Information System Development Life Cycle Models - High Level Models and Methodologies for Information Systems - page 33

working prototype of the product, which is then implemented so that the users can
test it and provide real-time feedback and experience. If improvements and changes
are found necessary, the prototype is revised and re

ned, and a new prototype is
released and implemented for testing. This subcycle will go on until the product is
generally accepted by the users and no longer requires substantial changes or
updates, at which time the final version is released (Carr and Verner 1997 ).

There are various types of prototyping, according to speci c needs of the project.

These can be summarized in three categories: exploration, experimentation, and
evolution (Floyd 1984 ).
The exploratory approach is centered on the premise that requirements are
thoroughly explored with each iteration. Under this category, we

find rapid
throwaway prototyping, essentially a method of delivering fast releases of the
product with each iteration, exploring needs and requirements with each version,
and perfecting the next version accordingly. Needs are assessed as the product is
used and tested. On the other hand, the spiral model, which we have previously
discussed, is another form of exploratory prototyping, where prototypes are
employed in successive stages of the development process, each following the
waterfall pattern (Carr and Verner 1997 ).
The experimental approach entails that a solution to the user

'

rst
proposed and then evaluated through experimental use. The use of simulation
programs and skeleton programming (delivering only the most essential features of
the system so the user can get a general idea of what the

is needs is

final product will be like)
falls under this category, but there are many other examples, as it is the most
common form of prototyping (Floyd 1984 ).
Finally, the evolutionary approach essentially describes development in suc-
cessive versions and is closest to incremental and iterative life cycle models, in that
its main goal is to accommodate the eventual changes in requirements and needs.
The prototype is used fundamentally to allow for easier contact with the product, in
order to pinpoint perceived needs. Each prototype is no more than a version of the
product, and each version serves as the prototype for the next one (Floyd 1984 ).
By using a form of the prototyping model, a development project can easily
adapt to changing requirements, because there is constant feedback. With each
iteration, or version of the product, the user will have the ability to test the prototype
and provide valuable input on its traits and requirements. This provides the model
with much higher probabilities of success, as well as low risks. On the other hand,
because there is not much emphasis on extensive documentation, and the product
evolves as it is created, the time frame for the development project is much shorter
than with rigid models (Sabale and Dani 2012 ).
However, prototyping models are weak on analysis and design planning. While
requirements are assessed as the product is developed in successive versions, there
is little control over costs and resources, which can dramatically increase the

financial cost of the project (Sabale and Dani 2012 ). Therefore, we can conclude
that prototyping is ideal for larger projects and particularly for user-centric ones.