Page 39 -

THE MODELLING PROCESS 19

common quantitative models used in management science. However, it must be
appreciated that such models and the process of creating suitable models is part of
the wider management science methodology that we discussed in Section 1.4.
Management science models are not plug-and-play solutions to management problems
(although some organizations do see, and use, them this way.) That is, it is not simply a
situation of choosing a model, plugging it into the problem and finding a solution.
Rather, model building in management science is both a science and an art. The
science comes from knowing what models are available, how they are typically con-
structed and used and what their limitations are. The art relatestothe processofadapting
the model to the business problem being examined – making the model fit the problem
situation as well as it can and also appreciating where some of this fit is less good. It must
be remembered that any model is a simplified version of reality – we are not trying to
capture the problem situation in all its complexity but rather simplify the problem down
to its key elements so that we can more easily make sense of it and better analyze it.
The modelling process typically consists of a number of iterative stages. Initial
model selection involves the management scientist identifying which model, or
models, seem best suited for the problem. This typically follows the Problem
Structuring stage of the overall methodology outlined earlier in Figure 1.2. Obvi-
ously, this assumes that the management scientist is aware of the different models
available. Clearly, this is one of the purposes of this text – to help you become aware
of the different quantitative models available. However, for the management science
practitioner, this stage is less obvious than it first appears. Management Science, like
all other academic disciplines, is constantly changing with new models and techni-
ques being developed. It is also worth realizing that in practice more than one model
may be used. For example, in order to build and use a revenue model we might first
need to build a forecasting model to forecast consumer reaction to price and volume
changes. Following initial model selection we then typically get involved in data
collection as the next stage. This will involve searching for and collecting the data needed
by the model we have decided to use. Different models have different data require-
ments and to some extent the availability of appropriate data may restrict the choice
of model. As we shall see, some models require a lot of accurate and reliable data –
they’re often referred to as data-hungry models. If this isn’t available, the management
scientist may have to choose another model which has fewer data requirements (or set
out to collect the data that the first-choice model needs if time and budget permit).
Assuming appropriate data is available, the next stage is model construction – building
an appropriate model for the problem. Once again, in practice, this is more difficult
than it seems. Any model is a simplified version of reality – in other words there are
certain aspects of the problem that we conveniently push to one side in order to build a
simpler picture of the problem situation we face. This often requires the modeller to
make certain assumptions and these assumptions can make all the difference between a
good model and a bad one. Sometimes these assumptions may be explicitly stated. In
other cases they may be implied. If we return to the Nowlin breakeven model that we
built in Section 1.5 there are no explicit assumptions stated. However, there are certain
assumptions implied in the model. These include:
l We assume that the data used – such as fixed cost and variable cost – is known
for certain, is accurate and is fixed and constant.
l We assume that customers will continue to buy the product at E5 no matter
how many we sell and no matter what our competitors might do.

Such assumptions may be necessary to allow us to build a suitable model but they
may not always be reliable assumptions – or rather they may be reliable only under
certain limited conditions. The assumptions made may affect the reliability and

Copyright 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has
deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.