Quantitative and Qualitative Decision Making
yenomrah
AD715OL_LectureNotes_L5L6.pdf
This is a single, concatenated file, suitable for printing or saving as a PDF for offline viewing. Please
note that some animations or images may not work.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
1 of 40 6/18/17, 11:26 AM
Week 3 Study Guide and Deliverables
Readings: Modules 5 and 6 online content
e-Reserve: Render, Ch. 1: Introduction to Quantitative Analysis (Module 5)
Render, Ch. 3: Decision Analysis (Module 6)
Discussions: Discussion 3: Quantitative Analysis and Managerial Decisions in an
Organization
With the help of one or several of the recommended tutorials for Week
3, discuss your experience and plans for applying analytical methods in
your Assignment 2 or in your current (or targeted) profession.
Exercises: Tutorial 02: Decision Trees in TreePlan
Tutorial 03: Scenario Analysis with Scenario Manager in Excel (Optional)
Tutorial 04: Introduction to Data Modeling and Visualization in Excel 2013
Power Pivot (Optional)
Tutorial 05: Risk & Sensitivity Diagrams in Excel Power Pivot (Optional)
Individual Exercise 3-1: Working with Tutorial 02 for AD715: Decision Trees
in TreePlan
In preparation for Assignment 2, (i) review the Tutorial for AD715: Decision
Trees in TreePlan, (ii) go to the V-Labs, open Excel 2013 Add-In TreePlan,
and repeat the steps as discussed in the tutorial, (iii) read case problem "New
Electric Razor" (Assignment 2) and complete Task 2.2 (without TreePlan) and
Task 2.4 (with TreePlan).
Individual Exercise 3-2 (optional): If you are already working extensively
with Excel spreadsheets, you may find the following OPTIONAL tutorials
useful for your professional development:
Tutorial 03 Scenario Manager in Excel;i.
Tutorial 04 Data Modeling and Visualization in Excel Power Pivot;ii.
Tutorial 05 Risk & Sensitivity Diagrams in Excel Power Pivotiii.
Assessments: Quiz 3 (covering Modules 5 & 6) opens on Saturday at 9:00 AM ET, closes on
Monday at 11:59 PM ET
Live
Classrooms:
Preparation W3 – Tuesday (Day 1) at 8:30-9:30 PM ET (will be recorded)
Module 5 – Introduction to Quantitative Analysis and Decision Making
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
2 of 40 6/18/17, 11:26 AM
After you complete this lecture, you will be able to:
Describe the quantitative analysis approach
Describe the differences between business analysis, data analytics, and business analytics
Understand the application of quantitative analysis in a real situation
Discuss possible problems in using quantitative analysis
Understand the process of managerial decision-making explained with the help of a business simulation
and a business running case
One of the main objectives of managers is to define business needs and to determine solutions to business
problems. To gain a better understanding of a complex business problem, we are using the formal concept of
breaking it into smaller parts. This formal concept or process is widely known as analysis and has been applied
for centuries in the study of different disciplines such as philosophy, mathematics, physics, engineering, music,
and many more. In management, the analysis is an aid to managerial decision making process, and is defined as
quantitative and qualitative.
Render (2015) notes that Quantitative analysis is a scientific approach to managerial decision making in which
raw data are processed and manipulated to produce meaningful information (p. 2).
In solving a problem, decision makers must considered both quantitative and qualitative factors. Typical
examples for quantitative and qualitative factors in managerial decision makings can be provided, as follows:
Quantitative factors are data that can be accurately calculated
Different investment alternatives
Interest rates
Inventory levels
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
3 of 40 6/18/17, 11:26 AM
Demand
Labor cost
Qualitative factors are more difficult to quantify but affect the decision process
The weather
State and federal legislation
Technological breakthroughs
The fact that the quantitative analysis is dealing with data is making it important in all areas of management.
One of the widely used data-driven approach is the business analysis, defined as ‘the practice of enabling
change in an organizational context by defining needs and recommending solutions that deliver value to
stakeholders’ (BABOK).
