Management Support Systems and DecisionMaking

1
Management Support Systems and Decision-Making
2
Supporting Managers with Information Systems
3
Models and Methods for Management Support

• To understand how computers support managers, it
  is necessary to understand what managers do.
• It is difficult to produce a standard job
  description for all managers.

4
Fundamental Functions of Management

• The traditional description of what managers do
  was first characterized by French industrialist
  Henri Fayol in his 1916 classic, Administration
  Industrielle et Generale. Fayol considered the
  manager's job as a composite of four separate
  functions
• Planning
• Controlling
• Leading
• Organizing

5
Fundamental Functions of Management - defined

• Planning - establishing goals and selecting the
  actions needed to achieve them over a specific
  period of time.
• Controlling - measuring performance against the
  planned objectives and initiating corrective
  action.
• Leading - inducing the people in the organization
  to contribute to its goals
• Organizing - establishing and staffing an
  organizational structure for performing business
  activities

6
Mintzbergs Studies of Managers

• Myth 1 The manager is a reflective
  systematic planner.
• Fact Study after study shows managers
  work at an unrelenting pace, that their
  activities are characterized by brevity, variety,
  and discontinuity, they are strongly oriented
  toward action, and dislike reflective activities.
• Myth 2 The effective manager has no regular
  duties to perform.
• Fact Managerial work involves
  performing a number of regular duties, including
  ritual and ceremony, negotiations, and processing
  of soft information that links the organization
  with its environment

7
Mintzbergs Studies of Managers

• Myth 3 The senior manager needs aggregated
  information, which a formal management
  information system best provides.
• Fact Managers strongly favor verbal
  media, telephone calls, and meetings over
  documents.
• Myth 4 Management is, or at least is quickly
  becoming, a science and a profession.
• Fact The managers' programs - to
  schedule time, process information, make
  decisions, and so on-remain locked deep inside
  their brains.

8
Classic Study of Managerial Work

• The classic study of managerial work was done by
  Mintzberg, who divided the managers roles into
  three categories
• 1. Interpersonal roles
• 2. Informational roles
• 3. Decisional roles

9
Management Roles

• Interpersonal Roles
• Figurehead, Leader, Liaison
• Informational Roles
• Monitor, Disseminator, Spokesman
• Decisional Roles
• Entrepreneur, Disturbance Handler, Resource
  Allocator, Negotiator

10
Mintzberg The Nature of Managerial Work
Formal Authority and Status
Interpersonal Roles
Informational Roles
Decisional Roles
11
Mintzbergs Management Roles

• Interpersonal Roles
• Figurehead - Carries out a symbolic role as head
  of the organization, performing duties of a legal
  or social nature.
• Leader - In the most widely recognized managerial
  duty, the executive is responsible for motivating
  and "activation" of subordinates, as well as
  staffing, training, promoting.
• Liaison - Develops and maintains a personal
  network of external contacts who provide
  information and favors.

12
Mintzbergs Management Roles

• Informational Roles
• Monitor - Seeks and receives a wide variety of
  special information (much of it current) to
  develop a thorough understanding of the
  organization and the environment. In this role,
  the executive serves as the nerve center of
  internal and external information about the
  organization.
• Disseminator - Transmits information received
  from outsiders or subordinates to other members
  of the organization. Some information is
  factual, some involves interpretation and
  integration of diverse value positions of
  organizational influencers. All information is
  to guide subordinates in decision making.
• Spokesman - Communicates information to outsiders
  on the organization's plans, policies, actions,
  results, etc. serves as the expert on the
  organization's industry.

13
Mintzbergs Management Roles

• Entrepreneur - Searches the organization and
  environment for opportunities and initiates
  "improvement projects" to bring about change
  supervises design of certain projects as well.
• Disturbance Handler - Responsible for corrective
  action when the organization faces important,
  unexpected disturbances.
• Resource Allocator - Allocates organizational
  resources of all kinds-in effect the making or
  approval of all significant organizational
  decisions.
• Negotiator - Represents the organization in major
  negotiations.

14
IS and Mintzbergs Roles
15
Information Support for Management

• Early information systems mainly supported the
  informational roles.
• The purpose of recent information systems is to
  support all three roles.
• We will explore the information support required
  for all roles, beginning with the decisional
  roles.
• The success of management depends on the
  execution of managerial functions such as
  planning, organizing, leading, and controlling.
  To carry out these functions, managers engage in
  the continuous process of making decisions.

16
Executive Activities and Information Support

• Handling Disturbances (42) - A disturbance is
  something that happens unexpectedly and demands
  immediate attention, but it might take weeks or
  months to resolve.
• Entrepreneurial Activity (32) - activities
  intended to make improvements that will increase
  performance levels. Improvements are strategic
  and long term in nature.
• Resource Allocation (17) - Allocating resources
  within the framework of the annual and monthly
  planning tasks and budgets
• Negotiations (3) resolve conflicts and disputes,
  either internal or external.
• Other Activities (6)

17
Introduction to Decision-Making

• A basic understanding of decision making is
  essential because most information systems are
  designed to support decision making in one way or
  another.
• We will survey some models and concepts of
  decision making and methods for deciding among
  alternatives.
• We will look at their relevance to information
  systems design.

18
Decision MakingPhases

• Herbert A. Simon (1960) proposed the most famous
  model of the Decision-Making process.
• 1. Intelligence
• 2. Design
• 3. Choice
• Some models of decision making include a 4th
  step Implementation.
• There is a flow activities from one phase, to the
  next. At any time there may be a return to a
  previous phase.

