© 2002 South-Western/Thomson Slide 1-1 Chapter 1 - Decision Making and Quantitative Modeling © 2002 South-Western/Thomson Learning™ Slides prepared by Jeff Heyl, Lincoln University
© 2002 South-Western/Thomson Slide Quantitative Business Modeling Definition of a ModelDefinition of a Model Benefits and Drawbacks of ModelingBenefits and Drawbacks of Modeling Types of ModelsTypes of Models Effective ModelersEffective Modelers
© 2002 South-Western/Thomson Slide The Modeling Process A Five-Step Modeling ProcessA Five-Step Modeling Process Step 1:Opportunity/ Problem RecognitionStep 1:Opportunity/ Problem Recognition Step 2:Model FormulationStep 2:Model Formulation Step 3:Data CollectionStep 3:Data Collection Step 4:Analysis of the ModelStep 4:Analysis of the Model Step 5:Implementation and Project ManagementStep 5:Implementation and Project Management
© 2002 South-Western/Thomson Slide Detailed Modeling Example Step 1:Opportunity/ Problem RecognitionStep 1:Opportunity/ Problem Recognition Step 2:Model FormulationStep 2:Model Formulation Step 3:Data CollectionStep 3:Data Collection Step 4:Analysis of the ModelStep 4:Analysis of the Model Step 5:Implementation and Project ManagementStep 5:Implementation and Project Management
© 2002 South-Western/Thomson Slide Software for Modeling 1.5The Structure of this Text Questions Experiential Exercises Modeling Exercises Case - Henry Ford Hospital I.Learning Curve Exercise II.Scoring Model Exercise
© 2002 South-Western/Thomson Slide 1-6 Influence Diagram for Car Purchase Invoice price Dealer profit Cost to insure Car price Comfort Acceleration Warranty Select car Fuel economy Maximize satisfaction Exhibit 1.1
© 2002 South-Western/Thomson Slide 1-7 Data Collected for Car Purchase Decision LegSpeed InsuranceRoomWarranty0 to 60 CarRatingMPG(Inches)(Months)Cost(Seconds) BMW Z3 1.9LBetter than $27, average average CorvetteAverage , PorscheAverage , Boxster Boxster Mustang GTMuch worse ,604n.a. convertible than average convertible than average FirebirdWorse than , convertible average convertible average Mercedes SLKAverage , Volvo C70Average , n.a. = not available Exhibit 1.2
© 2002 South-Western/Thomson Slide 1-8 Method for Scoring Cars on Each Criteria Criterion1 Point2 Points3 Points Cost> $36,000$28,000 – $36,000 $36,000$28,000 – $36,000< $28,000 Warranty36 months48 months Insurance ratingWorse than AverageBetter than average average average average Zero to 60 MPH> 8 seconds7 to 8 seconds 8 seconds7 to 8 seconds< 7 seconds Leg room< 42 inches42 – inches≥ 43 inches MPG< 20 MPG20 – MPG≥ 21 MPG Exhibit 1.3
© 2002 South-Western/Thomson Slide 1-9 Method for Scoring Cars on Each Criteria Criterion1 Point2 Points3 Points Cost> $36,000$28,000 – $36,000< $28,000 Warranty36 months48 months Insurance ratingWorse than AverageBetter than average Zero to 60 MPH> 8 seconds7 to 8 seconds< 7 seconds Leg room< 42 inches42 – inches≥ 43 inches MPG< 20 MPG20 – MPG≥ 21 MPG Exhibit 1.3 BMW Z3 Total Score =(3 x 15) + (2 x 10) + (1 x 15) + (3 x 25) + ( 2 x 15) =245
© 2002 South-Western/Thomson Slide 1-10 Quantitative Business Modeling Definition, qualities, and perceptions of models A simplified representation or abstraction of reality Important qualities – Validity, Usability, Value Positivists or relativists
© 2002 South-Western/Thomson Slide 1-11 Benefits of Modeling Models enable the compression of time Manipulation of a model is easier than manipulating the real system The costs of mistakes is much smaller Allows the consideration of risk Lower cost than experimenting with the real system
© 2002 South-Western/Thomson Slide 1-12 Benefits of Modeling Models enhance and reinforce learning Using the QBM process forces the use of rigorous thinking The use of mathematical models enables quick identification and analysis of a very large number of possible solutions Provides a better understanding of the real situation
© 2002 South-Western/Thomson Slide 1-13 Drawbacks of Modeling It is time consuming Managers might be reluctant to accept results Obtaining necessary data might be difficult, time consuming, expensive, or not even feasible It may be difficult to assess uncertainties
© 2002 South-Western/Thomson Slide 1-14 Drawbacks of Modeling An oversimplified model might lead to erroneous recommendations QBM can be expensive to undertake relative to the size of the problem Studies and may be abandoned or results ignored resulting in unproductive expense The common perception that if done on a computer, it must be correct
© 2002 South-Western/Thomson Slide 1-15 Types of Models Physical – least abstract, usually based on a different scale Analog – do not look like the real situation, but behave like it Hourglass, organizational charts, maps, stock market charts, graphs
© 2002 South-Western/Thomson Slide 1-16 Y = SUMPRODUCT(B$4:G$4,B7:G7) [copy to cells H8:H12] Car purchase decision Criteria: Ins. Rating MPG Leg Room Weights: Options Ins. Rating MPG BMW Z3 1.9L 3 A B C D Knobs x 1 = Number of orders per year x 2 = Quantity of safety stock Cost ($1,000) Dial Types of Models Mathematical – most abstract, yet easily manipulated for experimentation and prediction Independent variables – those we can control Uncontrollable parameters – governed by nature or outsiders Dependent variables – our measures of interest
