Decision Making

UNIT 1 DECISION MAKING

Objectives
After going through this unit, you should be able to
• appreciate the nature and variety of managerial decisions
• develop decision table logic for structured and programmed decisions
• understand the decision making process
• understand the relevance of various models of individual and organisational
decision making
• give examples of how information systems support various stages of decision
making.
Structure
1.1 Introduction to Decision Making
1.2 Structured Decisions
1.2.1 Decision Tables for Structured/Programmed Decisions
1.3 Unstructured Decisions
1.4 Additional Categories
1.4.1 Departmental. Inter-Departmental and Enterprise Decisions
1.4.2 Organisational and Personal Decisions
1.4.3 Individual and Group Decision Making
1.5 Decision Making Process
1.5.1 Simon's Decision Making Model
1.5.2 Massie's Decision Making Model
1.6 Information Requirements for Different Decision Making Stages
1.6.1 Information for Intelligence Phase
1.6.2 Information for the; Design Phase
1.6.3 Information for the Choice Phase
1.6.4 Information for the Implementation Phase
1.7 Rational Individual Models of Decision Making
1.7.1 Rational Model
1.7.2 Bounded Rationality and Satisficing
1.7.3 Muddling through Model
1.7.4 Psychological Types and Frames of Reference
1.7.5 Implications of the Models for Information Decisions
1.8 Organisational Models for Decision Making
1.8.1 Implications for Information System Design
1.9 Summary
1.10 Key Words
1.11 Self-assessment Exercises
1.12 Further Readings

1.1 INTRODUCTION TO DECISION MAKING

The field of decision making is vast. There have been many approaches to managerial
decision making.These have ranged from the strictly quantitative as typified by the
methods of operations research - to those based on human and organisational
behaviour. It is only recently that those working in the area of decision making have
started to combine approaches that could cater to the multiplicity of subjective and
objective factors and the multiplicity of criteria and objectives- some in conflict with
one another.
Decision making is an essential part of management. Some have even suggested that
management is synonymous with decision making. Managers are decision makers and 5
problem solvers. Whether a manager is involved in evaluating new opportunities or
Information for Decision eliminating long standing difficulties, decision making for management is essentially
Making problem solving. The process of deciding is intimately related to the whole process of
knowing (or or 'cognition'). Knowing "what the problem is" will assist in deciding "what
to do about it", that is finding the best solution. This is basic for each individual
Personally and also for a professional manager, whose main orientation is Upwards the
making of decisions. There' is always an implicit opportunity loss associated with poor
decisions. They are the foregone gains, profits or cost savings, which could have been
realised had a better decision been made.
However, the process by which one arrives at a decision is quite complex; in fact no
one process can he applied to all decisions. Over the decades, decision-making
methods have evolved from primitive to supersophisticated, ranging from the
instinctive/intuitive approach, to traditional precedent based approach, to the
commonsense approach, to the scientific method. A decision involves many intuitive
and deep-seated cognitive mechanisms that cannot be observed fully or directly
influenced. What can be influenced are the behaviour patterns, the analytical
procedures and the sequence of logic that are followed in making a decision. Ideals,
objectives and goals form the background against which decisions are made.
There are many types of decisions, some that are completely specified, some partly
specified and many unspecified-We shall discuss how computers have entered in
decision aiding process, who all are the end users of the systems, the different types of
decisions, the decision making process and the implications for the information
systems analyst. It would be our effort to minimise repetition of concepts or subjects
that we have already discussed in previous courses or modules. These, however, are
important and should be referred to for better understanding and appreciation.
Let us look at the different types of decisions that are encountered by managers. We
would like to categorise tnem primarily into two categories, structured and
unstructured.