19
Simons ModelFlowchart of Decision Process
Intelligence
Design
Choice
20
Intelligence Phase

• Searching the environment for conditions calling
  for decisions
• Data inputs obtained, processed, examined for
  clues to identify problems or opportunities
• Identify problems for opportunity situations
  requiring design and choice.
• Scanning the environment, intermittently or
  continuously, is important.
• Organizational objectives
• search and scanning procedures
• data collection
• problem identification
• problem classification
• problem statement

21
Examples of the Intelligence Phase

• Air traffic controller continuously scanning to
  detect problems in air space.
• Each time you start your car, there is a
  conscious or unconscious scanning (listening,
  checking gauges, etc.).
• Marketing executive makes periodic visits to key
  customers to review possible problems and
  identify new customer needs.
• A plant manager reviews daily scrap report to
  check for quality control problems.
• An executive reads the industry trade paper to be
  aware of events and changes in the environment.

22
Summary Intelligence Phase

• Intelligence activities result in dissatisfaction
  with the current state or identification of
  potential rewards from a new state.

23
Design Phase

• Inventing, developing, and analyzing possible
  courses of action
• This involves processes to understand the
  problem, to generate solutions and test solutions
  for feasibility
• Formulate a model.
• Set criteria for choice.
• Search for alternatives
• Predict and measure outcomes

24
Choice Phase

• Select an alternative from those available
• Select and implement a choice
• Solution to the model
• sensitivity analysis
• selection of best (good) alternatives(s)
• plan for implementation (action)

25
Comment on Simons Model

• Simons Model does not go beyond the choice
  phase.
• There are no steps for implementation, or
  feedback from the results of the decision.
• Although Simons model is the most famous, others
  have adapted it.
• Our textbook provides a similar model

26
(No Transcript)
27
Alter Textbook Model

• Decision-making is represented as a
  problem-solving process preceded by a separate
  problem-finding process.
• Problem-solving is the use of information,
  knowledge, and intuition to solve a problem that
  ha previously been defined.

28
An Alternative Model Rubenstein and
Haberstrohs

• 1. Recognition of problem or need for decision
• 2. Analysis and statement of alternatives
• 3. Choice among the alternatives
• 4. Communication and implementation
• 5. Follow-up and feedback of results

29
Slades Model of Decision Making
Identify Problem
Identify Alternatives
Choose Usual Action
Evaluate Alternatives
Choose Among Alternatives
Generate New Alternatives
Effect Choice
Abandon Problem
30
Summary

• All models indicate the same basic ideas
• 1. Problem finding - Identify situations where
  problems need to be solved.
• 2 Problem formulation - clearly state the
  problem.
• 3. Alternative Generation
• 4. Evaluate Outcomes.
• 5. Choice
• 6. Implement
• 7. Evaluate..

31
Problem Finding I

• It is the difference between existing state and
  the desired state
• The problem finder usually has an idea of the
  desired state ( a model)
• Compared with the reality and differences noted
• A Problem exists when there is a major difference

32
The role of models in decision-making

• A major characteristic of decision-making is the
  use of models.
• A model is a simplified representation or
  abstraction of reality.
• It is usually simplified because reality is too
  complex to copy.
• Basis idea is that analysis is performed on a
  model rather than on reality itself.

33
Pounds Categories of Models - Expectations
against which reality is measured

• Historical - expectation based on extrapolation
  of past experience.
• Planning - the plan is the expectation
• Inter-organizational - Models of other people in
  the organization (e.g. superiors, subordinates,
  other departments, etc.)
• Extra-organizational - models where the
  expectations are derived from competition,
  customers, professional organizations, etc.

34
Another classification of models

• Iconic Models
• Analog Models
• Mathematical Models
• Mental Models

35
Iconic and Analog Models

• Iconic (scale) models - the least abstract model,
  is a physical replica of a system, usually based
  on a different scale from the original. Iconic
  models can scale in two or three dimensions.
• Analog Models - Does not look like the real
  system, but behaves like it. Usually
  two-dimensional charts or diagrams. Examples
  organizational charts depict structure,
  authority, and responsibility relationships maps
  where different colors represent water or
  mountains stock market charts blueprints of a
  machine speedometer thermometer

36
Mathematical Models

• Mathematical (quantitative) models - the
  complexity of relationships sometimes can not be
  represented iconically or analogically, or such
  representations may be cumbersome or time
  consuming.A more abstract model is built with
  mathematics.
• Note recent advances in computer graphics use
  iconic and analog models to complement
  mathematical modeling.
• Visual simulation combines the three types of
  models.

37
Mental Models

• People often use a behavioral mental model.
• A mental model is an unworded description of how
  people think about a situation.
• The model can use the beliefs, assumptions,
  relationships, and flows of work as perceived by
  an individual.
• Mental models are a conceptual, internal
  representation, used to generate descriptions of
  problem structure, and make future predications
  of future related variables.
• Support for mental models are an important aspect
  of Executive Information Systems. We will discuss
  this in depth later.

38
Problem Formulation

• There is always the danger of solving the wrong
  problem.
• Here, you try to clarify the problem so that you
  work on the right problem
• Frequently, the process of clearly stating the
  problem is sufficient in other cases, reduction
  of complexity is needed.
• Some strategies to use for reducing complexity
  and formulating a manageable problem are shown in
  the next slide

39
Problem Formulation Strategies

• Determine problem boundaries (I.e. what is
  clearly part of the problem)
• Examine changes that precipitated the problem
• Break it down into smaller sub-problems
• Focus on controllable elements
• Relate to a previously solved class of problems,
  an analogy situation.

40
Alternative Generation

• A significant part of the process of
  decision-making is the generation of alternatives
  to be considered in the choice phase.
• This is a creative task and can be taught
• Can be enhanced by aids such as
• scenarios
• brainstorming
• analogies
• checklists, etc
• Requires Knowledge of the problem and its
  boundaries (domain knowledge), as well as
  motivation to solve the problem.

