Describes and illustrates the use of fuzzy ''partition'' statistical methods, based on fuzzy set mathematical principles, as a means of analyzing high dimensional data with categorical responses. Statistical procedures such as estimation, testing of hypotheses and goodness of fit evaluation are based upon maximum likelihood principles adapted to fuzzy set mathematics and logic.