© 2002 South-Western/Thomson Slide 1-17 Uses of Models Prescriptive – used to find the optimal solution within the assumptions of the model Enumeration Complete, exhaustive enumeration Algorithm
© 2002 South-Western/Thomson Slide 1-18 Uses of Models Descriptive – characterize things as they are Used to investigate outcomes or consequences Solutions not necessarily optimal Useful in predicting the behavior of systems under various conditions
© 2002 South-Western/Thomson Slide 1-19 Effective Modelers Internal skills: creativity, sensitivity to the client, and persistence Interpersonal skills: communication and teamwork Expertise in quantitative business modeling
© 2002 South-Western/Thomson Slide 1-20 A Five-Step Modeling Process Real-life opportunity or problem Step 1: Opportunity/ problem recognition Step 2: Model formulation Step 3: Data collection Step 4: Analysis of the model Step 5: Implementation and project management Validation Solution testing, verification Exhibit 1.8
© 2002 South-Western/Thomson Slide 1-21 The General Structure of a Model Uncontrollable factors (parameters) Mathematical relationships Independent (decision) variables Dependent variables Exhibit 1.9
© 2002 South-Western/Thomson Slide 1-22 Example of the Components Models Exhibit 1.10 DecisionDependentUncontrollable AreaVariablesVariablesVariables Financial Investment amountsTotal profitInflation rate investment Period of investmentRate of returnPrime rate investment Period of investmentRate of returnPrime rate Timing of investmentEarnings/shareCompetition Liquidity MarketingAdvertising budgetMarket shareDisposable income Number of modelsCustomer Competitor’s actions Zonal sales reps satisfaction ManufacturingProduction amountsTotal costMachine capacity Inventory levelsQuality levelTechnology Incentive planSpoilageMaterials prices AccountingAudit scheduleData processingLegal requirements Use of computers costTax rates Depreciation scheduleError rateComputer technology ServicesNumber of serversCustomer Demand for service satisfaction satisfaction
© 2002 South-Western/Thomson Slide 1-23 A Manufacturing Systems Model MachinesMethodsToolsLaborEnergy Dependent variables: Quantity Quality Profit Decision variables: What to produce When Who will work Where to stock Uncontrollable variables: Price of material Speed of machine Wages Legal requirements Environment Raw materials Finished products InputsProcessesOutputs Exhibit 1.11
© 2002 South-Western/Thomson Slide 1-24 A Simplified Model of a Manufacturing Situation Uncontrollable variables: 5, 2, and 50 (market prices, marketing limitation) Decision variables: x 1, x 2 (What quantities of products 1 and 2 should be produced?) Dependent variable: R = 5x 1 + 2x 2 (total revenue) Mathematical relationship: Maximize revenue Objective Constraint Subject to: x 1 + x 2 50 x 1 + x 2 50 Exhibit 1.12
© 2002 South-Western/Thomson Slide 1-25 Problem Classification Allocation situations There are a number of activities to be performed There are multiple ways to perform these activities Resources or facilities are limited Decision situations Waiting-line situations Predicting the behavior of a system
© 2002 South-Western/Thomson Slide 1-26 Decision-Making Categories IgnoranceUncertaintyRiskCertainty Increasing knowledge Exhibit 1.13
© 2002 South-Western/Thomson Slide 1-27 Analysis of the Model – Selecting an Alternative Generate Alternatives Predict the Outcome of Each Alternative Relate Outcomes to Goals Compare the Alternatives Select an Alternative
© 2002 South-Western/Thomson Slide 1-28 Audio Equipment Assembly Line Adopt video- based technology Learning rate Process technology Cross- training Task tenure Product design Worker forgetting Worker productivity Increase capacity of line Exhibit 1.15
© 2002 South-Western/Thomson Slide 1-29 Apartment Rental Problem Cost View Parking Size Apartment selection New apartment Shopping Entertainment
© 2002 South-Western/Thomson Slide 1-30 Apartment Rental Problem ShoppingEntertainment LocationCostSizeViewParkingConvenienceConvenience ABCDE
© 2002 South-Western/Thomson Slide 1-31 Apartment Rental Problem ShoppingEntertainment LocationCostSizeViewParkingConvenienceConvenience A$1,450650ExcellentAttended Average Good lot lot B900850AverageSelf-park Excellent Good lot lot C725725PoorOn-streetGood Excellent D875905Average Own GoodGood garage garage E1,130795GoodAttended PoorGood garage garage
© 2002 South-Western/Thomson Slide 1-32 Apartment Rental Problem Criterion [weight]1 point2 points3 points4 points Cost [15]> $1,200$1,000 - $1,200$800 - $1,000 $1,200$1,000 - $1,200$800 - $1,000< $800 Size (sq ft) [10] 1,000 View [10]PoorAverageGoodExcellent Parking [20]On-streetAttendedOwnAttended or self- lot garage garage or self- lot garage garage park lot park lot Shopping [10]PoorAverageGoodExcellent convenience convenience Entertainment [15]PoorAverageGoodExcellent convenience convenience
© 2002 South-Western/Thomson Slide 1-33 Apartment Rental Problem