41
Decision-Making Concepts
42
Decision Making Concepts

• Decisions differ in a number of ways.
• The differences affect the alternative generation
  process, and how a final choice will be made.
• The differences can also affect how information
  systems and information technology can support
  the process at any one of the stages.
• Four dimensions of decision types
• I. Knowledge of Outcomes
• II. level of structure/programmability
• III. criteria for the decision
• IV. level of decision impact

43
Decision Making Concepts IKnowledge of Outcomes

• Outcome - what will happen if a particular
  alternative or course of action is chosen
• Knowledge of outcomes is important with multiple
  alternatives
• Three types of knowledge with respect to outcomes
  are usually distinguished
• Certainty
• Risk
• Uncertainty

44
Knowledge of OutcomesThree Types

• Certainty
• Complete and accurate knowledge of outcome of
  each alternative. There is only one outcome for
  each alternative.
• Risk
• Multiple outcomes for each alternative and a
  probability can be assigned to each
• Uncertainty
• Multiple outcomes for each alternative and a
  probability cannot be assigned to each

45
Decision-Making Under Conditions of Certainty
Rationality

• If the outcomes are known and the values of the
  outcomes are certain, the task of the
  decision-maker is to compute the optimal
  alternative or outcome.
• Are we rational decision makers?
• There is ongoing argument pro and con
• People are said to be limited rationalists
• We might look for a limited number of
  alternatives and decide

46
RationalityExample

• A rational decision maker is expected to decide
  on the optimal alternative or outcome
• The optimal alternative is one that is related to
  some optimization criteria such as minimize cost,
  for example
• Thus the rational decision maker chooses the one
  that has the minimum cost
• Consider purchasing two products that are
  identical in all respects and appear equal in
  value
• All other things being equal, the rational
  decision maker chooses the one with the lower
  cost
• Rare, since all things are rarely equal

47
Decision Making under Risk

• Risk is when multiple outcomes of each
  alternative is possible and a probability of
  occurrence can be associated with each
• In such cases, the general rule is to pick the
  one that has the highest expected value

48
RiskExpected Value

• Which would you choose?
• Action 1 offers 1 probability of a gain of
  15,000, or
• Action 2 that offers 50 probability of a gain of
  400
• Solution use Expected Value
• Expected value is defined as the product of the
  outcome and the probability of the outcome
• Expected value outcome x probability

49
RiskExpected Value (contd.)

• Action 1 Expected Value 0.01 x 15,000 150
• Action 2 - Expected Value. 0.5 x 400 200
• Action 2 has the higher expected value
• The rational decision maker chooses the strategy
  that has the higher expected value
• OK strategy if the probability is known

50
Decision Making Under Uncertainty

• Uncertainty is the situation where the outcomes
  are known, but the probabilities are unknown
• One solution is to somehow assign the
  probabilities and then convert it to a problem
  under risk.
• Other decision rules are to minimize regret and
  to use the maximum and minimum criteria. We will
  look at these later.
• Uses Bayesian decision theory which recommends
  maximizing subjective expected utility, and on
  decision analysis which uses decision trees,
  payoff matrices, and influence diagrams to
  implement Bayesian Decision Theory.

51
Decision-Making Concepts II Programmed vs. Non
Programmed Decisions

• We have reviewed this with the Gorry and Scott
  Morton Paper discussed earlier
• Programmed Decisions - those that can be
  pre-specified by a set of rules or decision
  procedures
• Non-programmed Decisions - those that do not have
  any pre-established decision rule or procedures

52
Criteria for Decision-Making III Normative vs.
Descriptive Models

• Normative or Prescriptive - a model of decision
  making that tells the decision maker how to make
  a class of decisions. These have been developed
  by economists, management scientists, etc.
  Examples Linear programming, game theory,
  capital budgeting, statistical decision theory.
• In normative models the criterion for selecting
  among alternatives is maximization or
  optimization of either utility or expected value.

53
Criteria for Decision-Making III Normative vs.
Descriptive Models

• Descriptive - a model of decision making that
  describes how decision makers actually make
  decisions. They are used primarily by behavioral
  scientists.
• Descriptive models introduce the concept of
  satisficing.
• These two models introduce the Rational Approach
  as well as behavioral approaches.

54
Criteria for Decision-Making IV Level of
Decision Impact

• What are the consequences of the Decision?
• Will the consequences affect choice?
• What are the consequences under conditions of
  certainty, risk, or uncertainty?

55
Management Support Systems and Decision-MakingPar
t II
56
Views or Models ofIndividual and
Organizational Decision-Making
57
Views or Models of Decision-Making

• The Rational Manager View
• The Satisficing, or Process-Oriented View
• The Organizational Procedures View
• The Political View
• The Individual Differences View
• The Garbage Can Model

58
The Rational Manager View

• Oldest Theory to be proposed and studied in
  detail - it is a normative model.
• Based Heavily on Theory of Economic Man developed
  in economics and applied to management.
• Assumes organizational actors have complete
  knowledge of a decision scenario, and complete
  knowledge of their preferences.
• An exhaustive search is made of all possible
  alternatives.

59
Rational Manager - 2

• Consequences of alternatives are evaluated in
  terms of known preferences.
• An optimal choice can be selected.
• Proponents of cost-benefit analysis adopt this
  view.
• Model is highly normative, and has little
  descriptive support in true form.
• It is impractical and over-idealized.
• Influenced all other views of decision-making.

60
The Satisficing Viewpoint

• Simon was among first to attack the Rational
  Viewpoint.
• Most decision situations provide limited
  knowledge on some aspect of the problem.
• Impractical to think of generating all possible
  relevant alternatives for a situation.
• The bounded rationality of the human mind would
  make all of this information unassimable.
• Simon argues we tend to satisfice, or settle for
  a choice after a moderate amount of search.

