錯誤是]：內(nèi)銷產(chǎn)品單位貢獻毛收益為100-60=40元，外銷產(chǎn)品單位貢獻毛收益為75—50=25元，不合算。外銷產(chǎn)品的利潤為25*2-250=-200，更不合算。但是，實際上這家企業(yè)生產(chǎn)所投的固定成本已在內(nèi)銷產(chǎn)品中得到全額補償并有盈余70萬元，所以接受外商訂貨可使企業(yè)再凈賺利潤50萬元?？梢?，要是這家企業(yè)沒有其他更好的銷售機會，應(yīng)該作出接受外銷訂貨、增加產(chǎn)銷量的決策。[例]假設(shè)某電子器件廠的主產(chǎn)品的生產(chǎn)能力為10萬件，固定成本總額為250萬元，單位變動成本為60元。已有國內(nèi)訂單共8萬件，單價為100元。最近有一外商要求訂貨，但他出的單價僅為75元，訂量2萬件，并自己承擔(dān)運輸費用。由于這外銷的2萬件不需要企業(yè)支出推銷費和運輸費，這樣可使單位變動成本降至50元。該廠是否接受外商的訂貨呢？31風(fēng)險型決策方法－－決策樹法[例]某公司為投產(chǎn)某種新產(chǎn)品擬定兩個方案：一是建設(shè)規(guī)模較大的工廠，另一是建設(shè)規(guī)模比較小的工廠。假設(shè)兩者的使用期一樣，但建大廠需投資30萬元，建小廠只需投資20萬元。這種新產(chǎn)品未來的銷路有好壞兩種情況，它們出現(xiàn)的概率分別為0.7和0.3，相應(yīng)的損益值預(yù)測結(jié)果是：若采納建大廠方案，如果銷路好，則生產(chǎn)經(jīng)營這種新產(chǎn)品能帶來100萬元的收益，但如果遇到銷路差的情況，則要損失20萬元；若采納建小廠的方案，如果銷路好，經(jīng)營收益能達(dá)到40萬元，而如果銷路差，則只有30萬元的收益。試問哪一種方案更可取?32用決策樹的方法比較和評價不同方案的經(jīng)濟效果，需要進行以下幾個步驟的工作：

(1)根據(jù)決策備選方案的數(shù)目和對未來環(huán)境狀態(tài)的了解，繪出決策樹圖形。

(2)計算各個方案的期望收益值。首先計算方案各狀態(tài)枝的期望值（用方案在各種自然狀態(tài)下的損益值去分別乘以各自然狀態(tài)出現(xiàn)的概率）然后將各狀態(tài)枝的期望收益值累加，求出每個方案的期望收益值。33決策樹這是一種以樹形圖來輔助進行各方案期望收益的計算和比較的決策方法。決策樹的基本形狀如下圖所示。3412-30-206437銷路好P1=0.7銷路好P1=0.7銷路差P2=0.3銷路差P2=0.3100萬元-20萬元40萬元30萬元建大廠需投資30萬元，建小廠只需投資20萬元新產(chǎn)品未來銷路好壞出現(xiàn)的概率分別為0.7和0.3若建大廠，銷路好，100萬元的收益，銷路差，損失20萬元；若建小廠，銷路好，40萬元的收益，銷路差，30萬元的收益35Exhibit6–9 ExpectedValueforRevenuesfromtheAdditionofOneSkiLift（Robbins）

Expected

Expected × Probability = ValueofEach

Event Revenues AlternativeHeavysnowfall $850,000 0.3 = $255,000Normalsnowfall 725,000 0.5 = 362,500Lightsnowfall 350,000 0.2 = 70,000

$687,50036Decision-MakingConditionsUncertainty（不確定性決策方法）Limitedinformationpreventsestimationofoutcomeprobabilitiesforalternativesassociatedwiththeproblemandmayforcemanagerstorelyonintuition,hunches,and“gutfeelings”.Maximax（最大最大選擇）:theoptimisticmanager’schoicetomaximizethemaximumpayoffMaximin（最大最小選擇）:thepessimisticmanager’schoicetomaximizetheminimumpayoffMinimax（最小化其最大遺憾）:themanager’schoicetominimizemaximumregret.37Exhibit6–10 PayoffMatrix3815第3方案28第4方案相對收益最大值及選取的方案1428281418A41524152124A391818159A21114111413A1悲觀準(zhǔn)則（Y）樂觀準(zhǔn)則（X）B3B2B1

