Sensitivity analysis in covariance structure analysis with equality constraints

摘要: Influence functions are derived for covariance structure analysis with equality constraints, where the parameters estimated by minimizing a discrepancy function between assumed matrix and sample matrix. As special case maximum likelihood exploratory factor is studied precisely numerical example. Comparison made results of Tanaka Odaka (1989), who have proposed sensitivity procedure in using perturbation expansion certain eigenvalues eigenvectors real symmetric Also present paper gives generalization Tanaka, Watadani Moon (1991) to constraints.