The stages of systems analysis
An Iterative Model
+-----> Evaluation -+-----> Problem realization
| | |
Implementation | |
^ | V
| +--------> Solution search
| | |
"Final approach" | |
testing | V
^ | Primary evaluation
| (Leads to) | |
| | |
Design +---+ V
^ | +---> Problem definition
| V |
+--- Feasibility study <----------+
For
the company in question this forms the survival,
inquiry and sophistication
stages of growth, also known as the WHAT, WHY and WHERE stages, as embodied by
the questions “What shall we eat?”, “Why do we eat?” and “Where shall
we have lunch?”. It usually takes
groups large periods of time to pass through all three stages but under certain
high-pressure situations small groups can pass through this cycle with alarming
speed.
For
example this cycle is completed in less than 5 minutes when the technician is
faced with the last minute problem.
Problem realization / Solution search
This is the stage at which the
need is realized. This is the
customer / end user stage where they attempt to verbalize the perceived need.
Once this is done a solution is searched for “Call in Bodge N Scarper,
the famous*
computer systems solutions specialists”.
This sets the stage for the
professionals to enter:
Primary evaluation
Now our pros
attempt to distinguish the perceived problem and the users actual requirements.
Here the systems analyst will attempt to ascertain “exactly what do you
want”. This is the primary question at this stage.
If this is answered well by quality investigation and cool headed
question asking then the rest of the cycle will be a lot smoother.
Effectively the analyst is attempting to tune in to the mindset of the
customer who may hold any of a range of bizarre ideas.
This leads to
The Problem Definition
Now we know
what the problem is we can codify it. This
way we now know what the issue(s) to be dealt with are.
Basic ideas as to solutions are brainstormed to give a framework within
to work.
This then forms the aim for
the:
Feasibility study
This is where
the systems analyst tries to find out what options and solutions are feasible.
As mentioned the analyst will rapidly pass through all the
stages as each issue speeds through an identical but smaller life cycle – the
ideas life cycle. The analyst will
be asking himself questions not dissimilar to the social survival questions –
What? Why? And Where? For example:
What do we need? Why do we need it? Where shall we put it?
As mentioned (top)
under stress small groups can pass through these stages with extreme rapidity.
Hopefully our
analyst will do a great job of finding out how people want to relate to the new
system, what parts of their anatomy they might want to input data with and
thousands of other details the users might not know about themselves.
This stage should
provide all the information that will later be used to design screens and
forms and every other part of the system. Part
of the job is vast information gathering. For example: does anyone know if the clerks want nasally
insertable computers?
All work that
happens after this point will rest on the work done here.
The questions can be summed up as: "What do you want?" The survival
questions are asked.
Design
Having decided
on a specific set of goals it now comes time to design in detail.
The workload now passes over to the designers and programmers.
These workers will also face the questions of survival, inquiry and
sophistication (what, why and where). What
do you want? Why on earth do you
want that? Where else can I work?
Having dealt
with the issues of workers with too little vision the work is carried out in a
series of minor (test) implementations.
It has been assumed thus far that intrinsic continuous “self” testing
occurs: this is the natural ability of humans (as opposed to computers) to spot
problems before they arise and to dismiss obviously impractical and unworkable
solutions. This is called the
self-audit system. The best way to
test your designs is to test each element as you build it. This is known as continuous testing. However as the proud parent of a new system you are unlikely
to use it in a way that might cause errors.
So we fade into the final stage.
“Final approach” testing
This is where
even as the technicians are setting things up they are testing, logging reports
and altering minor parts to form a practical working solution.
It is done this way because no matter how good the design looks on paper
or how fantastically orgasmic you find symmetrical drawings these things never
quite work like that in real life. This
is called the Real Life Principle.
Thus we reach the middle of
what is normally called
Implementation
The putting it all to work, of the
whole thing. This is what Dull N Co
have been waiting for -the product. Hopefully
it is now a good working system.
But just in case
Evaluation
At this point in the cycle we evaluate what we have done,
see if it was all worthwhile and see what changes will be needed in the future.
Bodge N Scarper, on the other hand, leg it as fast as possible pausing
only to get paid and ask for directions out of the car park.
If we are running through this cycle in micro then we would
almost always branch back to an earlier stage applying the gained knowledge to
find a better solution. We do this
all the time; you just never knew it.
*
Infamous would perhaps be better.