Functions of DSS

There are five functions of a DSS facilitating managerial decision making. They are • Model building

• 'What if' analysis

• Goal seeking

Srinivas MBA : MIS : CH 5  2009 

136

• Risk analysis

• Graphical analysis.

(a) Model building: This allows decision makers to identify the most appropriate model for solving the problem on hand. It takes into account input variables, interrelationships among the variables, problem assumptions and constraints.

For example, a marketing manager of a television manufacturing company is charged with the responsibility of developing a sales forecasting model for color TV sets. A model builder uses a structured framework to identify variables like demand, cost and profit, analyze the relationships among these variables, identify the assumptions, if any (e.g., assume the prices of raw materials will increase by 5% over the forecasting period), and identify the constraints like the production capacity of the plant. All this information is then integrated by a system into a decision making model, which can be updated and modified whenever required.

(b) 'What-if' analysis : This is the process of assessing the impact of changes to model variables, the values of the variables, or the interrelationships among variables. This helps managers to be proactive, rather than reactive, in their decision making. This analysis is critical for semi-structured and unstructured problems because the data necessary to make such decisions are often either not available or incomplete. Hence, managers normally use their intuition and judgment in predicting the long-term implications of their decisions. Managers can prepare themselves to face a dynamic business environment by developing a group of scenarios (best-case scenario, worst-case scenario and realistic scenario).

(c) Goal seeking : It is the process of determining the input values required to achieve a certain goal. For example, house buyers determine the monthly payment they can afford (say for example, Rs. 5,000) and calculate the number of such payments required to pay the desired house.

(d) Risk analysis : It is a function of DSS that allows managers to assess the risks associated with various alternatives. Decisions can be classified as low risk, medium risk, and high risk. A DSS is particularly useful in medium risk and high risk environments.

(e) Graphical analysis : This helps managers to quickly digest large volumes of data and visualize the impacts of various courses of action. S L Jarvenpaa and G W Dickson studied the relative advantages and disadvantages of tabular and graphic output. They recommended the use of graphs when:

• Seeking a quick summary of data

• Detecting trends over time

• Comparing points and patterns at different variables

• Forecasting activities

• Seeking relatively simple impressions from a vast amount of information.

The researchers suggested that a tabular presentation be used when it is necessary to read individual data values.

Jarvenpaa and Dickson has also offered the following tips when choosing between the various types of graphs:

Srinivas MBA : MIS : CH 5  2009 

137

• Line or bar charts are preferred for summarizing data.

• Grouped line or bar charts are good for showing trends over time.

• Grouped bar charts are better than pie charts for presenting parts of a whole.

• Grouped line or bar charts are good for comparing patterns of variables.

• Use horizontal rather than vertical bars when comparing variables.

• Use single line or bar charts to compare individual data points between variables

• Put data values on the top of the bars in a bar chart for easier reading.

Note :

Types of decisions

There are three types of decisions

• Programmed decisions

• Non programmed decisions

• Semi programmed decisions

(1) Programmed decisions : The decisions in which a problem is solved by a predefined procedure or algorithm. These decisions are repetitive and routine in nature and are capable of being modeled mathematically in their entirety.

The examples of such decisions are :

(i) preparation of pay in accordance with the laid out regulations (II) inventory ordering.

To arrive at the programmed decisions, a solution manual to problems is prepared to help the users. Some characteristics of such decisions are :

(i) These decisions can be delegated

(ii) The cost of solving a problem is low compared to non-programmed rules (iii) Such decisions can be made with the help of the computer system.

(2) Non programmed decisions : These decisions are unstructured, occasional, of high consequence, complex and involve major commitments. There is no predefined program or set decision rule or algorithm available to solve these problems automatically.

The examples of such decisions are : (i) advertising budget

(ii) new product decisions

(iii) acquisition of capital projects

Some characteristics of such decisions are

(i) These decisions are novel and difficult to structure in logical mathematical terms.

(ii) These decisions cannot be delegated, and are based on management direction, thinking and deliberations, e.g., purchase of scarce and capital items under fluctuating price conditions cannot be delegated

(iii) Computers cannot be used directly for such decisions. However, they may be used to process large volumes of necessary data.

(3) Semi programmed decisions : In these types of decision, at least one but not more than two of the above stages can be handled by a well defined preset procedure. An example of such a decision is the intelligence phase, which is well structured, having diverse kinds of variance analysis. Here a comparison with a budget is undertaken in a well defined way to indicate the need for a decision. Subsequent stages of design and choice are, however, not handled by any set procedure.

Srinivas MBA : MIS : CH 5  2009 

138 Assignments :

1. What is decision making ? What are the types of decisions ?

2. What do you mean by decision support system ? What are their characteristics, types, components and functions ?