To learn more about the business analysis profession, visit the website of the leading association International
Institute of Business Analysis (IIBA) at www.iiba.org and review the pages:
About IIBA (over 28,000 individual members, and over 220 corporate members)
Chapters (over 110 Chapters, including over 33 in the United States)
Certification & Recognition
Professional certification is a designation earned by an individual identifying that they have
demonstrated a standard level of skills, experience, and expertise within their field. Both the
Certified Business Analysis Professional™ (CBAP®) and Certification of Competency in Business
Analysis™ (CCBA®) designations are professional certifications
Recognition Programs provide an opportunity for IIBA to partner with academic and professional
learning institutions to ensure that courses deliver the appropriate level of business analysis
knowledge and develop the proper skills to effectively prepare students to work in the business
analyst role. The Academic Certificate Program and the Academic Diploma Program, are IIBA
Recognition programs that offer varying levels of knowledge and skill development to meet the
needs of students and experienced professionals
The diagram below summarizes the results of a software application, created by Enfocus Solutions for helping
companies to maximize value through business analysis:
The business analysis is described in the lower part of the diagram as a six steps approach – situation
analysis, ideation to features, features to requirements, requirements to implementation, transition, and
benefits realization
The upper section of the diagram presents the evolution of the current performance over time for each
one of the steps of the business analysis
Maximize Value through Business Analysis
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
4 of 40 6/18/17, 11:26 AM
http://enfocussolutions.com
Another emerging data-driven discipline is the big data analytics, defined as the process of generating insight
from data, where big data is characterized as high-volume, high velocity, high-variety (including unstructured)
and/or high-veracity information assets.
Big is relative to corporation’s capabilities—by growing its ability, the challenges of big data diminish
Information is collected in new ways using new technologies
The IT (information technology) perspective on big data analytics is challenging corporations to wrestle the big
data into a warehouse by collecting, treating, and sorting high-volume, high-velocity, high-variety, and high
veracity data. Possible solutions to this challenge are to improve the operational efficiency for handling the data
by expanding in a new way hardware and software capabilities.
The business perspective on big data analytics is challenging managers to handle the explosion of information
extracted from the data and to adapt their business model to facilitate better analytics-based decisions.
Data analytics represents a wide spectrums of forms of data-driven insight, including
data manipulation
reporting and business intelligence
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
5 of 40 6/18/17, 11:26 AM
data mining, optimization, and forecasting
Data analytics incorporate a wide variety of techniques (such as manual analysis, reporting, predictive
models, time series models, or optimization models) that help understand
what happened
what will happen
why it happened
what is the best one could possibly do
Companies and other organizations and policy makers need to address considerable challenges if they are to
capture the full potential of big data. A shortage of the analytical and managerial talent necessary to make the
most of big data is a significant and pressing challenge and one that companies and policy makers can begin to
address in the near term.
By 2018, the United States alone faces
a shortage of 140,000 to 190,000 people with deep analytical skills
a shortage of 1.5 million managers and analysts to analyze big data and make decisions based on their
findings.
The shortage of talent is just the beginning. Other challenges include:
The need to ensure that the right infrastructure is in place;
The need to ensure that the incentives and competition are in place to encourage continued innovation;
The economic benefits to users, organizations, and the economy are properly understood;
Safeguards are in place to address public concerns about big data (McKinsey, 2011).
Individual Exercises Read the article ‘Data Scientist: The Sexiest Job of the 21st Century’, by T. H. Davenport and
D. J. Patil, in Harvard Business Review, October 2012, also downloadable from:
http://www.tias.edu/docs/default-source/Kennisartikelen/harvard_data-scientist-the-sexiest-
job-of-the-21st-century_2012.pdf?sfvrsn=0
1.
Review the article ‘Salary Trends for Data Science Professionals’ by V. Granville on March 4,
2015 in http://www.datasciencecentral.com/profiles/blogs/salary-trends-for-data-science-
professionals?utm_content=buffer80207&utm_medium=social&utm_source=twitter.com&
utm_campaign=buffer
2.
If you are interested in becoming Certified Analytics Professional (CAPR) please visit:
https://www.informs.org/Certification-Continuing-Ed/Analytics-Certification
3.
Another emerging data-driven business discipline is the business analytics, defined as ‘the process of
leveraging all forms of analytics to achieve business outcomes through requiring business relevancy, actionable
insight, and performance management and value measurement’ (Stubbs, 2014, p. 206).
