QUANTIFICATION OF CATEGORICAL DATA USING DIMENSIONAL ANALYSIS
In previously published works we demonstrated that a modification of Dimensional Analysis (DA) from physics, referred to as Reversed Dimensional Analysis (RDA), can be used to attribute a dimension to a non-physical variable provided measurement data are available, and that such a dimension allows for a meaningful interpretation (cf. Marinov, 2004, 2005). In this article we describe a further step in using DA beyond physics, namely in quantifying categorical data. A theoretical basis and a computational algorithm for such quantification are described. Computer simulations with model data of the kind X1=f(X2,X3), where one of the variables is considered categorical (up to fifteen distinct categories), show that the emerging scale of computed values resembles the structure of the scale of model values, and that the values of the categorical variable can be determined unambiguously.