1.2 STRUCTURED DECISIONS

These structured decisions are those that can be programmed. They are essentially
repetitive, routine and involve a definite procedure for handling them so that they do
not have to be treated each time as if they were new.
It has been seen that in general at the lowest level in the organisation viz., the
operational level and the managerial staff, deal mostly with such fairly well structured
problems. In the past most successful applications of information systems have come
in dealing with structured, operational and management control decisions. These fall
really in the domain of transaction processing and it is for this reason that some
researchers prefer to call such systems as TPS - Transaction Processing Systems,
rather than MIS - Management Information Systems.
Thus structured decisions which are also called programmable decisions involve
situations where the procedures to follow when a decision is needed can be specified
in advance. Therefore, such decisions are structured or programmed by the decision
procedures or decision rules developed for them. A structured decision could possibly
involve what is known as a deterministic decision or an algorithmic (step by step
procedure) decision. In this case, the outcome of a decision can be determined with
certainty if a specified sequence of activities viz., the algorithm, is performed. No
doubt many decision situations do not all come under deterministic criteria. There
might be probabilistic decision situations: Since, in this- case, enough probabilities
about possible outcomes, are known, a decision can be statistically taken or
determined with an acceptable probability of success.
Herbert A. Simon stresses the programmable part of the structured decisions and
accordingly terms `structured decisions' as `programmed decisions'. In the programmed
decisions, a problem (or one very similar) occurs frequently enough so that a fixed routine
or programme is established for solving it. Programmed decisions are in fact those that are
made in accordance with some policy, rule or procedure so that they do not have to be
handled de novo each time they occur. In fact these decisions are generally repetitive and
routine and are obviously the easiest for managers to make. It is for these reasons that
6 such managerial problems are relegated to the supervisory level. The supervisors fall
in the first entry ring of management. It gives persons of this level the slight `kick'
 Decision Making
or morale boost as they have entered into the 'management cateogry'. No doubt the
perceptions will have to change since they have moved across from the
worker/operator level to the other side of the table. Decisions implemented by the
supervisors might feel elated, but when compared to managers at the higher levels,
the supervisory level decisions are pretty straight forward and simple because of
their structured nature. These decisions vote in fact routine'decisions and they
require little deliberation from the top man's point of view.
These routine decisions are taken against a familiar background in everyday business
operations. Such decisions involve no extraordinary judgment, analysis or authority,
since they are to traverse through more or less fixed avenues. On account of the
experience gained or because of the trivial nature of the problems on hand,
management has already established a set of rules, policies and procedures. With the
organisation's goals, policies and processes established, the routine decisions
demand, on the part of managers, a power of selection of the best pathe, as the
connecting link between the given means and the established ends.
The structured decisions, often termed as programmed decisions, are labels that are
derived from the jargon of the computer field, where a program is defined as a plan for
the automatic solution of a problem. Programs are simply a string of instructions to
accomplish an assignment. However, it is well known that all problems do not lend
themselves to automatic programmed solutions. No doubt an informatlon system analyst
might be having his cherished dream of having all decisions in an organisation
programmed or fully automatic. Then perhaps there would be no dividing line as regards
variety of type of skills required by managers at various levels, You would recollect that
studies have shown that more technical skills are required at the lower levels
whereas more conceptual skills are required by managers at the higher levels, apart
from human skills that are required at all levels of management.
It has been seen that by far the greatest number of business decisions are repetitive
and routine ones. If this is true, then there is an overriding need to automate or '
program these decisions so that managers and executives could delegate such
problems to lower levels and have them made by one or more techniques of
programmed decisions. It is interesting to note that in some cases even up to 90% of
management decisions are routine ones.
Most programmed decisions are solvable by the `force of habit'. We go to our
offices, make decisions about the in-basket or in-tray correspondence, and take
dozens of actions daily that are "programmed" through sheer force of habit. In fact
this is gained by `experience' in an organisation. Whenever there are turnovers, it
is this valuable experience that is lost, which Could be quite costly. It usually takes
time and money for newer recruits to acquire this experience.
The `force of habit' in fact leads to certain traditions, conventions and practices
which might become a platform towards formal statement of rules, procedures,
policies, etc.
At this stage, it would be worthwhile to spend some time looking at what constitutes
a rule, procedure or a policy.
A rule is a specific policy statement about the conduct of certain affairs. For instance, a
rule may state that any person who comes in late to work will face disciplinary
action. It could he that if an employee is to start his office hours from 9.00 a.m., if he
comes by 9.15 a,m., he could be tolerated. Time between 9.15 am. to 9.30 a.m. could be
allowed at best say thrice in a month. Anything beyond these limits would
necessitate some disciplinary action. A rule is frequently used when confronting a
well structured problem. Rules are usually rigid which might specifically tell about
`dos and donts'. Rules are usually framed in a manner to take care of almost all
situations. However, we do have instances where the rule applies - `show me the man 1'11
show you the rule'. You often hear the term that `rules are silent on this issue'.
Indeed rules are quoted again and again depending on what is to be proved or
justified or rationalised in order to take the final decision.
Activity A

Could you enumerate the rules followed in your organisation by you in seeking a
temporary replacement for three days for an absent typing clerk. 7
Information for Decision …………………………………………………………………………………………
Making ………………………………………………………………………………………….
………………………………………………………………………………………….

A procedure is a series of steps that are sequential in nature and interrelated to

address a well structured problem. A procedure is a kind of direction in using a series of
steps in either diagnosing a problem or solving it. For instance, if your car is not
working, the mechanic has a manual of procedures to check the trouble and follow
procedures to correct it. He might like to first find out whether there is petrol in the
tank, If petrol is there, he might like to check the battery terminals to see whether they
are loose. With experience it is possible to have shortcuts in diagnosis, which might
sometimes appear to be done on a random basis or just on atrial and error' basis. Now
the company procedure could be written, oral or implied. Standard operating
procedures provide a means for indoctrinating and training new personnel and for
guiding experienced personnel in the performance of specific tasks. The procedure has
the additional advantage of forcing a certain amount of detailed planning, because it
cannot be adequately designed, reviewed, or implemented without careful thought.

