BA 210: Decision Making

1
BA 210 Decision Making
2
Part Three Special Topics
Individual
Organization
Decision-Making
Diversity
Planning
Power, Politics and Careers
Operations Management
Entrepreneurship
Innovation
Change
3
Admin

• Exam 2 Results Review
• Turning in Extra Credit Hold onto the slips!
• Slips can be turned in either of last two classes
  (Wed Apr 30 or Mon May 5), or to any TA during
  office hours, by 5PM Fri May 9.
• The latter option (TA office hours) should only
  be used if you participate in an experiment after
  class on May 5.

4
Roadmap Objective

• Central task of managers making decisions
• Objective understand how we make decisions
• Decision-making process
• Types of decisions
• Rational vs. Boundedly rational decisions
• Heuristics and cognitive biases
• Types of decision-makers
• Pareto diagrams

5
The Decision-Making Process
3. Allocation of Weights to Criteria
1. Problem Identification
2. Identification of Decision Criteria
4. Development of Alternatives
My salespeople need new computers
Price Weight Warranty Screen type Reliability
Reliability 10 Warranty 5 Weight
5 Price 4 Screen type 3
Compaq Gateway HP Sony Toshiba
6
Types of Problems and Decisions

• Well-Structured Problems - straightforward,
  familiar, and easily defined
• Programmed Decisions
• Routines Procedures, rules, and policies
• Minimize need for discretion facilitate
  efficiency
• Poorly-Structured Problems - new, unusual
  problems for which information is ambiguous or
  incomplete
• Nonprogrammed Decisions a unique decision to
  address (poorly-structured) problems that require
  custom-made solutions
• More frequent among higher-level managers
• Most real-world decisions have elements of both

7
Rational Decision Making

• Decisions are optimal (value-maximizing) given
  the constraints of the situation.
• Rational decision-making is the desired and
  default mode.
• But, assumptions of rational decision-making
  model are highly demanding (see next slide).
• So managerial decisions seldom meet all the tests

8
The imposing set of assumptions associated with
Rational Decision-Making
Understand main point only
9
Boundedly Rational Decision-Making

• Bounded Rationality behaving rationally within
  limitations imposed by cognitive abilities, and
  complexity of environment.
• Satisficing - Searching for and choosing an
  acceptable, or satisfactory response good
  enough - rather than trying to make the optimal
  decision.

10
Boundedly Rational Decision Process Why do we
make satisficing decisions?
Satisficing decision
Information processing biases
Uncertainty, Risk, Ambiguity

Perceived problem
11
Decision-Making in increasingly foggy conditions
Key concepts

• Certainty - outcome of every alternative is known
• Risk - able to estimate the probability of
  outcomes stemming from each alternative
• Uncertainty - not certain about outcomes and
  unable to estimate probabilities
• Ambiguity not certain about what the meaning of
  the data is in the first place
• Limited Information Because of time constraints
  and information costs you may not have the time,
  and/or not have the resources, to gather all the
  information you would like to have.

12
Ambiguous Information

• Ambiguous Information
• Information whose meaning is not clear allowing
  it to be interpreted in multiple or conflicting
  ways.

Young Woman or Old Woman?
13
Decisions in foggy conditionsWhich concept is
evident?

• Managers at Nike had an opportunity to acquire
  North Face, an outdoor gear manufacturer. Some
  managers through this was an opportunity to enter
  a booming market, whereas others thought it was
  too troublesome so that internal development of
  new products should be favored?
• Apple launches Newton (the first PDA). Whats
  the probability of success? Apple is first to
  market so there is no body of knowledge to
  estimate chances of success.
• Managers in biotechnology can draw on the
  experience of thousands of drugs to know that
  only there is only a 10 chance a drug will
  successfully pass advanced clinical trials.
• The sun will rise tomorrow.
• You are a purchasing manager at Ford. You have a
  month to choose a supplier for a small engine
  part. There are 20,000 auto suppliers in the
  United States alone

