R-mode and Q-mode influence analyses in statistical modelling: relationship between influence function approach and local influence approach

作者: Yutaka Tanaka , Fanghong Zhang

DOI: 10.1016/S0167-9473(99)00026-2

关键词:

摘要: Relationship between the influence analyses based on function (R-mode analysis) and Cook's local (Q-mode is discussed. The equivalent results are obtained under general conditions by these in statistical modelling without/with equality constraints, if same perturbation introduced both analyses. equivalence holds most normal theory multivariate methods including regression analysis, principal component canonical correlation exploratory confirmatory factor analyses, covariance structure analysis with or without constraints.

参考文章(18)
Frank R. Hampel, The Influence Curve and Its Role in Robust Estimation Journal of the American Statistical Association. ,vol. 69, pp. 383- 393 ,(1974) , 10.1080/01621459.1974.10482962
Yutaka Tanaka, RECENT ADVANCE IN SENSITIVITY ANALYSIS IN MULTIVARIATE STATISTICAL METHODS Journal of the Japanese Society of Computational Statistics. ,vol. 7, pp. 1- 25 ,(1994) , 10.5183/JJSCS1988.7.1
Yutaka Tanaka, SENSITIVITY ANALYSIS IN HAYASHI'S THIRD METHOD OF QUANTIFICATION Behaviormetrika. ,vol. 11, pp. 31- 44 ,(1984) , 10.2333/BHMK.11.16_31
L Shi, Local influence in principal components analysis Biometrika. ,vol. 84, pp. 175- 186 ,(1997) , 10.1093/BIOMET/84.1.175
Roderick P. McDonald, Yukihiko Torii, Shizuhiko Nishisato, Some results on proper eigenvalues and eigenvectors with applications to scaling Psychometrika. ,vol. 44, pp. 211- 227 ,(1979) , 10.1007/BF02293972
Shu-Jia Wang, Sik-Yum Lee, Sensitivity analysis of structural equation models with equality functional constraints Computational Statistics & Data Analysis. ,vol. 23, pp. 239- 256 ,(1996) , 10.1016/S0167-9473(96)00034-5
FRANK CRITCHLEY, Influence in principal components analysis Biometrika. ,vol. 72, pp. 627- 636 ,(1985) , 10.1093/BIOMET/72.3.627