Evaluating Alternatives Chapters 3

1
Evaluating AlternativesChapters 3 4

• Dr. Dale Masel
• April 13, 2002

2
Selecting a Project

• Need to meet objectives
• Need to work with available resources
• Need to stay on budget
• Need to meet schedule
• Need to balance all of the needs to choose the
  best all-around option

3
Risk Analysis

• Risk Something happens differently than was
  expected
• In project
• Dont meet schedule, budget, etc.
• In product
• Doesnt work like it was supposed to
• Size of risk depends on
• Cost if it happens
• Probability it will happen

4
Risk Analysis

• Before project starts
• What can go wrong?
• How likely is it?
• What can be done about it?
• Important to consider risks before each plan is
  selected
•  Some sources of risk
• People
• Equipment
• Act of God

5
Risk Analysis

• Steps
• Identify possible problems with each task
• Identify effect of problem on other tasks
• Develop plans to
• minimize effect of risk
• eliminate risk
•  Balance risk and cost
• Too expensive to eliminate all riskHow much is
  OK?
• Try to make risks less likely to occur

6
Risk-Benefit Analysis

• Formulation
• Identify risk
• Determine costs of reducing risk
• Analysis
• Probability of risk
• Effect of occurring
• Interpretation
• Evaluate results
• Make decision

7
Decision Trees

• Apply conditional probability to compare
  alternatives
• Decision node Can control result
• Chance node Random result

8
Decision Trees

• Branches on tree represent all possible outcomes
• Lead to all possible outcomes
• Determine expected value of all decisions
• Identify course of action with highest expected
  value

9
Developing a Decision Tree

• Main decision node at left
• Branches from node for all outcomes
• Determine chance of each random branch
• Continue until all branches have payoff
• Node rules
• Branches must be mutually exclusive
• Branches must be exhaustive

10
Solving a Decision Tree

• Start at right side of tree
• Determine expected monetary value (EMV) for
  chance nodes
• Determine EMV of decision nodes
• Work back to left side of tree

11
Decision Trees

• Advantages
• See relationships among alternatives
• Identify risks/weaknesses
• Disadvantages
• Significant amount of data collection
• Requires estimation of probabilities

12
Taking it all in
13
Reliability of your decision
Data Reliability
Model Accuracy
Randomness
14
Multicriteria Analysis

• Compare alternatives in several different areas
• Steps
• Define objectives
• List attributes

15
Multicriteria Analysis

• Complete cover all factors
• Important represent significant factors
• Measurable with utility function if necessary
• Familiar able to identify preferences
• Nonredundant each measure different thing
• Independent changing value of attribute
  shouldnt change its importance

16
Analytic Hierarchy Process

• Structured way to compare alternatives on many
  different characteristics
• Determine local priorities
• Relative importance of attributes
• Done by comparing all pairs of factors
• aij relative importance of attribute i vs.
  attribute j

17
Analytic Hierarchy Process

• Restrictions
• Can make hierarchy for problem
• All factors included in hierarchy
• aij 1/aji Reciprocity
• aij ? 8 Nothing is infinitely better

18
Analytic Hierarchy Process

• Sequence attributes from most important to least
  important
• Compare each pair
• Rank on scale of 1 to 9
• 1 equal importance
• 9 unquestionably more important

19
Group Decision Making

• Goals
• Balance opinions of group members
• Build consensus among group members
• Can be live or by correspondence
• Beware of
• Experts
• Power

20
Group Attitudes

• Uncooperative
• May be hurt by result
• Steadfast
• Dont want to compromise and change opinion
• Compromise
• Have opinions combined
• Build
• Work together to make decisions by consensus

21
Group Decision Support Systems

• Remove communications barriers
• Structure decision analysis
• Guide discussion
• Examples
• Electronic Messaging
• Electronic Bulletin Board
• Teleconference
• Document Sharing
• Agenda Setting
• Decision Modeling (e.g., trees, MAUT, AHP)