Special review: Evaluation of the exploratory factor analysis programs provided in SPSSX and SPSS/PC+
Gregory J. Boyle · The Journal of multivariate experimental personality and clinical psychology. · 1993
Given the frequent use of SPSSX and SPSS/PC+ exploratory factor analysis in analyzing multivariate psychometric data, it is germane to examine the limitations of the Factor program as it currently exists. Over recent years, the routines in these packages generally have been developed and expanded considerably. In particular, the exploratory factor analysis procedures have been greatly extended and enhanced with inclusion of additional estimation methods such as minimum likelihood, unweighted least squares, generalized least squares, and so on. Likewise, availability of a Scree plot of the latent roots against factor number has facilitated determination of the appropriate factor-extraction number. Provision of a Χ 2 goodness-of-fit test, as well as options for sorting and retaining factor loadings greater than a minimum level (usually ± .30), have added to the utility of the SPSSX and SPSS/PC+ statistical routines. These routines have undoubtedly facilitated and stimulated psychometric reports involving the use of factor analysis in the psychological literature.