191 / 330
臺大管理論叢
第
27
卷第
2
期
191
makers in selecting suppliers. This model incorporates a chain-of-interaction method for
determining the weight of attributes without subjective human judgment. The other AHP-
based decision-making approaches for the supplier selection problem can be found in Chan,
Chan, Ip, and Lau (2007) and Kuo, Lee, and Hu (2010).
Although the AHP is both technically valid and practically useful, it is not without
criticisms (e.g., Chan and Kumar, 2007; Wu and Barnes, 2011). For instance, AHP does not
consider the interaction of different attributes or risk and uncertainty of alternatives, and
assumes that the relative importance of attributes is known with certainty. Moreover, the
number of comparison tables can become very large if one uses a lot of comparison
attributes. This can lead to a tendency to exclude valid comparison attributes in order to keep
the number of calculations manageable.
The analytic network process (ANP) is a generalization of the AHP used in MCDM
(Saaty, 1996). AHP structures a decision problem into a hierarchy with a goal, decision
criteria, and alternatives, while the ANP structures it as a network. Both use a system of
pairwise comparisons to measure the weights of the components of the structure, and finally
to rank the alternatives in the decision. AHP assumes that in the hierarchy, the decision
criteria are independent of each other; and the alternatives are independent of the decision
criteria and of each other. However, in many real-world cases, there is interdependence
among the items and the alternatives. ANP does not require independence among elements,
so it can be used as an effective tool in these cases (Saaty, 2005).
Several previous works employed ANP to deal with the supplier selection problem.
Sarkis and Talluri (2002) considered the internal interdependence among supplier-evaluating
factors and used ANP to evaluate and select the best supplier with respect to organization
factors and strategic performance metrics. Taking into account the interrelationships between
evaluating criteria in the selection process, Bayazit (2006) and Gencer and Gürpinar (2007)
also utilized ANP to tackle the supplier selection problem.
2.4 Simple Multi-Attribute Rating Technique (SMART)
SMART allows direct entry of relative weight for each attribute/criterion through rates,
in the natural scales of the attribute. The weights of attributes can be assigned by decision-
makers (experts’ weights) or determined by different rules, such as Rank Order Centroid
(ROC) weights (Barron and Barrett, 1996). In order to convert different scales of criteria into
a common internal scale, SMART adopts a value function. The simplest choice of the value
function is a linear function, and in most cases, this is sufficient. However, to better capture