Business Analytics is the use of data-driven insight to generate value
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
6 of 40 6/18/17, 11:26 AM
Data analytics without business analytics creates no value – it simply answers questions
The business outcomes are typically tangible and/or intangible value of the organization
The real value of business analytics comes through balancing the external value and the internal value of
the organization
Examples for external value: economic returns, drives growth, delivers outcomes, improves quality
Examples for internal value: productivity returns, drives scale, enables outcomes, reduces costs
(Stubbs, 2011, p. 99–143 )
Business analytics is often broken into three categories:
Descriptive analytics – the study and consolidation of historical data
Predictive analytics – forecasting future outcomes based on patterns in the past data
Prescriptive analytics – the use of optimization methods
The diagram below summarized important questions, enablers, and outcomes for each one of the categories.
Sharda, ‘Business Intelligence and Analytics: Systems for Decision Support’, 10e 2015 Pearson
To address the shortage of the analytical and managerial talent necessary to make the most of big data and
business analytics, a number of universities are already offering programs in data science/business analytics. As
part of BU MET’s strategic commitment to offer programs that provide our students with knowledge and skills in
the areas of highest demand by the industry, we developed a new graduate certificate program in Applied
Business Analytics (scheduled for launch in Fall 2015).
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
7 of 40 6/18/17, 11:26 AM
The diagram bellow is presenting the four newly offered graduate courses in Applied Business Analytics. It is
important to stress that
the course AD 571 ‘Business Analytics Foundations’ is a prerequisite for each one of the remaining
courses (‘Enterprise Risk Analytics’, Marketing Analytics’, and ‘Web Analytics in Business’)
prior to the enrolment in the new program, students are asked to take the preparatory, non-credit online
course AD100 ‘Pre-Analytics Laboratory’ (entry fee $75), graded on a pass/fail basis (students are
allowed to take the test per unit until they pass it). For additional information visit http://www.bu.edu
/met/courses/graduate/data-analytics/pre-analytics-laboratory-ad-100/
The scientific quantitative analysis approach consists of seven linked to each other’s steps, described below as
follows (Render, 2015).
The first step is to develop a clear and concise statement of the problem to provide direction and meaning
This may be the most important and difficult step
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
8 of 40 6/18/17, 11:26 AM
Go beyond symptoms and identify true causes
Concentrate on only a few of the problems – selecting the right problems is very important
Specific and measurable objectives may have to be developed
Models are realistic, solvable, and understandable (i) mathematical representations of a situation, or (ii) physical
models of a project, or (iii) scale (design engineering) models, or (iv) schematic (drawings, charts, and/or
pictures) models, or any other representations of objects—see below.
The models that are used in quantitative analysis are based on mathematical relationships (which in most of the
cases are expressed in equations and inequalities). These models generally contain variables and parameters
Variable is a measurable quantity that may vary or is subject to change
It can be controllable or uncontrollable
Controllable variable is also called decision variable, and is generally unknown (e.g. how many
items should be ordered for inventory?)
Parameter is a measurable quantity that is inherent in the problem
Parameters are known quantities that are a part of the model (e.g. what is the cost of placing an
order?)
Models should be solvable, realistic, and easy to understand and modify
The required input data should be obtainable (input data is the available amount of data to be used
in the model).
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
9 of 40 6/18/17, 11:26 AM
Obtaining accurate input data for the model is essential:
Improper data will result in misleading results—see the GIGO diagram, below.
Data may come from a variety of sources—company reports, documents, employee interviews, direct
measurement, or statistical sampling
Developing a solution involves manipulating the model to arrive at the best (optimal) solution.
Common techniques are:
Solving equations (to be solved for the best decision)
Trial and error– trying various approaches and picking the best result
Complete enumeration– trying all possible values
Using analgorithm – a series of repeating steps to reach a solution
Both input data and the model should be tested for accuracy before analysis and implementation
Testing the data:
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
10 of 40 6/18/17, 11:26 AM
New data can be collected from different sources
These additional data can be compared with the original data
Statistical test can be employed to determine whether there are differences
If there are significant differences, more effort is required to obtain accurate input data
Testing the model:
The model may not be appropriate
The model can be checked to make sure that it is logical, consistent, and represent the real
situation
Analyzing the results starts with determining the implications of the solution
Implementing results often requires change in an organization
The impact of actions or changes needs to be studied and understood before implementation
A model is only an approximation of reality, therefore determining the sensitivity of the solution to changes in the
model and input data is important for the analysis of the results.
