Premium
ANALYTIC METHODS OF FACTOR ROTATION
Author(s) -
Warburton Frank W.
Publication year - 1963
Publication title -
british journal of statistical psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.157
H-Index - 51
eISSN - 2044-8317
pISSN - 0950-561X
DOI - 10.1111/j.2044-8317.1963.tb00207.x
Subject(s) - varimax rotation , oblique case , rotation (mathematics) , mathematics , maximization , normalization (sociology) , thurstone scale , covariance , statistics , geometry , mathematical optimization , psychometrics , linguistics , philosophy , cronbach's alpha , sociology , anthropology
Recent work on analytical methods of factor rotation applied to large‐scale data has emphasized the use of oblique factors and the simple structure criterion. The chief methodological advances have been (i) to use the variation of column (factor) loadings rather than row (test) loadings in order to reach a clearer simple structure (ii) the normalization of rows before rotation, in order to give the tests equal weight (this prevents complex tests, with high loadings in several factors, pulling the factors too close to each other) and (iii) the minimization of the covariance of the loadings (rather than the maximization of the variance) since this requires that high and low loadings occupy adjacent positions in the factor table in accordance with Thurstone's third condition for simple structure. The best method would appear to be Varimax in the orthogonal case, and Binormamin for oblique rotation. Maxplane is the closest to rotation by subjective methods.