Activity B

What would be the procedure followed by you to mend a leaking tap in your bathroom?

…………………………………………………………………………………………
…………………………………………………………………………………………
…………………………………………………………………………………………
…………………………………………………………………………………………
…………………………………………………………………………………………

As for a policy, it is a general guideline which sets up parameters for the judgment
within which .to operate and is general and judgemental in nature. For instance; the
policy of a company may be `the customer is always right'. But it is left to the manager
to rightfully interpret this policy in a particular situation. It is indeed possible that
there could be a bias, even perhaps an emotional outburst and sometimes an immature
and unrealistic response of the customer. Should the policy be implemented in true
letter and spirit? Should one think of the policy in the long-run or short-run? Now if
the case is to be considered under the structured programmed decision category, some,
definite answers would have to be provided rather than leaving it ambiguous, vague
and fuzzy.

1.2.1 Decision Tables for Structured/Programmed Decisions

Let us demonstrate the use of a decision table in the context of a programmed or

structured decision. Consider an accounts receivable process in which customers
accounts are examined with the purpose of producing a statement and a possible
reminder, of variable severity, for each account.

The analyst's first step is to decide upon the set of criteria applicable. By discussing
with various relevant groups of people, three unrelated possibilities, are discovered (1)
that within 30 days the amount exceeds Rs.5,000/- (2) that within 60 days the amount
exceeds Rs.2,000/- and'(3) there is still an amount to be paid on goods purchased more
than.60 days ago. Let us denote these condition stubs as AR1, AR2, respectively.

The possible answers to each of these questions is a sample yes (Y) or no (N). thus there
are a maximum of 23 = 8 rules corresponding to the possible combinations of answers.
These are placed in the condition entry section of the table, one per decision rule
column. There is no significance attached to the order in which the rules are written,
provided that all possible entries are recorded. However , it is sometimes possible to
combine to or more rules if it is known that the ensuring actions are all identical

8 By further questioning, the analyst establishes the complete list of all possible actions
and writes them in action stub, one per line. Suppose that, the time this section
 Decision Making
of the system is `activated' a prelitpinary operation has bypassed all those customers not
needing a statement, and has already produced a statement for those to whom one is
d ue . The only task remaining is to produce one or more of:
• A moderate reminder (Letter A)
• A more severe reminder (Letter B)
• A very Terse Post Script to either letter (the same in each case
• A special report on the customer for the sales manager (called Exception
Report).
Now the REPORT will no doubt be produced on some unit other than the one which
wilt print the letters and may consequently be produced anywhere in the action
sequence.
A terminal action (which leads on to the next operation to be performed) could also be
included (i.e., GO TO xxx where xxx is the name of some other decision variable).
Figure 1.1 gives the completed decision table for the programmed decision situation of
accounts-receivable.
CONDITION STUB DECISION
RULES
1 2 3 4 5 6 7 8
AR1 Rs. 5,000/- N Y N Y N Y N Y

AR1 Rs. 2,000/ N N Y Y N N Y Y

AR1 Rs. 0/- N N N N Y Y Y Y

ACTION STUB
X X
Letter A
(Moderate
Reminder)

Letter B (Server X X X X
Reminder)

Terse Post Script X X

Exception Report X

GOT TO NEXT X X X X X X X X

Figure 1.1 Accounts Receivable Programmed Decision Table

The above decision table can be used for writing a program in an appropriate.
computer language.
It is worthwhile to note the orderly nature of the inquiries into which the analyst is led.
In fact this is a most valuable discipline tending to regularise the otherwise random
questioning which may occur.
Having expressed the table in as concise a form as possible, a programmer can now
proceed directly with the `encoding' process, or use the table to construct a formal
flowchart, which may then serve as a basis for coding. There are available translation
programs which take a decision table as input, and produce as output a program
written in a language such as FORTRAN, COBOL or PL/I, or compiled code.
Example of this is the Rand Corporation's FORTAB and DBM's Decision Log
Translator, both of which produce FORTRAN programs. The DET4AB/65 produces a
COBOL program.
The accounts receivable structured example draws heavily from the illustration
provided by Brookes et al. One can refer to a large number of problem situations that
are amenable to the process just illustrated. Primarily the analyst would have to
understand the problem on hand by being able to determine a set of operating
conditions and the amenable decisions to take by following the set of alternative rules
that can he drawn up. 9
Information for Decision Decision tables are quite graphical in nature and facilitate communications between
Making user, analyst supervisors and affected parties. As we have seen, the decision table
expresses primarily a series of conditions; when the conditions are fulfilled, then a
rule associated with the condition is executed. A `header' is used to identify the table
and condition stubs describe the various conditions. As already explained earlier, a
rule is a procedure for checking the different conditions, and the action statement tells
what action to take when a rule is true. The table is read until the conditions for a rule
are met and the action described is taken. Then the next scan of the table begins.