61
Satisficing View - 2

• Since search is not exhaustive, heuristics or
  rules of thumb are used to identify solutions
  that are good enough most of the time.
• Heuristics reflect bounded rationality, i.e. a
  compromise between the demands of the problem,
  and the capabilities and commitment of the
  decision-maker.
• Simons Model has had wide discussion.
• His model, called Administrative Man, is a
  rejection of the Economic Man theory.

62
Empirical Research

• Empirical research has shown the importance of
  rationality and bounded rationality in
  organizational decision-making.
• Rationality and bounded rationality may be viewed
  at opposite ends of a continuum with the decision
  setting playing a contingency role.
• Threatening environments, high uncertainty,
  external control decreased rationality.
• The more complex or turbulent the environment,
  the less rationality used.

63
Empirical Research - 2

• Comprehensiveness - a desire to be rational,
  reflects how exhaustive and inclusive the
  decision process is in seeking alternatives.

HIGH
Comprehensiveness
LOW
STABLE
UNSTABLE
Environment
64
Organizational Process View

• Cyert and March extended Simons concept of
  bounded rationality to the organizational
  setting.
• Organizational Decision-Making in terms of
• formal and informal structure of the organization
• standard operating procedures
• channels of communication
• Choice is made in terms of goals, on the basis of
  expectations.
• The organization is a coalition of participants
  with disparate demands, focus of attention, and
  limited ability to attend to all problems
  simultaneously.

65
Organizational Process View - 2

• Bargaining among coalitions produces agreements
  which are the organizations goals.
• Organizational expectations arise from inferences
  from available information.
• Choice emerges as the selection of the first
  alternative that expectations identify as
  acceptable in terms of goals.
• Choice in the short-run is driven by standard
  operating procedures.
• Choice in the long-run driven by organizational
  goals.

66
The Political View

• Here decision-making is a personalized bargaining
  process among organizational units.
• Power and Influence determine the outcome of any
  situation.
• The players act in terms of no consistent set of
  strategic objectives, but rather according to
  their personal goals, stakes, interests.
• Organizational choice is the result of the
  pulling and hauling that is organizational
  politics.

67
The Political View - 2

• One must understand the realities of power and
  the compromises and strategies necessary to mesh
  interests and constraints of the players.
• Decisions are made to enhance the winner's
  conception of organizational, group, or personal
  interests.
• Allison argued that Politics is a process or
  conflict and consensus Building.

68
Empirical Research - 3 (1992)

• Procedural Rationality, or the desire to follow a
  rational approach to decision-making, is
  independent of the political process
• Decisions can be both procedurally rational and
  political, or neither.
• They can be independent dimensions of the
  organizational decision-making process.

69
The Political View - 3

• An important sub-model is the concept of
  incremental change - because there are so many
  actors involved in an organizational decision
  setting, clear, rapid progress is rarely possible
• The result of political bargaining and compromise
  is incremental change, i.e. decision-makers move
  to situations which are only slightly different
  from the current situation.
• Lindbloom talked of this in The Science of
  Muddling Through (1959).
• Muddling through is explicitly anti-utopian - it
  is the best we can do.

70
The Individual Differences View

• This view focuses on the individual
  decision-maker and his/her cognitive style,
  information processing and problem-solving
  behavior.
• Some individuals have specialized styles of
  decision-making which are effective in some
  contexts, and less so in others.
• Personal rationality is subjective, and behavior
  is determined by the manner in which individuals
  process information.

71
Individual Differences - 2

• In the organizational context, managers develop
  their own mental models of problems and issues.
• Decision-making can be a mixture of rationality
  and intuition, based heavily of experience and
  style.
• MS and OR are attractive to an analytic,
  systematic style. They may be less attractive to
  managers with a more intuitive style.
• Personality (what a person thinks) vs. Cognitive
  Style (how a person thinks).

72
Cognitive Style Dimensions

• Left brain
• words
• analytic
• sequential
• active
• realistic
• planned

• Right brain
• images
• intuitive
• simultaneous
• receptive
• imaginative
• impulsive

73
Empirical Research - 4

• Bobbit and Ford (1980) saw that executives had
  firm pre-dispositions about how the process of
  looking for ideas should unfold.
• Executive attitudes are influences by belief
  structures and past experience - pragmatic.
• Those with a low tolerance for ambiguity and high
  need for structure will adopt a decision process
  that has a narrow search zone.
• Risk propensity and risk perception also play
  roles, with risk propensity dominating decision
  situations. (Sitkin and Pablo, 1992).

74
The Garbage Can Model

• Proposed by Cohen, March, and Olsen in 1972.
• Appropriate for highly complex, unstable, and
  ambiguous environments called organized
  anarchies.
• Decisions result from a complex interaction
  between four independent streams of events
• problems, solutions, participants, and choice
  opportunities.

75
Garbage Can Model - 2

• The interaction of these events creates a
  collection of
• choices looking for problems
• issues and feelings looking for decision
  situations in which they might be aired
• solutions looking for issues to which they might
  be the answer
• decision-makers looking for work.
• The four streams are independent in nature and
  interact in a random fashion.
• A decision is made only when the four streams
  happen to interact.

76
Garbage Can Model - contd.

• Good decisions are made when this happens at the
  right time.
• Solutions represent the ideas constantly flowing
  through an organization.
• Solutions are used to formulate problems.
• Note that managers often do not know what they
  want until they have some idea of what they can
  get.