Basic types of analyzing the results include (short list, expect more in Lectures 7 to 12):
What-If-Analysis
In what-If analysis, a decision maker:
makes changes to variables or relationships among variables, and
observes the resulting changes in the variables of other variables.
The common strategy for the application of the what-if analysis in businesses is to make changes
to the input variables and to observe the resulting changes in the output parameters of the financial
model of a business (or project).
Sensitivity Analysis
Sensitivity analysis determines how much the results will change if the model or input data changes
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
11 of 40 6/18/17, 11:26 AM
If the model or the input data are wrong, the solution could be wrong
Sensitive models should be very thoroughly tested
Goal-Seeking Analysis
Ingoal-seeking analysis (how-can analysis), the decision maker sets a target value (goal) for a variable and
then repeatedly changes other variables until the target value is achieved.
This form of analytical modeling would help answer the question 'How we can achieve $X,XXX,XXX in net profit
after taxes' by repeatedly change of the values of prices and/or variable costs and/or quantities sold and/or fixed
costs in a financial model until the targeted result is achieved.
Optimization Analysis
Optimization analysis is a more complex extension of goal-seeking analysis.
The goal here is to find the optimum value for one or more target variables, given certain constrain.
To make this possible, one or several other variables are changed repeatedly, subject to the specified
constraints, until the decision maker discovers the best values for the target variables.
Implementation incorporates the solution into the company:
Implementation can be very difficult
People may be resistant to changes
Many quantitative analysis efforts have failed because a good, workable solution was not properly
implemented
Changes occur over time, so even successful implementations must be monitored to determine if modifications
are necessary.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
12 of 40 6/18/17, 11:26 AM
Quantitative analysis models are used extensively by real organizations to solve real problems.
In the real world, quantitative analysis models can be complex, expensive, and difficult to sell
Following the steps in the process is an important component of success
Individual Exercise Review the Case ‘Railroad Uses Optimization Models to Save Millions’
[Textbook, page 7]
Advantages of Mathematical Modeling
Models can accurately represent reality1.
Models can help a decision maker formulate problems2.
Models can give us insight and information3.
Models can save time and money in decision making and problem solving4.
A model may be the only way to solve large or complex problems in a timely fashion5.
A model can be used to communicate problems and solutions to others6.
Models Categorized by Risk
Mathematical models that do not involve risk or chance are called deterministic models
Deterministic means with complete certainty
All of the values used in the model are known with complete certainty
Mathematical models that involve risk or chance are called probabilistic models
Values used in the model are estimates based on probabilities
Note regarding deterministic and probabilistic models: Expect more in Lecture 6
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
13 of 40 6/18/17, 11:26 AM
Example: Pritchett’s Precious Time Pieces (Textbook, pp. 8 – 9)
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
14 of 40 6/18/17, 11:26 AM
BEP for Pritchett’s Precious Time Pieces
BEP = $1,000/($8 – $3) = 200 units
Sales of less than 200 units of rebuilt springs will result in a loss1.
Sales of over 200 units of rebuilt springs will result in a profit2.
Defining the problem
Problems may not be easily identified
Conflicting viewpoints
Impact on other departments
Beginning assumptions
Solution outdated
Developing a model
Fitting the textbook models
Understanding the model
Acquiring accurate input data
Using accounting data
Validity of the data
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
15 of 40 6/18/17, 11:26 AM
Developing a solution
Hard-to-understand mathematics
Only one answer is limiting
Testing the solution
Solutions not always intuitively obvious
Analyzing the results
How will it affect the total organization
Lack of commitment and resistance to change
Fear formal analysis processes will reduce management’s decision-making power
Fear previous intuitive decisions exposed as inadequate
Uncomfortable with new thinking patterns
Action-oriented managers may want “quick and dirty” techniques
Management support and user involvement are important
Lack of commitment by quantitative analysts
Analysts should be involved with the problem and care about the solution
Analysts should work with users and take their feelings into account
Let's first discuss what strategies are and how they vary across different levels of an organization.