Now to familiarise you with the topic a bit more, we illustrate two decision tables as
shown in Figure 1.2 and Figure 1.3 for the 'limited entry' and `extended entry' decision
table examples. The illustrations pertain to the logic for a `credit card' purchase
authorisation. In this example, a purchase under Rs.500/- is approved automatically.
Purchases between Rs.500/- and Rs.1,000/- are given authorisation number. Finally
for purchases over Rs.1,000/ we give an authorisation number and place a "hold" on
the customer's account for the amount of purchase. The decision tables shown in
Figure 1.2 and Figure 1.3 are self-explanatory.

CREDIT CARD AUTHORIZATION

CONDITION STUB DECISION RULES

2 3 4

Is purchase < Rs. 500/- Y N N. N

Is purchase between Y N N

Rs: $001 and Rs 1000/-

I5 purchase over Rs. 1000/. Y N

ACTION STUB X

Approve with no action

Give authorization no X X

Place hold on account X

Error X

Figure 1.2: Limited Entry Example

CREDIT CARD AUTHORIZATION

CONDITION STUB DECISION RULES

Is purchase ‘P” 1 2 3
P > Rs. 1000/- 500 < P < 1000 O < P < 5000

Approve with no action X

Give authorization X X

Place hold on account X

10 Figure 1.3: Extended Entry Example

 Decision Making
As can be seen the extended entry is somewhat more compact and allows us to
uselogical conditions as entries and save space. No doubt the tables shown in and
1.3 adequately describe the logic for the credit card example and that both the tables
are equivalent.
Care must be taken by the analyst that the rules themselves must be unique and
independent; they cannot and should not contradict one another; and only one rule can
apply in a given situation. It really does not matter in what sequence rules are
presented, since only one set of conditions can be satisfied at a time.
It would be interesting to make an observation here at this point. Many knowledge
based expert systems are also `rule-based'. They utilise the if-then' type of logic. If °a
set of certain conditions holds', then `the action or outcome will be...'

1.3 UNSTRUCTURED DECISIONS

Thus far we have been discussing the structured programmable decisions which are
very large in number and perhaps more easy to handle as compared to the
unstructured or relatively less structured decisions which we would now like to
discuss. Though fewer in number as compared to the structured situations, this
category of decisions is more repetitive in nature, usually `one-shot' occurrences for
which standard responses are usually not available. Hence they require a creative
process of problem solving which is specially tailored to meet the requirement of the
situation on hand. In fact managers at higher levels in an organisation are usually
faced with more such unstructured decision making situations. Some have aptly
described the situations as somewhat `strategic' in nature as compared to the `tactical
orientation of the structured decisions at the lower levels of management.
strategic decisions are non-repetitive, vital and important and aim at determining or
changing the ends or means of the enterprise.
Since each manager, in the case of such unstructured, non-programmed decisions, may
bring his own personal beliefs, attitudes and value judgments to bear on the;..''
decision process, it is possible for two managers to reach distinctly different solutions,
to the same problem, each claiming that he is acting rationally. In fact the ability to
make good non-programmed decisions helps to distinguish effective managers from
ineffective managers. Unstructured decisions are not simple. They are usually quite
complex in nature. We can't shy away from them for someone has to ultimately make
these decisions even though there are inherent, dangers when confronting ill-
structured problems. Non-programmed decisions are essentially new and unique.
They have to be often solved de novo: There is no tried and true method of handling
them. Unstructured decisions are-those in which the decision maker must provide
judgment, evaluation and insights into the problem definition.

1.4 ADDITIONAL CATEGORIES

There are many ways of categorising decisions as proposed by various persons from
time to time. We have already discussed structured and unstructured situations. Some
have termed them as programmed/programmable and non-programmed/non-
programmable decisions. Yet others have talked of routine/repetitive and non-
routine/non-repetitive decisions.
Some have stated the range of tactical and strategic decisions; others have termed
them as minor and major decisions. We can easily have just two classes of decisions
as follows
Class I : Unstructured, Non-programmable, Strategic, Major, Routine;
Repetitive, Complex, Long-run,

Class II : Structured, Programmed; tactical, Minor, Non-routine; Non-

repetitive, ,,Simple, Short-run.

