On The Use of Variance Designs in Q Methodology

Abstract

Misgivings concerning the use of variance design and analysis in the structuring and evaluation of Q samples derive from the basic assumptions of deductive methodologies. Several alternative uses of variance design, as discussed by Stephenson, are suggested. Criticism of Stephenson's use of factorial arrangement is based on a misunderstanding of his position, too strict a reliance on the logical properties of Q samples, and failure to realize the synthetic quality of self-referent statements. Principles are illustrated with empirical examples.