Multidimensional Scaling of User Information Satisfaction.

摘要

The objective of this study was to determine if multidimensional scaling reveals more about user perception of satisfaction with information systems than did factor analysis. Multidimensional scaling shed a different light on information satisfaction data, making them easier to visualize and interpret. While the differences were not substantial between multidimensional scaling and factor analysis, we concluded that the possibility of remarkably new insights gained through multidimensional scaling were well worth the small marginal cost of undertaking the analysis. Multidimensional scaling (MDS) provides an information technology (IT) manager with possible new perspectives in the analysis of user's satisfaction with information and with information systems. The technique probes for meanings locked in user satisfaction data that are not accessible by other analytic procedures. IT managers should be, in all cases, skeptical of contrived hypothesis testing factor analyses that deal with satisfaction data only at its face value. MDS gives managers a tool by which they can identify meanings beyond the obvious. Coupled with the careful and effective use of the semantic differential question format, MDS is a powerful means to escape the fatal flaw in data gathered by survey questionnaires: socially desirable responses.