1.4.1 Departmental, Inter-departmental and' Enterprise Decisions

Let us discuss decisions that could he departmental, inter-departmental and 11

enterprise decisions. For instance, sanctioning leave to an employee is a departmental
Information for Decision decision, but making a slight change in the design of the product is an inter-
Making departmental decision, and entering a new line of business is an enterprise decision,
The department level decision is strategic in nature whereas departmental level is
tactical in nature.
1.4.2 Organisational and Personal Decisions
When an executive acts formally in his expected role in an organisation, he makes
organisational decisions. However, when the manager takes a decision in his personal
capacity and not as. a member of the organisation, it is known, as a personal decision.
For instance, transfers that are effected by an executive are organisational decisions.
However, an executive who decides to leave the present organisation is making a
personal decision.
1.4.3 Individual and Group Decision Making
Decisions may be taken either by an individual or by a group or a Committee. It is
difficult to say which a better method of taking decisions is. Each method has its
strengths and weaknesses. We are all aware of the delaying tactics that are adopted by
institution of committees in the resolution of conflicts. On the other hand, it is argued
that two blockheads are better than one. There would be richer ideas and many more
alternatives generated in the process. But when it comes to assessment of selection,
there could be `lot of heat' generated. How does one resolve such a situation? Is it by
consensus? or by voting?
i) Individual Decisions
According to Simon, "It is impossible for the behavior it of a single isolated individual
to reach any high degree of rationality. The number of alternatives he must explore is
so great, the information he would need to evaluate them so vast, that even an
approximation to objective rationality is hard to conceive. The individual decision
maker is an individual human being - the one factor most vital and most difficult to
understand because of various factors such as age, perception, intelligence, experience
in a given area, confidence in decision making, time available, resource position,
upbringing, family background, and so on that could come into play.
The most mysterious factor is still this decision making individual human being. How
and why an individual acts in a certain way at a certain time. is still quite a mystery.
The individual decision maker could have quite an amount of prejudice and bias that
is inherent on account of perceptual processes which act as great filters. We only
accept what we want to accept and hence only such information filters down to our
senses, and secondly, the perception is highly subjective. The information gets
distorted to coincide with our pre-established beliefs, attitudes and values.

These are additionally `cognitive' constraints. Psychologically we are always

uncomfortable with decision making. We are never sure if our choice of the
alternative was correct and optimum, until the impact of the implication of the
decision has been felt. This makes us feel very insecure and could be one of the many
causes of `stress' in individuals leading to hypertension and other health
complications.
No doubt it is important to have adequate and accurate information about the '
situation for good quality decision making. However, it must be recognised that "an
individual has constraints of nature such as physical, psychological, sociological etc.
These limit the amount of information the individual can handle".
ii) Group Decisions
As for group decisions, these are: usually taken for major issues in order to secure
wider cooperation, .cceptability and-coordination. Usually in a group like situation,
the chances of subjective errors are reduced and more options are thrown open. If the
group is larger, such decisions suffer from unnecessary delays, deadlocks and petty
party politics.
Some advantages of group decisions are

12 i) Increased acceptance by those affected and hence implementation is easier.

 Decision Making
ii) Easier Coordination,
iii) Easier Communication.
iv) More information processed on account of availability of a larger number
of specialists in the group.
v) Group decision making is more democratic in nature,
vi) Participative group process builds up a training ground for subordinates
and others which allows for smooth handover when individuals leave ±° e
scene (no vacuum is created).
Let us now enlist some of the disadvantages. of group decisions.
i) Group decisions take longer.
ii) Groups can be indecisive.
iii) Groups can compromise.
iv) Groups can be dominated.
v) Group members may exhibit "focus effect" viz., the group may just focus on
one or a few suggested alternatives and spend all the time in evaluating
these and may never come up with other ideas thus limiting the choices. '

1.5 D E C I S I O N MAKING P R O C E S S
1.5.1 Simon's Decision Making Model
Let us now look at the decision making process as proposed by Herbert A. Simon. His
model is a conceptual framework that divides the decision making process into the
following stages or phases:
i) Intelligence Activities: At this stage, a search of the environment takes
place to identify events and conditions requiring decisions. Data inputs are
obtained, processed and examined for clues that may identify problems or
opportunities,
ii) Design Facilities: At this stage, alternative courses of action arc developed,
analyzed and evaluated. This involves processes to understand the problem, to
generate solutions, and to test solutions for feasibility,
iii) Choice & Implementation Activities: Here one has to select an alternative
as course of action from those available. A choice is made, implemented and
monitored.
Though intelligence, design, choice and implementation activities are sequential in
nature, the decision making process includes the ability to cycle back to a previous
stage as shown in Figure 1.4. Choice and implementation have been shown to be
separated for better understanding,
Which
Is there a problem ? What are the should you Is there choice
alternatives? choose? working ?

