Decision Making, Systems,

Modeling and Support

Introduction

“The test of a first rate intelligence is

the ability to hold two opposed ideas
in the mind at the same time, and still
retain the ability to function”

F. Scott Fitzgerald, 1936

Decision Making Defined
 Choosing among two or more alternatives
 The outcomes of which are not perfectly
known
 For the purpose of solving a problem or
achieving a goal
Decision Making
Requirement
 Problem or Opportunity
 Decision Making Procedure
 Knowing about the problem or opportunity
 Data/Information/Knowledge
 Decision Making Tools
 Decision Maker’s Expertise & Experience
 Computer-Based Information Systems
 Data, Information and Knowledge
5

IS View

Data
Raw
numbers
Information
and facts Knowledge
Processe
d
data Mix of contextual Actions
information values
and experience Informed by Results
knowledge
Delivering value
to organization
Business View
DIKAR Model (adapted from Ward and Peppard (2002)
Problem or Opportunity

 Decision Makers can be reactive or

proactive
 You can seek out not only opportunities
but also problems
Decision Context
 The setting in which the decision is made
 Management level affects time horizon,
decision autonomy, resources, etc.
 Organization Design: Organization, authority
levels, and command unity.
 Degree of Maturity: an established or
emergent situation affects issue familliarity.
 Degree of Concurrency: Decision may be serial
or concurrent.
Decision Context
 Decision Making Models in Organizational
Context:
 Rational Model
 Satisficing Model
 Program Model
 Political Model
 Garbage Can Model
Rational Model
 Classical model of analytical decision-
making
 Decision makers are rational; They operate
autonomously in a word where all
alternatives, consequences, and
probabilities are known.
 Optimal alternative can be found
 Normative: how decision should be made.
Satisficing Model
 Rational behavior is bounded by human
capacities and the limited, incomplete, and
imperfect knowledge most decision-
makers have.
 Bounded Rationality: the decision-maker can’t
have complete knowledge of all the
alternatives to the consequences of choice.
 Decision makers will opt for “good enough”
solutions, rather than optimal ones.
Organizational Program
Model
 Organizational decisions are often based on
standard operating procedures, conformance
to group norms, budget limitations,
reinforcement of past decisions, training,
motivational programs, reward structures, etc.
 Decision making behaviour often what
worked before.
 Changes in support for decision makers
should be made in light of this situation.
Political Model
 Organizational decisions are made by
people with competing or conflicting goals.
 Often, power and influence, rather than
ratioanility, determine outcomes.
 Negotiation, bargaining, conflict resolution,
and compromise are important to decision
making.
Garbage Can Model
 Decision are “consequences of
intersections of problems looking for
solutions, solutions looking for problems,
and opportunities for decision making.”
 Decision are the result of random,
stochastic events: chance, luck, and
timing.
Decision Models in Practice
 Research found that:
 Decision makers are boundedly rational
 Power wins battles of choice
 Chance matters
 Most decisions contain rational, political,
program, and garbage can aspects
 Most decision makers behave rationally in
some ways, and not in other ways.
Decision Types
 Easy / Hard decisions
 Important / Unimportant decisions
 Routine / Unique decisions
 Personal / Business decisions
 Fun / Troubling decisions
 And so on…
Decision Types
 Decisions in different functional areas
often have proprietary knowledge
requirements and solution approaches.
 Marketing
 Accounting
 Personal
 Manufacturing
 Etc…
Decision Types
STRUCTURED SEMI- UNSTRUCTURED
STRUCTURED
 Established situation * Emergent situation
 Ordinary * Novel, unique
 Repetitive * One-shot
 Issues understand * Issues elusive
 Knowledge available * Knowledge hard to
get
 Programmable * Creative
Decision Types
 Programmed Decisions
 Complete, unambiguous solution procedure
(algorithm) is available for making optimal
decision.
 Non-Programmed Decisions
 For semi- & unstructured decisions, general
solution strategies may be available, but
optimal or correct decisions are not guaranted
 DSS were specifically designed to support
non-programmed decisions.
Decision Types
 Unilateral Decisions
 Single person makes the decision, but may
require knowledge from others
 Negotiated / Bargained Decisions
 Multiple participants must agree on decision
 Negotiated decisions become more complex
as the number of participants increases
 Politics and power can play as much as role as
the decision itself
Solution Types
 Optimization
 Finding the best solution
 Satisficing
 Finding a good – but not necessarily the best
solution to a problem
 Heuristics (“rule of thumb”)
 Commonly accepted guidelines or procedures
 Usually find a good solution.
Decision Making Process

