Structural Equation Modeling With Robust Covariances

摘要: Existing methods for structural equation modeling involve fitting the ordinary sample covariance matrix by a proposed model. Since is easily influenced few outlying cases, standard practice of covariances can lead to inefficient estimates as well inflated fit indices. By giving proper weight each individual case, robust will have bounded influence function nonzero breakdown point. These properties estimators be carried over parameter in model if technically appropriate procedure used. We study such which replace context Wishart likelihood function. This easy implement practice. Statistical this are investigated. A index given based on sampling from an elliptical distribution. An estimating approach used develo...