A very wide variety of both psychological and statistical techniques have been brought to bear upon the analysis of personality inventories and questionnaires. A most recent addition to this repertoire of methods is K-way scale analysis, which is a generalization of multiple-factor analysis to the situation in which two or more distinguishable scores on each variable can merely be ranked. The method is further characterized by the fact that it makes no assumption beyond assuming the existence of underlying dimensions or scales which determine the responses to a given collection of items or variables. Specifically, it makes no assumption about the form of any distribution, while factor-analysis of tetrachoric correlations requires the assumption of normality throughout. Furthermore, K-way scale analysis does not assume that all equal-appearing responses are necessarily equivalent, although this assumption is implicit in the Loevinger method of analyzing a pool of items into more homogeneous subpools. The procedure and an example of K-way scale analysis are reported separately. It is the purpose of this paper to consider whether the more rigorous approach to the analysis of personality inventory material which K-way scale analysis provides will result in personality scales with better practical properties, such as high reliability, low intercorrelation, and the availability of certain derived scores. Evidence pertinent to each of these points is presented, but the primary emphasis is on two particular derived scores.