es
liser
ative
n en
lS Ce
ridéo
lules
lent,
e et
Létés
.f de
irin,
1° Où
> ces
dele
nées
image
jets
sont
les
aux
|'e et
on yv
'cine
Lque.
une
Lbles
SELECTION OF AN OPTIMUM STRUCTURED METHODOLOGY FOR
DEVELOPING AN INFORMATION SYSTEM; A GENERAL MODEL.
C.M. Paresi, M.M. Radwan, O.U. Essien
International Institute for Aerospace Survey and Earth Sciences ( ITC )
P.O. Box 6, 7500 AA Enschede, The Netherlands
KEY WORDS: Information Systems, Structured Development Methodologies, Optimum Methodology, Evaluation Model.
ABSTRACT
Where the overall intent is a realistic and organised control of the System Development process, with a resultant effective
Information System, then it becomes imperative that the Methodology chosen for the development of the System, should
be optimum for that Object System. The paper shortly reviews the characteristics of Structured Development
Methodologies. It then focuses on the determination of a general evaluation model criteria for selecting an optimum
Structured Development Methodology applicable to a large number of problem domains. An overview of the model is first
presented, followed by an explanation of the steps in the process of evaluation.
INTRODUCTION
In the development of Information Systems which could be
Administrative, Technical, or Managerial, Structured
Development Methodologies have been used to create
systems that give an effective communication between the
system users, the machines that support the systems
operations, as well as the human factors in information
exchange. The bottom line is an efficient and organised
control of the system development process.
The number of these Methodologies have of recent
increased considerably in the Information Science/
Technology scene. Some look similar in many respects,
some use techniques that overlap in concepts, the
advantages or otherwise of some are more apparent in
some types of Systems than the others, and some lay
more emphasis on some aspects of Information Systems
development.
Therefore, a research was initiated on the determination of
a general evaluation model criteria for selecting an
Optimum Structured Development Methodology applicable
to a large number of problem domains.
The philosophy of the model is based on the fact that it is
unlikely at the moment, to have a single methodology that
can satisfy all the phases and stages during system
development. Therefore for a particular Object System,
development phases/stages are identified according to their
relative importance. The methodologies are then examined
for how well they provide for the individual phases/stages.
Based on the peculiar circumstances of the Object System
under study, several contingencies/criteria can be
determined and embedded in a general framework through
which the methodologies would be evaluated.
The model makes provisions for levels of detail to be
examined in the Object System, the levels of decomposition
to which phases/stages of the Object System can be
reasonably subjected, and also a means of reducing
subjectivity inherent in this type of evaluation. It is possible
to carry out the evaluation either for each phase, or the
entire phases put together. It can also be decided using a
contingency sub-model, whether the development strategy
should be the use of only prototypes, specifying, or a
mixture of both. The model recognises the influences from
issues collateral to the Object System.
WHY STRUCTURED DEVELOPMENT
METHODOLOGIES?
Systems Development is a non-trivial task; it costs time and
money because of the complex nature of modern
Information Systems and therefore needs a clear step by
step approach in the process of development. This is the
concept behind Methodologies. It defines the pieces or
components, phases and activities that one finds in a
typical System Development project; and also the interface
between those components [Burch,1979]. The most
important factor for the need for Methodologies is the
limitation of the human mind to perceive and retain all
information it requires and to act on it promptly [Dippel,
1969].
A Methodology should have as general objectives, to :
Analyse the complexity of problems and simplify them;
achieve unity of System Architecture; establish and improve
interaction between users and System Development Team;
enable efficient parallel development of Sub Systems;
achieve sound data analysis and administration through a
well defined procedural framework [Elving & Kirchoff, 1991].
A Methodology should specify:
How a project is to be broken down into phases, activities
for each phase, outputs for each phase, when activities
should be executed, constraints to be applied, support tools
to be utilized, and how projects have to be managed.
Techniques (which can be diagrams, tables etc.) should be
recommended at various stages for carrying out particular
activities recommended by a Methodology.
257
