An Application of Multidimensional Scaling to the Prioritization of Decision Aids in the S-3A

摘要

This thesis presents an application of Multidimensional Scaling (MDS) used in the prioritization of ASW decision functions in the S-3A. The ASW decision space was divided into 14 discrete decision functions for the purposes of this analysis. The problem of developing a prioritization methodology was approached from two independent directions. First, an unconstrained sorting task was preformed to provide input to Multidimensional Scaling algorithm. The result of this analysis provided a three dimensional representation of the decision space with dimensional interpretation. Second, a series of ranking tasks were preformed to provide input to an Unfolding Analysis algorithm. The Generalized Distance Model was selected as the model most representative of the ranking data. The decision function coordinates for the MDS algorithm and the decision function coefficients for the Unfolding Analysis algorithm were combined in a regression-like equation to provide a prioritization methodology for the 14 decision functions of the S-3A ASW decision space. (Author)