A Comparison of the Direct Oblimin and Primary Parsimony Methods of Oblique Rotation

The Crawford‐Ferguson family of rotation criteria is extended to the oblique primary factor case. This method of rotation is referred to as the primary parsimony criterion, and it is conceptually and empirically compared with the direct oblimin method. Results of this study indicate that the advantages of the primary parsimony criterion are as follows: ( a ) the rationale for this approach is well understood; ( b ) it is a natural extension of the orthomax criterion for orthogonal rotation; ( c ) it has parameters for altering the properties of the solution; and ( d ) it does not allow collapse of the factor space. Considering all special cases, its results are superior to those obtained with the direct oblimin criterion.