Environmental Engineering Reference
In-Depth Information
3.
Linear additive models
: An additive func-
tion combines individual values of an option
into one overall value. It multiplies the value
score on each criterion by the weight before
adding all those weighted scores together.
The weighted sum model (WSM) is a com-
monly used approach.
4.
Pairwise comparisons
: This process con-
sists of developing a linear additive model
and deriving weights and scores based on
pairwise comparisons between criteria
and alternatives (Saaty, 2008). Procedures
using pairwise comparisons include the
Eigen value method, the Anayltic Hierarchy
Process (AHP) and the Delphi method.
5.
Multi-Attribute Utility Theory (MAUT)
:
Utility theory gives decision-makers the
ability to quantify the desirability of certain
alternatives. It uses mathematical func-
tions—utility functions—to represent the
decision-makers' preferences given the
relative performance of each option toward
specific criteria. It does allow the calculation
of a single number index, U, which represents
the overall valuation of decision-makers for
a particular option. The most commonly
used functions are linear, crooked linear or
multiplicative (Winston, 1994).
Users should note that the most complex
methods are not necessarily the most appropriate
in all cases. Although pairwise comparisons are
commonly used in environmental evaluations
for the purpose of computing weights, industry
professionals—notably consultants from Jacobs
Consultancy—and professors at the University of
California, Berkeley express their concerns about
using the AHP with pairwise comparisons on an
airport master plan project. In their experience,
stakeholders fail to recognize the legitimacy of
such a method when its outcome differs from
their initial opinion. Two frequent issues arise:
first, participants often do not have an analytical
background and therefore see the tool as a 'black
box'; second, many voters fail to meet consistency
requirements when inputting comparisons and
weights, making the process inaccurate. While
pairwise comparisons derive weights in a quasi-
independent manner, the process is not practical
when considering more than a few criteria; using 20
Table 2. Strengths and shortcomings of commonly applied MCA procedures
Strengths
Shortcomings
Features of the proposed methodology
inspired from each MCA procedure
MCA Procedures
• Performance matrix
Straightforward and di-
rect inspection
Mostly qualitative
Inclusion of a performance matrix ac-
ceptable for projects with a low level
of complexity
• Outranking process
Direct elimination of
alternatives that do not
match specific criteria
Approach focuses on thresholds
and does not consider overall
picture
Inclusion of critical threshold values
to ensure that maximal impact values
are not exceeded
• Linear additive model
Simple and easy to imple-
ment
Unable to handle qualitative
criteria
Inclusion of a weighted sum model
to determine separately a cost per-
formance for quantitative scores and
a utility performance for qualitative
scores
• Pairwise comparison
Simple and easy to imple-
ment
Unreliable with larger numbers
of criteria
Inclusion of a weighting process
with different weight determination
procedures
• MAUT
Appropriate for qualita-
tive criteria
Utility is not an optimal descrip-
tion of quantitative criteria
Inclusion of utility functions for
qualitative criteria
