Adjustments for Kurtosis in Factor Analysis with Elliptically Distributed Errors
Journal of the Royal Statistical Society: Series B (Statistical Methodology) · 1987 · Volume 49 · Issue 3
DOI: 10.1111/j.2517-6161.1987.tb01705.x
Abstract
Simple corrections to normal theory tests of fit of the factor analysis model are provided for the situation where common factors have arbitrary distributions and errors have a distribution from the elliptical class. It is shown that normal theory estimators are asymptotically efficient in this situation and appropriate modifications for the asymptotic covariance matrix of estimators are provided. Factor analysis models with structured error covariance matrices are also considered.
Keywords
asymptotic distributions, elliptical distributions, factor analysis, maximum likelihood, minimum discrepancy, robustness, tests of fit
Enable JavaScript for citation tools and additional paper interactions.