Intelligence Design Choice Implementation

1.5.2 Massie's Decision Making Model

Many step-wise models of the rational decision making process have been proposed
over the years. One such proposal by Massie has been found to be most
representative as a five stage procedure as follows:
i) Understand situation
ii) Diagnose and define problem
iii) Find alternatives
iv) Select action and 13
v) Secure acceptance of decision
Information for Decision Figure 1.5 captures the gamut of the decision making process and is quite self-
Making explanatory. You would have by now seen the similarities in the two decision making
processes that we have just discussed. However, Simon's model seems to be more
preferred to in literature and as such you would find some of the material in this unit
also leaning on this model, when discussing on the implications and requirements of the
various stages of the decision making process by information system analysts.

We shall not get into concepts of data and information which have already been
discussed in MS-7 (Information Management and Computers) Unit No.16 on
Information Needs and its economics. Suffice it to say that information is the vital
resource for managerial decision making.

14
 Decision Making
1.6 INFORMATIONREQUIREMENTS FOR DIFFERENT
DECISION MAKING STAGES
Let us now look into the information requirements for the various stages described in tne
earlier section

1.6.1 Information for Intelligence Phase

Information Systems can help in the intelligence stage by providing information about
external and internal conditions. Intelligence entails scanning the environment, either
intermittantly or continuously, depending on the situation to identify potential
decision situations. For example:
a) a marketing executive makes periodic visits to key customers to review
possible problems and identify new customer needs;
b) Sales analysis reports can be furnished to managers periodically, when
exceptional sales situations occur, or on demand. These help managers
toidentify the status of sales performance. Also information from market
research studies and external. databases could also help managers (at higher
levels) identify changes in consumer preferences or competitive activity;
c) A design. engineer might attend a trade show at the International Trade Fair to
observe new materials that may or could be incorporated in failure product
designs.
The Intelligence phase and its activities result in some type of dissatisfaction with
current state or alternatively aid in the identification of potential rewards for a new
state The intelligence phase is the `tricky' phase and. embraces the unstructured non-
programmed category of decisions that we have discussed earlier in this unit.
A major information system capability is needed at this stage. It should be possible to
provide situation-specific information to managers when they make ad-hoc inquiries
that could be unique and often unscheduled. All this should be in addition to some of
the exception reports that might be churned out on a weekly basis.
1.6.2 Information for the Design Phase

The design phase involves designing of several possible solutions to the problems and
evaluation of the alternate courses of action. Here more carefully specified and
directed information activities and capabilities focused on specific designs are
required. This stage calls for quite a deal of creativity and innovation. Idea generation
and idea engineering could play a useful role in this stage of decision making.
Techniques like Brainstorming, Nominal Group Technique etc.; could be utilised.
Here also some decisions might fall in the category of highly structured, programmed
'situations and move towards semi-structured or unstructured non-programmed
decision situations. We have already discussed the use of decision tables for
structured situations. Decision trees could also be utilized by laying out graphically
the alternative courses of action that are within the control of the decision maker.

As for the semi-structured or unstructured decisions, decision support systems

(DSS) and expert system (ES) can provide assistance to managers. Models of
business operations can be developed with decision support software., including
advanced statistical, management science, and modeling packages, or less complex
spreadsheet programs. These packages and modules can then be used to manipulate
information collected in the intelligence stage to develop and evaluate a variety of
alternatives.

1.6.3 Information for the Choice Phase

In this stage a final selection of a particular course of action has to be made out of the
various alternatives generated in the preceding design stage. Here a manager can use 15
information tools that can calculate and keep track of the consequences, costs and
Information for Decision opportunities provided by each alternative designed in the previous stage.
Making
Information systems should help managers select a proper course of action. An
information system is most effective if the results of design are presented in a
decision-impelling format. The final c h o i c e would depend whether there is a
single criteria or objective on which it is to be decided or whether the decision
situation is one which involves multiple criteria and objectives. Most decision making
situations fall under the multiple criteria category, which is more difficult and
complex and operations researchers are just about getting into this field. The manager
is hardly an optimiser now; he believes in satisfying. Worse still, one has to account
for factors both subjective and objective, quantifiable and non-quantifiable, tangible
and intangible. A technique called Analytic Hierarchic Process (HP) developed by
Saaty is finding a great deal of application in such situations and seems to be
becoming a great boon for decision makers. An `expert choice' software is available
for the purpose of prioritisation of alternatives.
The quality of the choice stage depends very much on the quality of inputs made from
the previous twol stages - intelligence and design phases. It is possible that the
manager, even though at the choice stage, might like to refer and return to the
previous stages and reopen the issues for more data or alternatives etc.
Information systems can help managers in the choice stage in various ways. Managers
can be provided with summarised and organised information emphasising major
points such as major assumptions, resource requirements and expected results of each
decision alternative. Some type of a `what-if simulation analysis could be established.

