|
Although
not necessary for every feasibility study application, utility
curves are a good technique for translating diverse criteria
to a common scale. In cases where an attribute is difficult to
quantify or measure, the evaluation might establish a ratio
scale or use the analytic hierarchy process to establish a
relative scale for utility values.
The ratio and relative scaling approaches identify the
"best" alternative through a structured comparison
of alternatives. These approaches are most valuable in
considering non-technical parameters (such as cost,
development time, political saleability) where only subjective
(high, medium, low) evaluation is possible.
For example, if safety was an essential feasibility study
criterion, it could be included as a relatively scaled
attribute using a paired comparison process and scaling
methodology.
Utility curves for a given feasibility study must use
consistent scales (e.g., between 0 and 1) so as not to
inadvertently weight the scores. These models also must assume
the independence of criteria. The "zero point" of
each curve indicates the level of performance which no longer
provides value to system performance or effectiveness.
Graphic utility curves are not necessary for every criterion.
Where linear relationships are assumed between utility and
performance, simple tables can be established. Tabular scoring
plans could replace graphic charts for any criterion; however,
some fixed plan for scoring performance evaluations must be
established before the evaluations are conducted.
FEASIBILITY
STUDY REPORTS
|
|
|