14
Boundedly rational decision-making usually
involves heuristics

• Heuristics
• Saves effort if the heuristic is right,
• If heuristic is wrong, can have cognitive bias
  systematically repeated errors that arise from
  the decision-making process in use.
• Analyse the heuristics you use, and assess their
  appropriateness

15
Example of Decision-Making Heuristics
Availability Bias

• Which causes more deaths annually in the U.S.?
• Lung Cancer or Motor Vehicle Accidents
• Emphysema or Homicide
• Tuberculosis or Fire and Flames

16
Additional Key Cognitive Biases

• Prior Hypothesis Bias strong prior beliefs about
  a relationship between variables influences
  decisions even when evidence is they are wrong.
• Representativeness
• Illusion of Control The tendency to overestimate
  ones own ability to control activities and
  events.
• Escalating Commitment Committing considerable
  resources to project and then committing more
  even if evidence shows the project is failing.

17
Examples of Cognitive biases

• CanGo
• Nike has had a near-fanatical emphasis on doing
  things the Nike Way. Nike has repeatedly
  stumbled recently, being late in responding to
  market change. For example, Nikes long-standing
  emphasis on performance led it to devote too many
  resources to expensive 140 shoes, rather than
  midrange 60-90 shoes that are the majority of
  revenues.
• Montgomery Ward shelves plans for expansion after
  World War II because CEO Avery believed, as with
  World War I, there would be a depression after
  the war that depressions always follow wars.
  Sears bet differently and won big.
• Autocratic Nissan CEO Kawamata runs firm in
  highly centralized fashion and makes some
  spectacular mistakes, notably changing the
  companys name from Datsun to Nissan.
• In 1970s, U.S. steelmakers respond to low-cost
  competition from minimills and foreign
  steelmakers by increasing their investments in
  their existing, technologically obsolete
  steelmaking plants rather than changing
  technologies more fundamentally.

18
Decision-Making Styles
Analytic
Conceptual
Directive
Behavioral
19
Pareto Diagrams to Aid Decisions

• Example of Visual display of data
• By type, time, geographic location, person
• Reasons for using

20
Firestone Clearly Knew Something Was
Amiss Firestone knew it had a problem with its
treads There is also a mountain of statistics
that, to the trained eye, reveals a disturbing
pattern A chart distributed inside the company
by Firestones financial analysts early this year
showed that nearly 60 percent of claims against
the company in 1999 were for tires manufactured
during the Decatur strike.
21
Pareto Diagram Data Categorized by Defect
19
18
Pareto principle
15
14
10
Frequency
Source Adapted from E. S. Fine. Pareto diagrams
get to the root of process problems. Quality,
October 1996, available at www.qualitymag.com/1096
ql (January 16, 1998).
2
Rough Finish
Under- size
Over- size
Out of round
Dis- colored
Misc.
Defect
22
Pareto Diagram Data Categorized by Shift
35
Source Adapted from E. S. Fine. Pareto diagrams
get to the root of process problems. Quality,
October 1996, available at www.qualitymag.com/1096
ql (January 16, 1998).
9
Frequency
7
7
3
6
2
5
0
4
Mon Tue Mon Wed Thu
Wed Tue Fri Thu
Fri
1 2 2 1
1 2 1
1 2 2
Defect
23
Pareto Diagram Data Categorized by Size, First
Monday Shift
18
13
Source Adapted from E. S. Fine. Pareto diagrams
get to the root of process problems. Quality,
October 1996, available at www.qualitymag.com/1096
ql (January 16, 1998).
Frequency
2
2
Under- size
Over- size
Out of round
Misc.
Defect
24
Next Up Planning

• Read after class the sections we cover from
• Ch 7 - Foundations of Planning
• T123 Skip 6, 10
• Ch 9 Planning Tools and Techniques
• T123 Skip 2, 3, 11, 12, 15
• CanGo Planning module
• Complete the introduction, concepts, exercises
  and resolution sections of the Planning module
• Grin and bear it

25
Results
Availability bias ease of recall leads to
over-estimating
26
Results
Availability bias ease of recall leads to
over-estimating
27
Results
Availability bias ease of recall leads to
over-estimating