TY - JOUR
AU - Widaman, Keith F.
AB - Multi~~ariate Beha,ior-a1 Research, 25 (1 ), 89-95 Copyright O 1990, Lawrence Erlbaum Associates, Inc. Bias in Pattern Loadings Represented by Common Factor Analysis and Component Analysis Keith F. Widaman University of California at Riverside In this comment, I will consider the first two of the data analytic issues discussed by Velicer and Jackson (1990). The substance of these two issues is (a) that common factor analysis and component analysis provide quite similar representations of empirical data, and (b) that differences between common factor analytic and component analytic representations arise only if too many factors or components are retained. In the following, analytic expressions for the basis of the differential bias in common factor and component pattern loadings are offered for population data of the type used in recent Monte Carlo studies (Guadagnoli & Velicer, 1988; Snook & Gorsuch, 1989); the analytic expressions are then evaluated through analyses of the population data. The goal here is to argue that the first two data analytic issues discussed by Velicer and Jackson are not as easily resolved as they have supposed. The Mathematical Basis,for Bias in Pattern Loadings Both component and common factor analysis are used to obtain reduced rank representations of 
TI - Bias in Pattern Loadings Represented by Common Factor Analysis and Component Analysis
JF - Multivariate Behavioral Research
DO - 10.1207/s15327906mbr2501_11
DA - 1990-01-01
UR - https://www.deepdyve.com/lp/taylor-francis/bias-in-pattern-loadings-represented-by-common-factor-analysis-and-zliE0XHHAr
SP - 89
EP - 95
VL - 25
IS - 1
DP - DeepDyve
ER -