B企業(yè)可能的反應(yīng)A企業(yè)的策略A企業(yè)在對手三種不同反擊策略下的收益狀態(tài)及方案選擇39最大后悔值最小化準(zhǔn)則

考慮到?jīng)Q策者在選定某一方案并付諸實施后，如果在未來實際遇到的自然狀態(tài)并不與決策時的判斷相吻合，這就意味著當(dāng)初如果選取其他的方案反而會使企業(yè)得到更好的收益。

這種情況無形中表明，這次決策存在一種機會損失，它構(gòu)成了決策的“遺憾值”，或稱“后悔值”。這里，“后悔”的意思是：你選擇了一種方案，實際上就放棄了其他方案可能增加的收益。所以，決策者將為此而感到后悔?！白畲蠛蠡谥怠弊钚』瘺Q策準(zhǔn)則就是一種力求使每一種方案選擇的最大后悔值達(dá)到盡量小的決策方法。40最大后悔值最小化準(zhǔn)則習(xí)題分析

7A4

最大后悔值中的最小值及選取的決策方案282124相對收益最大值7076281418A4131300152124A3151061518159A21717711111413A128-B321-B224-B1最大后悔值B3B2B1

B企業(yè)的可能反應(yīng)A企業(yè)的策略41Decision-MakingStyles（決策方式）DimensionsofDecision-MakingStylesWaysofthinking（思維方式）－理性的與直覺的不同Rational,orderly,andconsistentIntuitive,creative,anduniqueToleranceforambiguity（模糊承受力）－一致型和某種順序的需要與同時處理許多不同想法的不同Lowtolerance:requireconsistencyandorderHightolerance:multiplethoughtssimultaneously42Decision-MakingStyles(cont’d)TypesofDecisionMakersDirective（命令型）－快速、有效率的、有邏輯的Useminimalinformationandconsiderfewalternatives.Analytic（分析型）－謹(jǐn)慎，具有適應(yīng)和處理新情況的能力Makecarefuldecisionsinuniquesituations.Conceptual（概念型）－能夠?qū)で蠼鉀Q問題的創(chuàng)造性方案Maintainabroadoutlookandconsidermanyalternativesinmakingdecisions.Behavioral（行為型）－尋找決策的接受Avoidconflictbyworkingwellwithothersandbeingreceptivetosuggestions.43Exhibit6–12 Decision-MakingMatrix44Exhibit6–13 CommonDecision-MakingErrorsandBiases45Decision-MakingBiasesandErrorsHeuristics（啟發(fā)法）Using“rulesofthumb”tosimplifydecisionmaking.OverconfidenceBias（自負(fù)）Holdingunrealisticallypositiveviewsofone’sselfandone’sperformance.ImmediateGratificationBias（即時滿足）Choosingalternativesthatofferimmediaterewardsandthattoavoidimmediatecosts.46Decision-MakingBiasesandErrors(cont’d)AnchoringEffect（錨定效應(yīng)）Fixatingoninitialinformationandignoringsubsequentinformation.SelectivePerceptionBias（選擇性認(rèn)知）Selectingorganizingandinterpretingeventsbasedonthedecisionmaker’sbiasedperceptions.ConfirmationBias（證實）Seekingoutinformationthatreaffirmspastchoicesanddiscountingcontradictoryinformation.47Decision-MakingBiasesandErrors(cont’d)FramingBias（取景效應(yīng)）Selectingandhighlightingcertainaspectsofasituationwhileignoringotheraspects.AvailabilityBias（可獲得性）Losingdecision-makingobjectivitybyfocusingonthemostrecentevents.RepresentationBias（典型性）Drawinganalogiesandseeingidenticalsituationswhennoneexist.RandomnessBias（隨機性）Creatingunfoundedmeaningoutofrandomevents.48Decision-MakingBiasesandErrors(cont’d)SunkCostsErrors（沉沒成本）Forgettingthatcurrentacti