摘要: Interdependence among explanatory variables is a common condition for sociological analyses. It may markedly affect the stability of estimates parameters obtained from least-squares regression. Multicollinearity viewed as problem which poses two questions analyst: how severe multicollinearity and what its effect on analysis? The determinant correlation matrix measure severity multicollinearity. Haitovsky's chi-square statistic permits assessment null hypothesis that singular. This paper demonstrates need this test through an examination published matrices. suggested use Haitovsky be routine in any analysis attempts estimation regression analysis.
sagepub.com
sci-hub.se 参考文章(7)
Robert R. Glauber, Donald Eugene. Farrar, Multicollinearity in Regression Analysis; the Problem Revisited ,(2011)
Yoel Haitovsky, Multicollinearity in Regression Analysis: Comment The Review of Economics and Statistics. ,vol. 51, pp. 486- 489 ,(1969) , 10.2307/1926450
Herbert L. Costner, Edgar F. Borgatta, Sociological Methodology, 1969. American Sociological Review. ,vol. 34, pp. 950- ,(1969) , 10.2307/2095986
Otis Dudley Duncan, CONTINGENCIES IN CONSTRUCTING CAUSAL MODELS Sociological Methodology. ,vol. 1, pp. 74- ,(1969) , 10.2307/270881
David Knoke, A Causal Model for the Political Party Preferences of American Men American Sociological Review. ,vol. 37, pp. 679- ,(1972) , 10.2307/2093579
David R. Heise, Problems in Path Analysis and Causal Inference Sociological Methodology. ,vol. 1, pp. 38- ,(1969) , 10.2307/270880
H. M. Blalock, Correlated Independent Variables: The Problem of Multicollinearity Social Forces. ,vol. 42, pp. 233- 237 ,(1963) , 10.1093/SF/42.2.233