Strategy is:
a plan, a "how," a means of getting from here to there1.
a pattern in actions over time; for example, a company that regularly markets very expensive products is
using a "high end" strategy
2.
a position; that is, it reflects decisions to offer particular products or services in particular markets3.
a perspective, that is, vision and direction (Mintzberg, 1994).4.
Corporate strategy: Determines the businesses in which a company will compete
Business strategy: Defines the basis of competition for a given business level
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
16 of 40 6/18/17, 11:26 AM
Functional strategy
Is used to back up the corporate and the business strategies;
Is customized to a specific functional area of the company, such as marketing, research and development,
manufacturing and operations, human resources, finances.
Competitive strategy is about being different:
It means deliberately choosing a different set of activities to deliver a unique mix of value;
It is a combination of the ends (goals) for which the firm is striving and the means (policies) by which it is
seeking to get there (Porter, 1996, pp. 39–74).
In summary, a well-developed strategy contains five components, or sets of issues:
Strategic scope: the number and types of industries, product lines, and market segments the company
competes in or plans to enter.
1.
Goals and objectives: desired levels of accomplishments on one or more dimensions of performance,
such as volume growth, profit contribution, or return on investment – over specified time periods for each
of those businesses and product-markets and for the organization as a whole.
2.
Resource deployments: deciding how financial and human resources are to be obtained and allocated,
across businesses, product-markets, functional departments, and activities within each business or
product-market.
3.
Identification of sustainable competitive advantage: how the organization will compete in each business
and product-market within the domain, and how it will position itself to develop and sustain a differential
advantage over current and potential competitors.
4.
Synergy: enabling firm's businesses, product-markets, resource deployments, and competences
complement and reinforce one other and as result the total performance of the related businesses is
becoming greater than the sum of its parts (Mullins & Walker, 2013, p. 39–40).
5.
Strategic Management Process is a sequential set of analyses and decisions (choices) that can increase the
likelihood that a company will choose and implement a strategy that generate competitive advantages.
The diagram below is presenting the managerial decision-making in a company as an interactive analytical
modeling process and will be repeatedly used as a starting point for the topics covered in the remaining weeks of
this course.
Strategic Management Process
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
17 of 40 6/18/17, 11:26 AM
This diagram is also used as a starting point for the reviewing of the decision support tools and their applications
in AD715. The framework of the strategic management process is explained with the help of a business running
case, a business simulation, and several tutorials:
Business Running Case ‘Investing in a New BrewPub? (Conceptual Study by the Owner)’
Click here to download the Business Case
Click here to access a database with industry resources, needed for a speedy understanding of
important external and internal driving forces of the beer industry (U.S. market)
1.
Business Simulation 'Strategies and Decision Support in Organizations'
Click here to review the "Tutorial for AD715: Introduction to the Business Simulation, Strategies and
Decision Support in Organizations" and the business simulation. The diagram below is presenting
the navigation page of the business simulation.
2.
Selected topics from different functional business areas of the company:3.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
18 of 40 6/18/17, 11:26 AM
Marketing Management and Decision Making: Lecture 7
Innovation and Technology Management and Decision Making: Lecture 8
Operations Management and Decision Making: Lecture 9
Financial Management and Decision Making: Lecture 10
Performance Management, Data Visualization, Analysis and Decision Making: Lecture 11
Organizational and HR Management and Decision Making: Lecture 12
Students are asked to learn, to understand, and to demonstrate the complexity of the managerial decision
making process in the form of:
a managerial report, due last week of the class, completed as Assignment 3: What is the Rationale for
Investing in a New BrewPub? Conceptual Study by the Owner of an Existing Restaurant/Tavern (Based
on a running case and a business simulation package)
1.
a project presentation, due last week of the class, based on Assignment 3 (executive summary of the
results & recommendations & lessons learned)
2.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
19 of 40 6/18/17, 11:26 AM
International Institute of Business Analysis. (2015). A guide to the business analysis body of knowledge, 3.
http://www.iiba.org/babok-guide.aspx
McKinsey Global Institute. (2011, May). Big data: The next frontier for innovation, competition, and productivity.
Mintzberg, H. (1994). The rise and fall of strategic planning. New York: Free Press.
Mullins, J. & Walker, O. (2013). Marketing management. Marlborough, MA: McGraw-Hill/Irwin.