1.6.4 Information for the Implementation Phase

This is the final stage of the decision making process. It is concerned with
implementing and monitoring. When the choice is made in the previous stage, the role
of the system changes to the, collection of data for further feedback and assessment.
The information systems must help managers monitor the successful implementation
of the decision. Here managers can use a reporting system that delivers routine reports
on the progress of a specific solution. Some of the difficulties that arise are resource
constraints, and possible ameliorating actions. Support systems can range from full-
blown Management Information Systems to much smaller systems and project planning
(PERT/CPM based) software operating on micro-computers. Feedback about business
operations affected by a decision helps a manager assess the decision's success or
failure, and whether follow-up decisions are needed.

1.7 Rational individual models of decision making

Because organisations are made up of individuals, it makes sense to build information
systems that facilitate individual decision making, wherever possible. Different
assumptions have been made while proposing some model or the other.

1 . 7 . 1 Rational Model

Some economists, mathematicians and management experts believe that the decisions
are always made rationally because the decision maker:
• knows his or her objectives and ranks them in order of importance;
• knows all possible alternative solutions to the decision problem;
• knows the relative pros and cons of each alternative; and
• chooses the alternative that maximises attainment of the objective.
Such idealistic assumptions have their share of criticism. Decision makers are not so.
well informed as to consider all the alternatives or to know all consequences.
Moreover, it is not just a singular objective that is pursued in real life situations. Still,
despite these criticisms, the rational model remains a powerful and attractive model of
16
human decision making.
 Decision Making
1.7.2 Bounded Rationality and Satisficing
Instead of searching for all of the alternatives and consequences (unlimited
rationality) as in the preceding model, people limit the search process to
sequentially ordered alternatives (alternatives that are not radically different from
the current policy. Wherever possible they avoid new, uncertain alternatives and
rely instead on tried-and-true rules, standard operating procedures) and programmes.
Individuals and/or organisations have multiple goals, some in conflict with each other
a l s o . Even goals are prioritised or placed in a hierarchy. In this way rationality is
bounded. Simon proposes the `satisficing' approach rather than the `optimising
approach'.
1.7.3 Muddling through Model
In 1959, Lindblom proposed the most radical departure from the rational model in his
article on the "science of muddling through". He described this method of decision
making, as one of "successive limited comparisons". Here values are chosen at the
same time as policies, and no easy means-end analysis is possible. For instance
labour and management can rarely agree on values, but they can agree on specific
policies. Because of the limits on human rationality, Lindblom proposes, incremental
decision making', or choosing policies most like the previous policy. Non-
incremental policies are apolitical (not likely to bring agreement among important
groups) and dangerous as nobody knows what they will lead to. Lastly, it is argued
that choices are not "made". Instead decision making involves a continuous process
in which final decisions are always being modified to accommodate changing
objectives, environments, value preferences and policy alternatives provided by
decision makers.
1 7.4 Psychological Types and Frames of Reference
The psychologists have given an additional perception to the rationality concept.
They say that humans differ in how they maximise their values as well as in using
the frame of reference for interpreting information and making choices. Here we
hear about the `cognitive style' that refers to underlying personality dispositions in
the treatment of information, the selection of alternatives and the evaluation of
consequences. Systematic thinkers impose order on perceptions and evaluation;
intuitive thinkers are more open to unexpected information and use multiple models
and perspectives when evaluating information. Neither is more rational than the
other.
Some studies have found that humans have a deep seated tendency to avoid risks.
when seeking gains but to accept risks in order to avoid losses. In other words, people
are more sensitive to negative outcomes than to positive ones.
1.7.5 Implications of the Models for Information Systems
Decision making is not a simple process and hence some guidelines need be kept in
mind when designing information systems. The following characteristics of
information systems could be encouraged :
− they are flexible and provide many options for handling data and evaluating
information.
− they are capable of supporting a variety of styles, skills and knowledge.
− they are capable of changing as humans learnand clarify their values.
− they are powerful in the sense of having multiple analytical and intuitive
models for the evaluation of data and the ability to keep track of many
alternatives and consequences.

1.8 ORGANISATIONAL MODELS FOR DECISION

MAKING
Just as we discussed rational individual decision making, it is useful to think also of 17
organisational decision making. The models of organisational choice have been
Information for Decision portrayed in Table 1.
Making
Table 1: Models of Organisational Choice

Name Basic Concept Inference Pattern

Rational factor Comprehensive Organisations select goal(s) examine all alternatives and
rationality then choose a policy that maximizes the preference
function.

