Impact prediction, evaluation and mitigation - Introduction to Environmental Impact Assessment - page 154

• Economic components
Group 2
(e.g. indigenous population <45 years old)
various
• Social
• Physical
• Economic components
Figure 5.7 Simple matrix
identification of distribution of
impacts.
weightings. Some of the limitations of such approaches have already been noted in
Chapter 4.
Other methods in the multi-criteria/multi-attribute category include decision analysis,
the goals achievement matrix (GAM), multi-attribute utility theory (MAUT) and
judgement analysis. Decision analysis is the operational form of decision theory, a theory
of how individuals make decisions in the face of uncertainty, which owes its modern
origins to Von Neuman & Morgenstern (1953). Decision analysis usually involves the
construction of a decision tree, an example of which is shown in Figure 5.8. Each branch
represents a potential action, with a probability of achievement attached to it.
The GAM was developed as a planning tool by Hill (1968) to overcome the perceived
weaknesses of the PBS approach. GAM makes the goals and objectives of a project/plan
explicit, and the evaluation of alternatives is accomplished by measuring the extent to
which they achieve the stated goals. The existence of many diverse goals leads to a
system of weights. Since all interested parties are not politically equal, the identified
groups should also be weighted. The end result is a matrix of weighted objectives and
weighted interests/agencies (Figure 5.9). The use of goals and value weights to evaluate
plans in the interests of the community, and not just for economic efficiency, has much to
commend it. The approach also provides an opportunity for public participation.
Unfortunately, the complexity of the approach has limited its use, and the weights and
goals used may often reflect the views of the analyst more than those of the interests and
agencies involved.
MAUT has gained a certain prominence in recent years as an evaluation method that
can incorporate the values of the key interests involved (Bisset 1988, Edwards &
Newman 1982, Parkin 1992). MAUT involves a number of steps, including the
identification of the entities (alternatives, objects) to be evaluated, and the identification
and structuring of environmental attributes (e.g. noise level) to be measured. The latter
may include a “value tree” with general objectives (values) at the top and specific
attributes at the bottom. The ranking of attributes is by the central stakeholder/expert
group whose values are to be maximized. Attributes are scaled and formal value or utility
models developed to quantify trade-offs among attribute scales and attributes. For further
reference, see Parkin (1992) for an outline of the main steps and an application of a
“relatively” simple and well-proven version of MAUT known as the SMART method.