Porter, M. (1996). On competition. Boston: Harvard Business School.
Render, B., Stair, R. M., Hanna, M. E., & Hale, T. S. (2015). Quantitative analysis for management (12th ed.).
Pearson Edition.
Stubbs, E. (2011). The value of business analytics: Identifying the path to profitability. Hoboken, NJ: Wiley.
Stubbs, E. (2014). Big data, big innovation: Enabling competitive differentiation through business analytics.
Hoboken, NJ: Wiley.
Module 6 – Decision Analysis and Decision Support in Business
After you complete this lecture, you will be familiar with the following concepts:
Managerial Decision Making Process: the Decision Analysis Prospective
Familiarity with the Types of Decision Making Environments
Decision Making Under Uncertainty with no Prior Knowledge of the Probabilities of the States of Nature
Decision Making Under Risk – Decision Making with Probabilities
Develop Accurate and Useful Decision Trees
Use Software Applications for Payoff Tables and Decision Tree Problems - MS Excel Add-Ins TreePlan
Disclaimer Lecture 6 is an abridged version of ‘Quantitative Analysis for Management’, 12e, by Render, Stair,
Hanna and Hale Copyright ©2015 Pearson Education, Chapter 3: Decision Analysis, pages 65–95
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
20 of 40 6/18/17, 11:26 AM
Based on your familiarities with the six steps of the managerial decision making process (covered in Lecture 1)
and the steps of the quantitative analysis (covered in Lecture 5), we can define the following decision analysis
framework to the decision making process:
Clearly define the problem at hand1.
List the possible alternatives2.
Identify the possible outcomes or states of nature3.
List the payoff (typically profit) of each combination of alternatives and outcomes4.
Select one of the mathematical decision theory models5.
Apply the model and make your decision6.
The diagram below is presenting this framework.
The demonstration of the decision making process as a step-by-step analytical approach will be presented with
the help of a business running case ‘Thompson Lumber Company – see the diagrams below.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
21 of 40 6/18/17, 11:26 AM
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
22 of 40 6/18/17, 11:26 AM
In this lecture, we will discuss three basic types of decision making environments:
Decision making under certainty - The decision maker knows with certainty the consequences of every
alternative or decision choice
Decision making under uncertainty - The decision maker does not know the probabilities of the various
outcomes
Decision making under risk - The decision maker knows the probabilities of the various outcomes
Decision making under certainty - The decision makers know with certainty the consequence of every
alternative or decision choice.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
23 of 40 6/18/17, 11:26 AM
Let us first consider approaches to decision making that do not require knowledge of the probabilities of the
states of nature. Decision makers are using these approaches in cases where they have little confidence in the
assessment of the probabilities and are making their judgment and selecting the most appropriate approach from
the following five possible options:
Maximax (optimistic)
Maximin (pessimistic)
Criterion of realism (Hurwicz)
Equally likely (Laplace)
Minimax regret
Maximax (optimistic) vs Maximin (pessimistic)
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
24 of 40 6/18/17, 11:26 AM
Criterion of Realism (Hurwicz)
Equally Likely (Laplace)
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
25 of 40 6/18/17, 11:26 AM
Minimax Regret Decision (Opportunity Loss)
Decision making under risk is a decision situation in which several possible states of nature may occur, and the
probabilities of these states of nature are known.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
26 of 40 6/18/17, 11:26 AM
In the approach called decision making with probability the decision maker subjectively is assessing the
probabilities of the values of the states of the nature \(P(X_i)\).
The number of state of nature \(i = 1, 2, 3,\ldots,N\) is different than 1 but one and only one state of nature can
occur in one particular period of time (e.g. day, week, month).
When there are several possible states of nature and the probabilities associated with each possible state of
nature are known, it is possible to determine the expected monetary value (EMV) for each alternative.