77
Individual Aspects of Decision-Making
78
Human Expectations

• Humans display a variety of responses in decision
  making. Some are related to individual
  differences such as cognitive style, others re
  related to expectations.
• Role of expectations can be partially explained
  by
• theory of cognitive dissonance
• commitment theory
• theory of anticipatory regret

79
Theory of Cognitive Dissonance

• propagated by Leon Festinger
• explains behavior after a choice is made
• Selected alternative has some negative features
  and rejected ones have some positive features
• Decision maker has feelings of mental discomfort
  following a decision because of recognition of
  above
• Second-Guessing
• Ex. Purchase of car

80
Cognitive Dissonance - 2

• Customers might need to be bolstered about their
  decision
• Hence, sales procedures follow up a sale with a
  congratulatory letter to bolster the effect of
  cognitive dissonance reduction

81
Theory of commitment

• If the person knows the decision is not revocable
  (firm commitment to decision), then decision time
  increases and processes will be more careful
• Having spent time making decision, the decision
  maker is reluctant to change it

82
Theory of anticipatory regret

• The decision maker anticipates the regrets that
  might occur
• This inhibits the decision maker from making a
  decision without contemplating the consequences
• Can be used to lessen post-decision regret
  thinking about consequences before they happen
  reduces the psychological impact when hey happen.

83
Behavioral Aspects of Organizational
Decision-Making
84
Behavioral Aspects of Organizational Decision
Making

• Many Issues Related to the Organizational
  Procedures Viewpoint and the Political Viewpoint
• quasi-resolution of conflict
• uncertainty avoidance
• problemistic search
• organizational learning
• incremental decision making

85
Quasi-Resolution of Conflict

• An organization can be considered as coalition of
  members having different goals and unequal power
  t influence organizational objectives.
• There are conflicts among the goals of the
  various members (e.g. production, sales,
  inventory).
• Conflicts need to be resolved thru
• local rationality
• acceptable-level decision rules
• sequential attention to goals

86
Uncertainty Avoidance

• Organizations live in uncertain environments
• This theory assumes that organizations will seek
  to avoid risk and uncertainty at the expense of
  expected value
• A decision maker will be willing to accept a
  reduction in the expected value in exchange for
  an increase in the certainty of the outcome

87
Uncertainty Avoidance - 2

• Thus, a decision maker will choose a 90 chance
  of making 10 over a 12 chance of making 100
• The second alternative has a higher expected
  value
• The decision though, is the first alternative
• Major benefit - reduction in uncertainty

88
Uncertainty Avoidance - 3Legal Methods

• Short-run feedback and reaction cycle
• short feedback cycle allows frequent new
  decisions and thus reduce need to be concerned
  about future uncertainty.
• Negotiated environment
• organization seeks to control its environment
  through industry-wide conventional practices
  (sometimes just as restrictive or collusive
  behavior)
• long-term supply contracts, etc.

89
Problemistic Search

• Search for solutions is problem-stimulated
• Little planned search for solutions not motivated
  by problems
• Simple rules
• search locally close to present symptom
• if this fails, expand search to vulnerable areas
  before moving to other areas

90
Organizational Learning

• Organizations exhibit adaptive behavior over time
• They change their goals and revise problem search
  procedures on the basis of experience
• Aspiration levels for goals are assumed to change
  in response to results obtained
• Plans tend to reflect aspiration levels
• Information systems are an important factor in
  reconciling achievement level and aspiration level

91
Incremental Decision Making

• Decision making in organizations is confined to
  small changes from existing policy and procedures
• Emphasis is on correcting or improving existing
  policies and actions
• Emphasis on consensus
• Called muddling through by Lindbloom

92
Decision Making Under Psychological Stress
93
Decision Making under Stress

• based on the conflict-theory model of Janis and
  Mann (1977)
• Decision making causes stress but here the
  characteristic is that all the alternative
  courses of action appear to have serious
  undesirable outcomes.
• Symptoms of such conflict are
• apprehensiveness
• hesitation
• vacillation
• distress
• Decisions are made using coping patterns

94
Coping Patterns

• Used in emergency situations such as a flood or a
  fire
• Can be extended to situations where there exist
  serious threats
• Four Questions that determine the typical coping
  pattern.

95
Coping Patterns - Questions

• Q1 Are the risks serious in the absence of
  change?
• Q2 Are the risks serious if change is made?
• Q3 Is it realistic to hope for a better
  solution?
• Q4 Is there sufficient time to search and
  deliberate?

96
Coping Patterns - 3

• If answer to Q. 1 is yes, then next is relevant
• If answer to Q. 2 is yes, then go to Q. 3
• If answer to Q. 3 is no, then the coping pattern
  may be defensive avoidance
• If answer to Q. 3 and Q. 4 is yes, then the
  coping strategy can be a vigilant process of
  search, appraisal, and contingency planning.
• If answer to Q. 4 is no, (e.g. a fire) then the
  coping pattern may be hypervigilance

97
Hypervigilance

• Typical response to disasters
• The decision maker focuses on the expected
  unfavorable consequences and fails to process
  information indicating that they may not happen.
• Pressure is felt to take immediate action.
• Hastily choose without considering the overall
  result or other possible actions.

98
Defensive Avoidance

• This coping pattern is most appropriate for the
  design of information systems and decision
  support systems.
• Marked by decision maker avoiding exposure to
  disturbing information, wishful thinking,
  distortion of information received and selective
  inattention.
• If risk of postponing decision is low,
  procrastination is chosen.
• If not, buck passing is tried.
• Bolstering is used beforehand in the lack of
  complete information
• After decision, bolstering is used to reduce
  cognitive dissonance.

99
Defensive Avoidance - 2

• Some example bolstering tactics
• Exaggeration of favorable consequences
• minimizing unfavorable consequences
• Denial of adverse feelings
• Exaggeration of remoteness of action that will be
  required following decision
• Assuming lack of concern by society (iit is a
  private decision).
• Minimizing of personal responsibility.

100
Defensive Avoidance - 3

• This pattern can also be observed in a group
• Janis coined the term groupthink for collective
  defensive avoidance
• E.g. industry that fails to react to vigorous
  price, quality and design competition by foreign
  competitors
• Symptoms of groupthink - see next slide

101
Groupthink Symptoms

• Illusion of invulnerability - The company is
  large and powerful and has customer loyalty.
• Collective Rationalization - No one can match our
  research.
• Belief in the inherent morality of the group -
  The managers are the best trained and preserve
  traditional values.
• Stereotypes of outgroups - The competitors
  products are inferior. They can not provide
  service.