In t e llige n ce

D e sign
M o n it o r in g

Ch o ice

Im p le m e n t at io n
Intelligence
 Finding the problem / set the goal
 Classify the problem:
 Programmed, Non-programmed
 Decompose problem into subproblems
 Assign problem ownership
 Gather needed data / information
Intelligence
 Since Simon’s original description of this phase
in 1965, other researchers too have separated
problem recognation from problem solving:
 Decision Recognition and diagnosis
 Mintzberg at al. 1976
 Noticing, Interpreting, and Incorporating Stimuli
 Kiesler & Sproull 1982
 Scanning, Noticing, and Constructing Meaning
 Berthon, Pitt & Morris, 1998
 Managers operate as Scanners, Trackers or Probers
 Zmud 1990
Intelligence
 Data collection issues:
 Data are not available
 Obtaining data may be expensive
 Data may not be accurate or precise enough
 Important data that influence the results may
be qualitative
 Information overload
 Outcomes (or results) may occur over an
extended period
Intelligence
 Problem classification
The conceptualization of a problem in an attempt to
place it in a definable category, possibly leading to a
standard solution
 Problem decomposition
Dividing complex problems into simpler subproblems
may help in solving the complex problem
 Problem ownership
The jurisdiction (authority) to solve a problem
Design
 Generate decision alternatives
 Establish criteria for evaluation of
alternatives
 Evaluate decision alternatives
 Modeling is commonly used for alternative
generation and evaluation
Choice
 Selection of most profitable, agreed-upon, most
acceptable, least offensive
 Plan for implementation & monitoring should be
part of the choice
 Selection of alternative
 Many times goals or criteria conflict, particularly when
others are affected
 The mathematically optimal solution may therefore
not always be implementable
 Outcomes of any given alternative can rarely be
predicted with certainty, hence the need for Sensitive
Analysis
Choice
 Sensitive Analysis:
 What will happen to decision outcomes when
reasonable changes occur in internal or
external factors?
 How will decision outcomes behave over time?
 Mathematically optimal alternatives may be
very sensitive to optimal conditions!
Implementation
 Implementation of a decision usually
requires others in the organization to
agree and cooperate
 Implementation requires understanding of
organizational culture and organizational
change
Organizational Culture
 Culture – is a set of major understandings
and assumptions shared by a group or
organization.
 Organizational culture – consists of major
uderstandings and assumptions for a
business, corporation, or organization.
Organizational Change
U n fr e e zi n g

How do organizations M o v in g
plan for, implement,
and handle change?

R e fr e e zi n g

Lewin & Schein’s

Theory of Change (1947)
Monitoring
 Not really a phase, but rather a continuous
organizational activity.
 Examination of internal and external
indicators of decision results.
 Use this information to steer current
decision or to initiate new decision
processes (feedback!)
 Dovetails into Intelligence phase.
The Decision-Making Cylce
Decision Maker &
Cognitive Styles
 Who makes the decisions?
 Individual decision makers vary in
 Intelligence,
 Knowledge
 Training
 Experience
 Personality, and
 Cognitive style
The Decision Maker
Cognitive & Decision-Making
Styles
 What: Assessment of differences in
personality as related to decision making
 Why: different decision making styles may
require different types of decision support.
 Cognitive Style: how do people make
decisions?
 Information Gathering style
 Information Evaluation style
Cognitive Styles
 Information Gathering Style
 Preceptive: individuals filter data based on
deviations from or conformities to their
expectations.
 Receptive: individuals focus on individual pieces
of data rather than on preconceived relations or
holistic models.
Cognitive Styles
 Information Evaluation Style
 Systematic/Analytic: individual structures the
problem and the available data into a method
which is expected to lead to a solution.
 Intuitive/Heuristic: individual examines a variety of
methods and solutions using a trial-and-error
approach.
Cognitive Styles
Cognitive Styles