Bureaucratic Organisational Goals are determined by resource constraints and

output existing human & capital resources. Standard operating
procedures are combined into programs, programs into
repertoires; these determine what policies will be
chosen. The primary purpose of the organisation is to
survive; uncertainty reduction is the principal goal.
Policies are chosen that are incrementally different from
the past.

Political Political outcome Organisational decision result from political

competition; key players are involved in a game of
influence, bargaining, and power. Organisational
outcomes are determined by the beliefs and goals of
players, their skills in playing the game, the resources
they bring to bear and the limits on their attention and
power.

Garbage can Non-adaptive Most organisations are non-adaptive, temporary and

organisational disappear over time. Organisational decisions result
programs from interactions among streams of problems, potentials
actions, participants and chance.

Source: Kenneth C. Laudon & J.P. Laudon, Management Information Systems: A Contemporary
Perspective. Collier Macmillan Pub. Co. U.K., 1998 Pg. 141 Table 5.8

1.8.1 Implications for Information System Design

As can be seen from Table 1, the designers of systems just can't think of individual
decision making models but would have to consider organisational decision making.
Systems must do more than merely promote decision making. They must also
include the notion of making individual managers better managers of existing
routines, better players in the bureaucratic struggle for control of an organisation's
agenda, and better political players. In fact also. Systems should help bring a
measure of power to those who can attach the right solution to the right problem.

1.9 SUMMARY
Management and decision making are complex activities that involve many
dimensions of human behaviour. Early classical models of management stressed the
functions of managers namely planning, organising, staffing, coordinating,
reporting, and budgeting. Depending on the level at which managerial decision
makers are, they perform a different mix of managerial functions. There are
primarily three levels of management and decision making termed as strategic,
technical/tactical and operational decision making.

In conclusion, management decision making quality depends on the vital input of

information so as to support the functions that a manager performs, the levels at
which the decision maker is, and on the type of decisions, whether structured or
otherwise.

1.10 KEY WORDS

Business Systems Analyst: A systems analyst tied to an end-user business area with
18 specialised understanding of the business information requirements of that
 Decision Making
functional area. The analyst is responsible for translating those specific business
requirements into information systems for that functional area.
Cognitive Styles : Basic patterns in how people handle information and control
problems.
Decision Making Process : A process of intelligence design, choice and
implementation of a particular course of action (Simon's Model).
Decision Table : A method of documenting decision rules in matrix or tabular form,
showing a set of conditions, and the actions that can be taken on these conditions.
Procedures : Set of instructions used by people to complete a task.
Structured Decision : Type of decision that is repetitive, routine and can be
structured and programmed.
Unstructured Decision : One in which-the decision maker must provide judgment,
evaluation, and insight because the decision problem is novel, ndn-routine, and has no
agreed-upon procedure for solving it and is usually unstructured and non-
programmable.

1.11 SELF-ASSESSMENT EXERCISES

1) Define structured and unstructured decisions. .Give four examples of each.
2) Discuss the rational individual models of decision making. What are the
implications of these models to information system analysts?
3) Describe the organisational choice models. How would the design of
systems be affected by the choice of model employed?
4) Prepare a decison table based on the information given below:
An educational institute wishes to make a statistical listing of all seniors. If the
seniors. If the senior is a male and a veteran, he will have a "V" printed after his
name. A male student will have his marital status printed.

If the student is a female and single, she will have her name and phone number
printed if she is over 20 and less than 26 years old. Single females under 20 will
only have their names printed. Single females 26 and over will have their date of
birth printed. All married females will have their marital status printed.

1.12 FURTHER READINGS

Brookes, C.H.P., Grouse, P.J., Jeffrey, D.R. and Lawrence, M.J., 1982. Information
Systems Design, Prentice Hall of Australia.
Chandan, J.S., 1987. Management Theory and Practice, Vikas Publishing House:
Delhi.
Gordon, D. and M. Olson, 1985. Management Information Systems: Conceptual
Foundations, Structure and Development, McGraw Hill: USA
Laudon, K.C. and J.P. Laudon, 1988. Management Information Systems: A
Contemporary Perspective, Collier Macmillan: U.K.
Lucas, Jr., H.C., 1982. Information Systems Concepts for Management, McGraw
Hill International: USA.
Murdick, R.G., Ross, J.E. and Claggett, J.R., 1987. Information Systems for
Modern Management, 3rd Edition, Prentice-Hall India Eastern Economy Edition.
O'Brien, J.A., 1991. Management Information Systems - A Managerial End User
Perspective, Galgotia Publications (P) Limited: Delhi.
Tripathi, P.C. and P.N. Reddy, 1983. Principles of Management, Tata McGraw
Hill: Delhi.
Voich, Jr., Mottice, H.J. and Shrode, W.A., 1975. Information Systems for
Operations and Management, South Western Publishing Company: USA. 19