203 / 330
臺大管理論叢
第
27
卷第
2
期
203
available in the literature for supplier evaluation and selection. In this research, SMART was
combined with the utility theory to take into account buyers’ risk-taking behaviors. Two
different types of weights: experts’ weights and ROC weights were also considered and
compared. Moreover, the presented case study examined the supplier selection method in
three different scenarios: market demand is increasing, decreasing, and remains stable.
Most, if not all, of the previous works on supplier selection did not consider different
(external) market demand scenarios; suppliers’ attributes are their only concern. In the
presented case study, when applying the proposed method, supplier A is the best choice when
market demand is increasing and remains stable, while supplier B is the best when market
demand is shrinking. The results show that when market demand is (or expected to be)
increasing, risk-seeking buyers (manufacturers) may be more inclined to enlarge their
forecasts and use suppliers with larger capacities and/or flexibilities, whereas risk-averse and
risk-neutral buyers may or may not be willing to adjust their forecasts and change suppliers.
On the other hand, when market demand is decreasing, risk-averse buyers may reduce their
forecasts and select suppliers with lower inventory costs (to avoid further losses), whereas
risk-seeking and risk-neutral buyers may or may not switch suppliers. Such difference
highlights the importance of incorporating buyers’ risk-taking behaviors and of considering
market demand in the supplier selection problem.
This research proposed an approach which combines SMART with the utility theory for
supplier selection. It will be interesting to compare the proposed approach with other
methods developed in the literature, such as DEA. Note that our approach features the
incorporation of the utility function in the SMART to consider buyers’ risk-taking behavior
in supplier selection, whereas it may not be straightforward for those existing methods to
account for buyers’ risk-taking behavior. For instance, the incorporation of buyers’ risk-
taking behavior in DEA models may lead to nonlinear models that are difficult to solve; the
AHP method allows experts (or buyers) to determine the relative importance between
different attributes but cannot reflect buyers’ risk-taking behavior. In other words, significant
modifications would have to be made for those existing methods to address buyers’ risk-
taking behavior, which is not a trivial task to our current study. Moreover, other approaches,
such as swing weights, AHP and ANP, can also be employed to determine the weights of
attributes and criteria, instead of experts’ and ROC weights used in this research. In this
study, the survey results are obtained from the internal experts of the case-study company. It
will be interesting if opinions from external experts can also be solicited and analyzed, which
may provide more general results with the attributes considered in this study.