Geoscience Reference
In-Depth Information
Bourdieu suggests using double-entry tables to comparatively identify objects
such as different types of educational institutions, martial arts or newspapers. One
row in the table is assigned to each institution. In an inductive initial stage, a new
column is opened whenever a property needed to characterize one of the individuals
is found. This results in posing the issue of the presence or absence of such property
for all the others. In a
final stage, repetitions possibly introduced are eliminated.
By proceeding in this way, a functional or structural feature is associated with
each column in such a way to retain all the features
that allow for
discriminating more or less rigorously the different institutions. In these instances,
preference will be related to the presence of desirable features or to the absence of
undesirable ones.
In this framework, a numerical representation for the preference according to
each criterion may be built by assigning the value 1 to the presence of a desirable
attribute,
and only those
1 to the presence of an undesirable one and 0 to absence of the attribute,
whatever its type.
When we use it for comparison purposes, this evaluation will be inaccurate
because the presence or absence of even the simplest attributes may be subject to
discussion. It will vary, for instance, the value that the presence or absence of a
feature may have to make the object useful for the evaluator.
2.1.1 General Criteria Features
The decision becomes a bit more complex when the comparison, rather than being
based on the ability to satisfy certain conditions, is based on the usefulness for a
particular purpose to have different amounts of a certain attribute, where utility
grows with the amount possessed. Thus, applying a criterion consists in evaluating
such ability or utility.
Keeney ( 1992 ) suggests that for a candidate to effectively become a criterion an
analysis of its properties should be conducted. In this examination, the criterion
candidate must prove to be controllable, measurable, operational, isolable, concise,
and understandable within the decision context. These and other properties must be
judged with respect to the alternatives to be evaluated. These alternatives may have
a high level of complexity, formed, for instance, by considering distinct results from
the same experiment as more or less satisfactory.
2.1.2 Criteria Combination
Different relationships between the criteria determine different algorithms for the
combination of the assessments according to them. The
final results depend on the
combination algorithm chosen as much as on the evaluations according to the
multiple criteria.
Search WWH ::




Custom Search