102
Groupthink Symptoms - contd.

• Direct pressure on dissenters -demotion or firing
  of managers who disagree on a subject.
• Self-censorship - The subject of foreign
  competition is never put on the table by anyone
  in the group.
• Illusion of unanimity - No one is objecting, so
  everyone must agree that foreign competition is
  not serious.
• Self-appointed mind guards - evidence that
  contradicts the thinking of the group is removed
  as it moves up the organization.

103
Groupthink Example PATCO Strike of 1981

• Illusion of Vulnerability - The air system can
  not survive long without air traffic controllers
  (ATCs). Plans to replace them will not work.
• Collective Rationalization - The oath not to
  strike wasnt binding in this case, even though a
  strike was illegal.
• Belief in Inherent morality of the group - The
  strike for higher pay is morally justified
  because ATCs are responsible for more lives now.
• Stereotypes of Outgroups - The government is a
  typical bureaucracy. Reagan is just bluffing on a
  threat to fire us. No one has listened to our
  complaints.

104
Groupthink Example PATCO Strike of 1981 - contd.

• Direct Pressure of Dissenters - John Feydon,
  President of PATCO until 1980, forced to resign
  because he did not support strike.
• Self-censorship - Quotes such as Doubts seemed
  in the minority...The union is tight, almost
  like a family. 20 of strike force returned to
  work.
• Illusion of Unanimity - Other unions offered
  token support for PATCO. AFL/CIO privately was
  critical of PATCOs strike.
• Self-Appointed Mind Guards - Negotiators claimed
  there was no alternative but to strike.

105
Deciding Among Alternatives
106
Introduction to Methods

• Numerous method help one decide among
  alternatives.
• They generally assume that all alternatives are
  known or can be know, even though the search
  process often stops well before all feasible
  alternatives have been examined.

107
Optimization Techniques Under Certainty

• All alternatives and their outcomes are known.
  The computational problem is to choose which one
  is optimal.
• Use optimization techniques
• systems of equations
• linear programming
• integer programming
• dynamic programming
• queuing models
• inventory models, etc.
• Capital budgeting analysis
• Break-even Analysis

108
Mathematical Programming

• Mathematical Programming is the name for a family
  of tools designed to solve managerial problems in
  which the decision maker must allocate scarce (or
  limited) resources among various activities to
  optimize a measurable goal.
• Example Distribution of machine time (the
  resource) among various products (the activities)
  is a typical allocation problem

109
Sample Linear Programming

• XYZ corporation makes servers. A decision must
  be made. How many servers should be produced
  next month in the Boston plant? Two types of
  servers are considered S-7 requires 300 days of
  labor and 10,000 in materials S-8 requires 500
  days of labor 15,000 in materials. The profit
  contribution of S-7 is 8,000 whereas that of S-8
  is 12,000. The plant has a capacity of 200,000
  days per month while the material budget is
  8,000,000 per month. Marketing requires that at
  least 100 units of S-8 be produced.
• Problem How many units of S-7 and S-8 should
  be produced?

110
The Model

• Decision Variables X units of S-7 to be
  produced Y units of S-8.
• Result Variable The total profit. The
  objective is to maximize total profit.
• Objective function
• Total Profit 8,000X 12,000Y
• Constraints
• Labor Constraint 300X 500Y lt 20,000 (in
  days)
• Budget Constraint 10,000X 15,000Y lt 8,000,000
  (in dollars)
• Marketing Requirement X gt 100 (in units).

111
Optimization Techniques

• Computer algorithms and programs are readily
  available to handle many problems of this class.
• The major problem is to construct the model
  correctly.
• Reference other books on Optimization,
  Mathematical Programming, or Operations Research,
  or Management Science for a further discussion of
  these models and their application.

112
Statistical Decision Theory

• Decision Theory provides a rational framework
  for choosing between alternative courses of
  action when the consequences resulting from
  choice are imperfectly known.
• The necessity of making decisions in the face of
  uncertainty is an integral part of our lives.
• The theory provides techniques for mathematically
  evaluating potential outcomes of alternative
  actions in a given decision situation.
• In all cases, the decision-Maker has an objective
  (e.g. maximize profit).
• Two methods Payoff Matrix and Decision Tree.

113
Statistical Decision TheoryPayoff Matrix

• The payoff matrix consists of rows for the
  alternatives or strategies available and columns
  for the conditions that affect the outcomes
• Each cell contains the payoff (the consequences)
  in dollars if that strategy is chosen and that
  state occurs
• If it is known with certainty which state will
  prevail, then choose the strategy that has the
  highest payoff for that state
• This is simply the strategy of maximizing
  expected utility.