EMV is known as the most popular method for making decisions under risk where we are choosing the
alternative with the highest expected monetary value – see the diagrams, below.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
27 of 40 6/18/17, 11:26 AM
Expected Value of Perfect Information (EVPI) places an upper bound on what you should pay for additional
information and can be presented as follow:
EVPI = EVwPI – Best EMV
Where
Expected Value with Perfect Information EVwPI is the long run average return if we have perfect
information before a decision is made
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
28 of 40 6/18/17, 11:26 AM
Expected opportunity loss (EOL) is the cost of not picking the best solution
Construct an opportunity loss table
For each alternative, multiply the opportunity loss by the probability of that loss
for each possible outcome and add these together
Minimum EOL will always result in the same decision as maximum EMV
Minimum EOL will always equal EVPI
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
29 of 40 6/18/17, 11:26 AM
The previous examples have illustrated how to apply the decision making criterion when the payoffs are to be
maximized. The following example illustrates how the criterial are applied to problems in which the payoffs are
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
30 of 40 6/18/17, 11:26 AM
costs that should be minimized.
Problem (Text, p. 75–77):
A department will be signing three year lease for a new copy machine and three different machines are being
considered
For each of the machines, there is a monthly fee (incl. monthly fee & charge per each copy)
The department has estimated that the number of copies/Mo could be 10,000 or 20,000 or 30,000
The monthly cost for each machine based on the offers and the three levels of activities is shown in the
table below
Which machine should be selected?
To determine the best alternative, a specific criterion must be chosen. In the diagrams below we will demonstrate
the costs minimizations approach with the help of different criterions, discussed earlier in this lecture.
Criterion of Realism
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
31 of 40 6/18/17, 11:26 AM
Equally Likely Criterion
EMV Criterion
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
32 of 40 6/18/17, 11:26 AM
EOL Criterion
Any problem that can be presented in a decision table can be graphically illustrated in a decision tree:
Decision tree is most beneficial when a sequence of decisions must be made
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
33 of 40 6/18/17, 11:26 AM
All decision trees contain decision points/nodes and state-of-nature points/nodes
At decision nodes one of several alternatives may be chosen
At state-of-nature nodes one state of nature will occur
Categories and Definitions:
State of Nature: one and only one of all possible states that will occur
Influence Diagram: graphical device that shows the relationships among the decisions, the chance
events, and the consequences for a decision problem
Payoff: the consequence resulting from a specific combination of a decision alternative and a state of
nature
Payoff Table: a table showing payoffs for all combinations of decision alternatives and states of nature
Decision Tree: a graphical representation of the decision-making process which is based on the natural
or logical progression that will occur over time
Five Steps of Decision Tree Analysis Analyzing problems with decision trees involves five steps:
Define the problem1.
Structure or draw the decision tree2.
Assign probabilities to the states of nature3.
Estimate payoffs for each possible combination of alternatives and states of nature4.
Solve the problem by computing expected monetary values (EMVs) for each state of nature
node
5.
Structure of Decision Trees – Basic Rules:
Trees start from left to right
Trees represent decisions and outcomes in sequential order
Squares represent decision nodes
Circles represent states of nature nodes
Lines or branches connect the decisions nodes and the states of nature
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
34 of 40 6/18/17, 11:26 AM
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
35 of 40 6/18/17, 11:26 AM
Thompson's Complex Decision Tree
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
36 of 40 6/18/17, 11:26 AM
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
37 of 40 6/18/17, 11:26 AM
Expected Value of Sample Information Demonstrated with the Help of the Business Running Case 'Thompson Lumber Company'
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
38 of 40 6/18/17, 11:26 AM
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
39 of 40 6/18/17, 11:26 AM
Excel Decision Tree Add-Ins
Lumenant by www.lumenaut.com
PrecisionTree 6 by www.palisade.com
Risk Solver Platform by www.solver.com
1.
Decision Tree in IBM SPSS
URL: www-03.ibm.com/software/products/en/spss-decision-trees
Demo: www-01.ibm.com/software/analytics/media/spss/decisiontrees
2.
Top Decision Tree Analysis Software Products 2014 (ranked by Capterra, www.capterra.com):
Analytica by www.lumina.com
1000Minds by www.1000minds.com
Blaze Advisor by www.fico.com
Decision Support Software by www.logicnets.com
DPL 8 Direct by www.syncopation.com
Simulation Modelling by www.rapidmodeling.com
Spotfire by www.tibco.com
VisiRule by www.lpa.co.uk
3.
Week 3 https://learn.bu.edu/bbcswebdav/pid-4715354-dt-content-rid-16560270_...
40 of 40 6/18/17, 11:26 AM
