This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Chapter 4 Decision Making Chapter • • • • • Nature of decision making Management science Tools for decision making Computer based methods Qualitative methods Advanced Organizer
Managing Engineering and Technology
Management Functions Planning Decision Making Organizing Leading Controlling Managing Technology Research Design Production Quality Marketing Project Management Personal Technology Time Management Ethics Career Relation to Planning Relation
Managerial decisionmaking is the Managerial process of making a conscious choice between two or more rational alternatives Weighing the cost/benefits of viable Weighing alternatives and picking the correct one Decision making is an essential part of Decision the planning process Occasions for Decisions Occasions
• Required by superiors • Delivered by subordinates • Self initiated Q: what is the risk of selfinitiated decisions? Types of Decisions Types
• Routine • Structured • Clear data • Clear criteria • Specific alternatives • Computerize NonRoutine • Unstructured • Unclear data • Subjective • No clear set of alternatives • Intuition In the end, all decisions are subjective Rational Decisions Rational
• Optimizing (maximizing) the outcome by choosing the best from the choices • We like to think we make rational decisions but there are 3 problems. Problems w/ Rational Decisions Problems
1. Rationality requires complete understanding of the consequences. 2. Since reality is in the future we must imagine it. 3. Only a limited number of the possible alternatives are known. Reality of Decision Making Reality
• We act within bounded bounded rationality • Accepting a choice that is “good enough” • This is called satisficing. satisficing Management Science Management Characteristics
• Systems view of the problem • Significant interrelated variables • Team approach • Mixed backgrounds • Emphasis on use of formal mathematical models and statistical and quantitative techniques Modeling Modeling
• Model large systems (systems engineering) • Abstraction or simplification of reality • Contains only the essential features for a particular analysis • Most are mathematical • Simple to complex set of equations Management Science Management Process (5 steps)
Real World
1. Formulate Problem (Objectives, variables, constraints) 2. Model World
Construct simple yet realistic mathematical model Test model – past to present – revise as needed Derive a solution from the model 3. 5. Apply model to real system  document effectiveness  revise as required 4. Categories of Decision Categories Making Tools
• Certainty • Linear Programming, Computer Solutions • Risk • Expected Value, Decision Tree, Variance • Uncertainty • Maximin, Maximax, Hurwicz, Insufficient Reason Payoff Table Payoff
Alternative State of Nature/Probability N1 p1 A1 A2 … Ai … Am O11 O21 … Oi1 … Om1 N2 p2 O11 O22 … Oi2 … Om2 … … … … … … … … Nj pj O1j O2j … Oij … Omj … … … … … … … … Nn pn O1n O2n … Oin … Omn A : alternatives N: future state of nature P: probability of state of nature occurring O: outcome (payoff, benefit gained) Decision Making Decision Under Certainty
• Certain of the future state of nature (at least assume future state is certain)
• Choose alternative that gives most favorable outcome • Linear Programming
• Mathematical function to maximize benefit (or minimize cost) subject to constraints • Used to determine optimal allocation of an organization’s limited resources • Computer Solutions
• Many variable analysis for Linear Programming Linear Programming Linear
• State the problem • Decision Variables • Objective function • Constraints Decision Making Decision Under Risk
• A number of possible future states exist and there may not be one that results in best outcome • Expected Value
• • Sum of the probabilities of each outcome Probability plays the role of a substitute for certainty • Decision Trees similar (branch representation) • Queuing
• Identifying optimum number of people needed to reduce overall cost Decision Making Under Risk: Decision computer simulation
• Utilizes probability distributions to represent stochastic activities • Specific instances of variables are selected from the distribution based on random variables • The model is run many times Decision Making Under Decision Risk: Risk as Variance Risk
• Projects have the same mean, therefore, need to look at the expected values of the variances • Greater the variability of the expected value, the greater the risk Expected Values for X and Y are equal. V(X) < V(Y) Project Y has more risk, so choose Project X Decision Making Decision Under Uncertainty
• Several future states of nature exist but probability is not known • Maximax • Optimistic • Offers the highest possible outcome • Maximin • Pessimistic • Alternative with the “least bad” outcome Decision Making Decision Under Uncertainty
• Hurwicz approach • Between the Maximax and Maximin • Principle of insufficient reason Pj=1/n • All states same chance so set probabilities as same for each • Minimax regret • Alternative with the smallest difference between “best” and “worst” Uncertainty (Drill Example) Uncertainty
• Maximax
• • • • • • Best outcome A2 ($9,300,000) Best worse outcome A3 ($0) – loose nothing but try because could gain Coefficient of optimism @0.2 A2: 0.2*(9,300,000) + 0.8*(500,000) A3: 0.2*(1,250,000) + 0.8*(0) Expected value with probabilities equal • Maximin • Hurwicz • Equally likely • Minimax
• • • • Look at what you could lose if you choose not to do it A1: loose the max you could gain A2: loose the drilling cost if the well is dry A3: loose the difference of striking the well but sharing the wealth ComputerBased Computer Information Systems
• Integrated Data Bases
• Information sharing • Management Information Systems
• Structured problem • Decision Support Systems
• Less structured • Expert Systems
• Artificial intelligence with experts Discrete Event Simulation Discrete
• Computer modeling of discrete events Example Example
We want to build a simulation model of a drill press in the shop. The press is used at random, depending what jobs people are doing. We will model each shop worker as a customer of the equipment. Assume customers enter the drill press area every 6 minutes and their job takes 6 minutes. Then life is fine, assuming no down time occurs for the press. Service Schedule of Drill Press Service
Cust 1 2 3 4 5 6 7 8 9 10 11 12 time between arrivals 6 6 7 1 3 service time 3 1 6 2 5 7 arrival time 0:00 00:06 00:12 00:19 00:20 00:23 service begin 0:00 00:06 00:12 00:19 00:21 00:26 service end 00:03 00:07 00:18 00:21 00:26 00:33 mach. idle time 0 3 5 1 0 0 worker waiting time 0 0 0 0 1 3 Examples of DES Examples
• ProModel • ServiceModel Qualitative Approaches to Qualitative Decision Making
• So far we have addressed the classical, analytical techniques • Seven new and seven old tools Affinity diagrams Affinity
The use of an affinity diagram is to provide a structured way to organize textual data. Taking an intuitive approach is ok for simple projects, but most benefit from the use of an affinity diagram. Process for an affinity Process diagram
• Write the terms onto cards or Postit Notes • Group the terms into clusters that seem to go together  no talking • If fights break out, make a duplicate card for another pile • After sorting, begin discussion. Make adjustments. Combine redundant terms. Process for an affinity Process diagram
• Operationally define each cluster. Terms that are more abstract than the others often make good descriptors. • Watch out for “world hunger” and “peace on earth” cards, they tell you nothing. ie a good product, best in its class, second to none Process for an affinity Process diagram
• Now arrange the clusters into groups. Operationally define those. • You now have a primary, secondary, and tertiary description of terms. Implementation Implementation
Decisions can be the easy part, after that is implementing and sustaining (revisiting, modification, etc…)
Homework: 41,2,4,9,10,11 ...
View
Full
Document
This note was uploaded on 04/13/2010 for the course EM 510 taught by Professor Stauffer,l during the Spring '08 term at Idaho.
 Spring '08
 Stauffer,L
 Engineering Management

Click to edit the document details