114
General Payoff Matrix
States of Nature
n1
n2
Strategies
n3
n4
S1
S2
S3
115
Example 1 The Anniversary Problem
You are suddenly driving home from work in the
evening when you suddenly recall that your
wedding anniversary comes about this time of
year. In fact, it seems quite probable, (but
not certain), that it is today. You can still
stop at the local florist and buy a dozen roses,
or you may go home empty-handed and hope the
anniversary lies in the future. What do you do?
116
Possible Outcomes (States of Nature)
Decision Alternatives (Strategies)
It IS NOT Your Anniversary
It IS Your Anniversary
SPOUSE SUSPICIOUS AND YOU ARE OUT 50
Buy Flowers
DOMESTIC BLISS
Do Not Buy Flowers
SPOUSE IN TEARS AND YOU IN DOGHOUSE
STATUS QUO
Anniversary Problem Payoff Matrix
117
Decision Tree for Anniversary Problem
DOMESTIC BLISS
Anniversary
NOT Anniversary
Buy Flowers
50 LOSS ANDSUSPICIOUS WIFE
DOGHOUSE
Anniversary
Do Not Buy Flowers
NOT Anniversary
Decision Point
STATUS QUO
Resolution of Uncertainty
118
Example 2 Fast Service Restaurant
An entrepreneur is deciding among three
alternatives for a fast-service restaurant that
she owns (1) leave as is (2) refurbish t to
improve layout (3) or re-build completely to add
capacity and improve layout.
There are three significant, independent
conditions (assume only one can occur) that
affect the possible profit (payoff) from each
alternative strategies. These conditions are
(1) a competitor may open on a nearby property
(2) a proposed highway re-routing will change
the traffic passing by (3) conditions will stay
approximately the same as they are. What should
the entrepreneur do?
119
Payoff Matrix (in Thousands of )
New Competitor 0.20
Highway Rerouting 0.30
Same - 0.5
Strategies
Do Nothing
2
0
-1
Refurbish
4
-3
3
7
2
Rebuild
-10
120
Analysis with Knowledge

• If we assume conditions remain the same, Rebuild
  is the best strategy. (Payoff 7,000).
• If probabilities are assigned, using a criteria
  of maximizing expected value
• Do Nothing (0.5)(2) (.2)(0) (0.3)(-1) 0.7
  or 700
• Refurbish (0.5)(4) (0.2)(3) (0.3)(-3) 1.70
  or 1,700
• Rebuild (0.5)(7) (0.2)(2) (0.3)(-10) 0.90
  0r 900
• Therefore, refurbishment is the best choice.

121
Statistical Decision Theory Imperfect Knowledge
of Consequences

• If there is uncertainty about the probabilities
  of the various conditions, then you can use one
  of several rules for deciding
• minimize regret -select strategy which minimize
  the sum of regrets for the strategy
• maximin - Select strategy which has highest
  payoff if the worst state of nature occurs
  (pessimistic).
• maximax - Select strategy which has highest
  payoff if most favorable state of nature occurs
  (optimistic).
• Each one of these rules has been criticized in
  the literature. They have disadvantages if
  applied as a general decision rule. You must
  decide if the rule is appropriate for the
  situation.

122
Regret Definition

• The regrets are the differences between the best
  payoff for a state of nature and the other
  outcomes.
• To compute a matrix of regret, subtract the value
  in each entry in a column from the highest value
  in the column.
• Sum the rows to compute the regrets for each
  action (assuming the payoff matrix has columns
  for states and rows to show actions)

123
Analysis with Imperfect Knowledge

• Minimize Regret
• Do Nothing 5 3 0 8
• Refurbish 3 0 2 5
• Rebuild 0 1 9 10
• Therefore, action which minimizes regrets is to
  Refurbish.
• Maximin Rule Identify state with worst payoff,
  and choose strategy with least unfavorable
  payoff. Therefore, choose to Do Nothing.
• Maximax Rule Rebuild

124
Statistical Decision Theory

• When decisions must be made under uncertainty,
  the emphasis is on Bayesian decision theory which
  recommends maximizing subjective expected
  utility.
• Bayesian decision theory provides a framework in
  which all available information is used to deduce
  which of the decision alternatives is best
  according to the decision makers preferences.
• Distinguish between a good decision and a good
  outcome

125
The Concept of Utility

• Not all outcomes can be compared in terms of
  dollars. Dollars and other measures work well in
  a narrow range of values, but not at extremes
  (e.g. overtime pay).
• Here money is used as a substitute measure of the
  outcomes utility.
• Whereas utility may be linear in a certain range
  in comparison to money, it generally is not under
  all ranges.

126
Utility vs. Money
UTILITY
MONEY
The Linear Assumption of Money for Utility in
this Narrow Range
127
Indifference Curves

• Any two possible outcomes can be compared, and
  generally one can say which one is preferred.
• In some cases they may be equally desirable, in
  which case you are indifferent.
• Example You may prefer a weeks vacation in
  Florida rather than paid double time to work a
  week extra.
• There is a tradeoff here between two
  value-properties. (e.g. leisure time vs. money).

128
Indifference Curves
Money
I3
I2
I1
Leisure Time
129
Other Alternative Selection Techniques

• Ranking, Weighting, or Elimination by aspects -
  often used to evaluate competitive bids.
• Game Theory (for conflict bargaining) - when one
  decision unit (player) gains, the other loses.
• Classical Statistical Inference
• sampling
• probability distributions
• regression and correlation analysis
• testing of hypotheses

130
Rational Choice and the Framing of Problems

• Alternative descriptions of a problem often give
  rise to different preferences.
• Example Consider the following statistical
  information provided on two alternative
  treatments of lung cancer. The same statistics
  are presented in terms of survival rates and in
  terms of mortality rates to two groups of
  respondents.

131
Rational Choice and the Framing of Problems -
(contd.)

• Example Survival Frame
• Surgery Of 100 people having surgery, 90 live
  through the post-operative period, 68 are alive
  at the end of the first year and 34 are alive at
  the end of five years.
• Radiation Therapy Of 100 people having
  radiation therapy all live through the treatment,
  77 are alive at the end of one year, and 22 are
  alive at the end of five years.

132
Rational Choice and the Framing of Problems -
(contd.)

• Example Mortality Frame
• Surgery Of 100 people having surgery 10 die
  during surgery or the post-operative period, 32
  dies by the end of the first year, and 66 die by
  the end of five years.
• Radiation Therapy Of the 100 people having
  radiation therapy, none die during the treatment,
  23 die by the end of year one, and 78 die by the
  end of five years.

