Selection criteria
are weighted by the decision maker according to their relative
importance in determining the effectiveness of alternatives. To
ensure the objectivity of the subsequent analysis, weighting
factors developed by the decision maker may be withheld from the
analysts who do the performance evaluation.
Weighting follows a logical
breakdown. Effectiveness measures are examined for their
contribution to objectives for the system; then each criterion
is weighted according to its perceived contribution to the
effectiveness measures. The extent of the breakdown required is
determined by:
a.
The level
at which
performance evaluation is possible
b.
The level at which separate performance specifications
have
been
established.
Numerical weights are given
to reduce the effect of evaluator bias on the analysis.
Numerical weighting allows the decision maker to obtain an
objective assessment of
the alternatives.
In addition, numerical
treatment facilitates comparison among criteria that are not
related.
Decisions are based on
quantified criteria whether that quantification is subconscious
(unsystematic and un-documentable) or objective/ numerical
(systematic and documentable). Where a program maintains an
overall system effectiveness model and has operational data,
this weighting process can be very objective. In cases where
such a foundation is not available, decision support techniques
can be used to render subjective evaluations more reliable.
