A DECISION SUPPORT FRAMEWORK

(BY GORY AND SCOTT-MORTEN, 1971)

A DECISION SUPPORT FRAMEWORK –
CONT.
 Degree of Structuredness (Simon, 1977)
 Decision are classified as
 Highly structured (a.k.a. programmed)
 Semi-structured
 Highly unstructured (i.e., non-programmed)
 Types of Control (Anthony, 1965)
 Strategicplanning (top-level, long-range)
 Management control (tactical planning)
 Operational control
TYPES OF DECISIONS
 Structured processes are routine & typically
repetitive problems for which standard
solution method exist…all phases structured.
 Unstructured processes are fuzzy, complex
problems for which there are no cut & dried
solution methods…all phases unstructured
 Semi structured fall between having some
structured elements & some
unstructured…..trading bonds, setting
marketing budgets for consumer products.
TYPE OF CONTROL
 Strategic Planning….defining long range goals
& policies for resource allocation
 Management control….acquisition & efficient
use of resources in the accomplishment of
organizational goals
 Opeartional control…efficient & effective
execution of specific tasks.
DECISION SUPPORT MATRIX
 Initial purpose of the matrix suggest
different type of computerized support to
different cells in the matrix
 Later..suggested that for semistructured &
unstructured conventional MIS & MS tools are
insufficient
 Human intellect & a different approach to
computer technologies are necessary.
 Proposed use of supportive IS…DSS
COMPUTER SUPPORT FOR
STRUCTURED DECISIONS
 Structured problems: encountered repeatedly,
have a high level of structure
 It is possible to abstract, analyze, and classify
them into specific categories
 e.g.,
make-or-buy decisions, capital budgeting,
resource allocation, distribution, procurement, and
inventory control
 For each category a solution approach is
developed => Management Science
MANAGEMENT SCIENCE
APPROACH
 Also referred to as Operation Research
 In solving problems, managers should follow the
five-step MS approach
1. Define the problem
2. Classify the problem into a standard category (*)
3. Construct a model that describes the real-world
problem
4. Identify possible solutions to the modeled problem
and evaluate the solutions
5. Compare, choose, and recommend a potential
solution to the problem
WHY ARE STRUCTURED DECISIONS ALSO CALLED
PROGRAMMABLE DECISIONS?
 In structured decisions, all of the factors
affecting the decision and its outcome are
known and it is also known how to produce a
given outcome given the relevant
information.
 In this way, such decisions could be
converted to a simple programme of actions
that a computer could execute.
 Hence structured decisions are called
programmable.
AUTOMATED DECISION MAKING
 A relatively new approach to supporting decision
making
 Applies to highly structures decisions
 Automated decision systems (ADS)
(or decision automation systems)
 An ADS is a rule-based system that provides a
solution to a repetitive managerial problem in a
specific area
 e.g., simple-loan approval system
AUTOMATED DECISION MAKING
 ADS initially appeared in the airline industry
called revenue (or yield) management (or
revenue optimization) systems
 dynamically price tickets based on actual demand
 Today, many service industries use similar
pricing models
 ADS are driven by business rules!
AUTOMATED DECISION-MAKING
FRAMEWORK
 A good example of structured decision could be
calculation of tax.
 Indeed, the goal is clearly defined — the amount of
money to be taxed.
 If the profit and the circumstances of how it was
obtained are known, then information is obtainable
and manageable.
 There are laws that specify precisely how to pay tax.
 Thus, the procedures are known.
 The context is well–defined as the problem is not
unique (most people and organisations
pay taxes).
 COMPUTER SUPPORT FOR
UNSTRUCTURED DECISIONS
 Unstructured problems can be only partially
supported by standard computerized quantitative
methods
 They often require customized solutions
 They benefit from data and information
 Intuition and judgment may play a role
 Computerized communication and collaboration
technologies along with knowledge management
is often used
 For an example of unstructured decision we may
consider a fashion clothes company that is
looking to design a new product line.
 Although the goal is set, it is uncertain what
will be in fashion by the time the product line is
ready.
 The outcome is uncertain. Although it is possible
to see some repetitions in fashion over time,
the new product is supposed to be original.
 Thus, the problem appears in unique context. It
is hard to assess what is required to design a
product that will be successful.
COMPUTER SUPPORT FOR
SEMI-STRUCTURED PROBLEMS
 Solving semi-structured problems may involve a
combination of standard solution procedures and
human judgment
 MS handles the structured parts while DSS deals
with the unstructured parts
 With proper data and information, a range of
alternative solutions, along with their potential
impacts
 GORRY AND SCOTT MORTON’S FRAMEWORK FOR
DECISION SUPPORT (1971)
Control Operational Managerial Strategic Planning Technology
Control Control Support Needed
Decision
Structured e.g. e.g. short- e.g. financial MIS,
accounts term management mathematical
receivable, forecasting models,
order entry transaction
processing
Semi-structured e.g. e.g. credit e.g. mergers DSS
production evaluation and
scheduling acquisitions
Unstructured e.g. e.g. e.g. new DSS, ES,
selecting a recruiting technology Neural
cover for a an development Networks
magazine executive
IMS3001 – BUSINESS INTELLIGENCE SYSTEMS – SEM 1 , 2004
Technology MIS, Manageme
16 EIS, ES,
 A(n) ________ is a rule-based system that
provides a solution, usually in one functional
area, such as finance or manufacturing, to a
specific repetitive managerial problem.
 A) decision support systems (DSS)
 B) transaction processing system (TPS)
 C) online analytical processing (OLAP) system
 D) automated decision system (ADS)
Solving semistructured problems may involve a
combination of both standard solution
procedures and human judgment.
Management science can provide models for
the portion of the decision-making problem
that is ________.
 A) controlled
 B) semistructured
 C) unstructured
 D) structured
 Gorry and Scott-Morton suggested, for
example, that for semi structured decisions
and unstructured decisions, conventional
________ and management science (MS)
tools are insufficient.

 Structured problems, which are encountered

repeatedly, have a high level of ________.
 ________ is the application of scientific
approach and mathematical models to the
analysis and solution of managerial decision
situations.
 Automated Decision Systems (ADS) is a
business ________ system that uses
intelligence to recommend solution to
repetitive decisions (such as pricing).
 Accounts receivable and accounts payable
processes are highly structured, which makes
them easy to program.