Combining DEMATEL with ANP to modify multidimensional scaling in identifying the similarities of e-shopping stores

Abstract

While many studies have discussed the purchase behavior for e-shopping stores, there is little study focusing on mechanism which consumers accurately and reasonably identify the similarities of e-shopping stores. Therefore, we introduce the approach which combines DEMATEL with ANP to modify multidimensional scaling to address this concern. In traditional MDS, the average summation is combined with the proximity matrices of the objects for each criterion (perceived price of stores), which may neglect the interdependence and feedback relationships between the criteria. The weights and relationships of different criteria are treated as equal and independent to obtain similarities that are often counterintuitive because the human mind has a much more complex judgment system that is able to account for different weighting. To address these problems, we propose an approach that combines DEMATEL with ANP. An empirical study is used to compare our approach with traditional MDS. Key words: Multidimensional scaling (MDS), Analytic network process (ANP), Decision Making Trial and Evaluation Laboratory (DEMATEL), multiple criteria decision making (MCDM), Purchase intention, online shopping.