 Again, why is this relevant:

- Match individual decision maker with DSS
interface which matches his or her cognitive style
- Since no cognitive style is superior, have the DSS
provide a counterbalance to the decision maker’s
cognitive style
Cognitive Styles
 Huber’s “Much Ado About Nothing” (1983)
- Cognitive styles can’t be reliably measured

- Other psycho-social and task aspects are relevant too

- Reinforcing a cognitive style may reinforce existing

decision maker biases
- Opposing a cognitive style is technically not yet
feasible and possibly unethical
 Following Huber’s article:
- Emphasis on cognitive processes, rather than on
cognitive styles
- Realization of decision making preferences: the MBTI!
Myers-Briggs Type Indicator
 Based on Jungian Psychology and
Phenomenology: the theory that current
perceptions, attitudes, beliefs, and world
views are shaped by our experiences.
 Preferences on 4 dimensions.
 No type is superior, they are merely different.
Myers-Briggs Type Indicator
 Orientation to Life or “World View”
- I – Introversion: interests are primarily in the inner
world of concepts and ideas.
- E – Extraversion: interests involve the outer world
of people and things.
 Ways of perceiving:
- S – Sensing: direct use of the sense
- N – Intuition: indirect perception by way of the
unconscious, such as hunches, women’s intuition,
or artistic creativity.
Myers-Briggs Type Indicator
 Ways of Judging:
- T – Thinking: a logical process that tends to strive
for impersonal findings.
- F – Feeling: a process of appreciating things that
are personal and subjective.
 Dealing with the World
- J – Judging: need to come to closure, need for a
solution. As soon as enough data is available for a
decision, perception shuts down.
- P – Perceiving: tends not to judge, keeps
perception open to collect more data.
Myers-Briggs Type Indicator
 The correct administration and evaluation of
the MBTI test requires an organizational
psychologist or counselor.
 Cheaper, computerized versions (such as the
Keirsey Temperament Sorter Test) are
avaiable but not as sophisticated.
Myers-Briggs Vs Keirsey
Keirsey Indicator Results
Decision Making: Common
Flaws
 Why are mistakes made?
 Limits placed on decision-making
- Cognitive limits: decision makers are only
human
- Economic limits: decision making is expensive
- Time limits: speed thing up
- Competitive demands: staying off the joneses
Decision Making: Common
Flaws
 Primacy effects: once people develop an
opinion about something or a frame of
reference for analyzing an issue, it is often
difficult for them to move from that position
 Recency effects: making decisions based on
recently seen information
 Availability bias: making decisions based on
the most easily available information
Decision Making: Common
Flaws
 Poor framing: allowing a decision to be
“framed” by the language or context in which
it is presented
 Strategic anchoring: the initial assessment of
the decision sets the tone for all that follows.
Whoever sets this ‘anchor’ is at an advantage
Decision Making: Common
flaws
 Poor probability estimation: overestimating
the probability of events that are familiar or
dramatic, are under one’s control, or are
beneficial, while understanding the
probability of negative events.
 Escalation phenomena: finding it difficult to
abandon already adopted courses of action,
and ignoring feedback indicating the course
of action is failing
Decision Making: Common
Flaws
 Overconfidence: Both experts and general
public tend to be overconfident about the
accuracy of what they know
 Association bias: trying to repeat past
successes by choosing strategies more
related to a past situation than the current
one
 Groupthink: the need to maintain group
cohesion overpowers the group’s desire to
make the best decision