ABSTRACT
Historically the centroid method was commonly used for initial factoring until the advent of high-speed electronic computers permitted the more accurate but computationally much longer principal-axes method to supersede it. As the number of variables increases, and especially as the number of common factors increases, the Spearman approach, based on explicit theorems and the tetrad-triad procedures, becomes increasingly complex and soon becomes impractical. The first factor-analyst who recognized the problems fully, and who devised procedures to solve them, was L. L. Thurstone, though some of his early work was foreshadowed by the studies of E. Webb and Garnett. The objective of initial factoring is to factor in such a manner that the loadings on the first factor a shall be in aggregate as large as possible, and the loadings on each successive factor shall be as large as possible given the larger loadings on all previous factors.
