Rotation is used in almost all exploratory factor analysis (EFA) studies. There are numerous rotation strategies that can be employed in these various applications. This paper reviews the various rotation choices in EFA studies, including confirmatory rotation, and presents criteria useful in selecting rotation methods in various analytic…

Research supported by Office of Naval Research contract, N00014-67-A-0305-0012

Provides normative data about the distribution of one measure of similarity of factor loadings, the congruence coefficient, through a Monte Carlo Technique. Matching of "chance" factor patterns was done by the method of Tucker. Statistical tests of the results, based on similarities of the method to canonical and multiple correlation,…

Four prominent oblique transformation techniques--biquartimin, direct oblimin, promax, and the Harris-Kaiser procedure--are described and compared as to optimality of solution. Additionally, two newly-developed procedures--falling into the category designated as Case III by Harris and Kaiser (1964)--are presented and included in the comparisons.…

Rotation of canonical components is contrasted with factor rotation and the concept of communalities in canonical analysis is considered. Three possible strategies for rotating canonical components are described: single rotation of all components, separate rotation of each set, and sequential rotation of one set then the other. Constraints,…

In this paper a rather general theory of oblique factor rotation is outlined. The main results are formulated as four theorems. Necessary and sufficient conditions are derived for two factor matrices to admit identical factor structures and/or factor patterns with factors having unit variances. These conditions are expressed in terms of…

The oblimax, promax, maxplane, and Harris-Kaiser techniques are compared. (Author)

Under some circumstances, it is desirable to compare the factor patterns obtained from different factor analyses. To date, the best method of simultaneously achieving simple structure and maximum similarity is the technique devised by Bloxom (1968). This technique simultaneously rotates different factor patterns to maximum similarity and varimax…

Stephens et al. (EJ 055 112) committed a serious methodological error in holding chronological age constant in their IQ measures and allowing it to vary in their Piagetian developmental measures. This error is unrelated to differences in factor rotation methods used by these authors and was not answered in their reply. (CTM)

To investigate whether student attitudes toward basic school subjects were multidimensional, responses of 876 students in grade 6 to a preliminary pool of 72 items from the Survey of School Attitudes (SSA) were factor analyzed. If attitudes are unidimensional, as suggested by the four scores yielded by the SSA (one each for reading/language arts,…

These authors cite their own study of the relationship between traditional measures of intelligence and Piagetian measures of cognitive development in support of Glass and Stephens' contention that there are important qualitative differences. They question Humphreys' and Parsons' conclusions on both substantive and factor theoretical grounds. (CTM)