Page 12 - DMGT209_QUANTITATIVE_TECHNIQUES_II
P. 12
Unit 1: Quantitative Techniques for Managers
Since statistical decision theory also uses probabilities (subjective or prior) in analysis, therefore Notes
it is also called a subjectivist approach. It is also known as Bayesian approach because Baye’s
theorem is used to revise prior probabilities in the light of additional information.
Self Assessment
Fill in the blanks:
3. ……………include all those devices of analysis and synthesis by means of which statistical
data are systematically collected and used to explain or describe a given phenomena.
4. The …………………are those which have already been collected by another organisation
and are available in the published form.
5. Statistical decision theory relies heavily not only upon the nature of the problem on hand,
but also upon the ………………
1.4 Models in Operations Research
In this Section we are presenting several classifications of OR models so that you should know
more about the role of models in decision-making:
Purpose
A Model is the representation of a system which, in turn, represents a specific part of reality
(an object of interest or subject of inquiry in real life). The means of representing a system may
be physical, graphic, schematic, analogy, mathematical, symbolic or a combination of these.
Through all these means, an attempt is made to abstract the essence of reality, which in turn, is
quite helpful to describe, explain and predict the behaviour of the system Thus, depending upon
the purpose, the stage at which the model is developed, models can be classified into four
categories.
1. Descriptive model: Such Models are used to describe the behaviour of a system based on
certain information.
Example: A model can be built to describe the behaviour of demand for an inventory
item for a stated period, by keeping the record of various demand levels and their respective
frequencies.
A descriptive model is used to display the problem situation more vividly including the
alternative choices to enable the decision-maker to evaluate results of each alternative
choice. However, such model does not select the best alternative.
2. Explanatory model: Such models are used to explain the behaviour of a system by
establishing relationships between its various components.
Example: A model can be built to explain variations in productivity by establishing
relationships among factors such as wages, promotion policy, education levels, etc.
3. Predictive model: Such models are used to predict the status of a system in the near future
based on data.
Example: A model can be built to predict stock prices (within an industry group), for
given any level of earnings per share.
LOVELY PROFESSIONAL UNIVERSITY 7