133
Framing Example Results

• The inconsequential difference in framing
  produced a marked effect. The overall percentage
  of respondents who favored radiation therapy rose
  from 18 in the survival frame to 44 in the
  mortality frame.
• Radiation Therapy appears better than surgery
  when stated as a reduction of the risk of
  immediate death from 10 to 0, rather than as an
  increase from 90 to 100 in rate of survival.
• Framing effect was similar for physicians,
  business students, and a group of clinic patients.

134
What do We learn from Framing?

• Normative models of choice, which assume
  invariance of preferences, can not provide an
  adequate descriptive account of choice behavior.

135
Sample Planning Models
136
Planning Models

• A planning model is a method for structuring,
  manipulating, and communicating future plans.
• Simple Profit Model
• Sales input variable
• Cost of Sales 0.4 x sales
• Gross Margin sales - cost of sales
• Operating expenses input variable
• Profit before taxes gross margin - operating
  expenses
• Taxes 0.48 x profit before taxes
• Net Profit profit before taxes - taxes

137
Sample Profit Plan
Sales 100,000 Less cost of Sales (
40,000) Gross Profit 60,000 Less
Operating Expenses ( 52,000) Profit Before
Taxes 8,000 Less Taxes (
3,840) Net Profit 4,160
138
Use of Planning Models

• Model building can begin with simple models
  calling for inputs of major, high-level items.
• Subsequent model development can expand the
  details f the model to calculate the high-level
  items from more basic input. Example
• Selling Expense 0.10 x sales
• Advertising expenses 0.05 x sales
• Interest expense 0.10 x average long-term debt
  0.12 x average short term loans
• bad debt expense 0.01 x accounts receivable
  balance at beginning of period
• administrative expense input variable
• operating expense selling advertising
  interest bad debt administrative expense.

139
Use of Planning Models - contd.

• These models are are characteristic of Managerial
  Accounting.
• Individual terms can be estimated using
  techniques of statistics based on past history.
• Planning Models provide opportunities for what
  if scenerios.

140
Summary and Relevance of Decision-Making Concepts
for Information Systems Design
141
How Information Systems Might Help Counteract
Common Flaws in Decision Making
POOR FRAMING Description Allowing a decision to
be influenced excessively by the language used
for describing the decision How an information
system might help Provide information
encouraging different ways to think about the
definition of the issue RECENCY
EFFECTS Description Giving undue weight to the
most recent information How an information system
might help Provide information showing how the
most recent information might not be
representative
142
How Information Systems Might Help Counteract
Common Flaws in Decision Making

• PRIMACY EFFECTS
• Description Giving undue weight to the first
  information received
• How an information system might help Show how
  some information is inconsistent with the first
  information received
• POOR PROBABILITY ESTIMATION
• Description Overestimating the probability of
  familiar or dramatic events underestimating the
  probability of negative events
• How an information system might help Make it
  easier to estimate probabilities based on
  pertinent data

143
How Information Systems Might Help Counteract
Common Flaws in Decision Making
OVERCONFIDENCE Description Believing too
strongly in ones own knowledge How an
information system might help Provide
counterexamples or models showing that other
conclusions might also make sense ESCALATION
PHENOMENA Description Unwillingness to abandon
courses of action decided upon previously How an
information system might help Provide
information or models showing how the current
approach might give poor results
144
How Information Systems Might Help Counteract
Common Flaws in Decision Making

• ASSOCIATION BIAS
• Description Reusing strategies that were
  successful in the past, regardless of whether
  they fit the current situation
• How an information system might help Provide
  information showing how the current situation
  differs from past situatioins
• GROUPTHINK
• Description Bowing to group consensus and
  cohesiveness instead of bringing out unpopular
  bias
• How an information system might help Provide
  information inconsistent with the current
  consensus and prove its relevance

145
Support for the Intelligence Phase

• The search process involves an examination of
  data both in predefined and in ad hoc ways.
  Information systems support should provide both
  capabilities.
• Scanning of internal and external databases for
  opportunities and problems.
• Filtering should be used to avoid information
  overload.
• Information system should scan all data and
  trigger a request for human examination of
  situations apparently calling for attention (e.g.
  examination of key indicators and critical
  success factors).

146
Support for the Intelligence Phase - contd.

• Routine and ad-hoc reports can aid in the
  intelligence phase (e.g. summarization,
  comparison, prediction, confirmation).
• Various models should be included in the scanning
  and report layouts (e.g. historical, planning,
  etc.).
• Either the system of the organization should
  provide communication channels for perceived
  problems to be moved up the organization until
  they can be acted upon.

147
Support for the Design Phase

• The information system should contain decision
  models to process data and generate alternative
  solutions.
• It should assist with checklists, templates of
  decision processes, scenarios, etc.
• The models should assist in analyzing the
  alternatives.

148
Support for the Choice Phase

• An information system is most effective if the
  results of design are presented in a
  decision-impelling format.
• Presentation of alternatives. Use of appropriate
  methods depending on presence of certainty, risk,
  uncertainty.
• When the choice is made, the role of the system
  changes to the collection of data for further
  feedback and assessment.

149
Relevance for Information System Design

• Provide support for decision-makers in
  semi-structured and unstructured situations by
  bringing together human judgment and computerized
  information. Structured problems are easily
  handled by methods of management science and
  operations research.
• Provide support for model development, whether
  formal models or mental models.
• Provide tailorability to style of the
  decision-maker.
• Recognize that organizations place constraints on
  the decision-maker. Rationality is not always an
  option.

150
Relevance for Information System Design - (contd.)

• Recognize that uncertainty is a part of life.
• Recognize that stress is a part of life.
• Looks at methods for sensitivity analysis
  what-if analysis and goal-seeking analysis,
  and other methods for deciding among
  alternatives.
• Promote organizational learning.
• Provide knowledge components for very difficult
  problems.
• Specific examples will be discussed when we look
  at specific examples for compute-based support of
  decision-making.

151
Fi