CORRELATION WEIGHTS IN MULTIPLE REGRESSION

被引：0

作者：

Waller, Niels G. ^{[1
]}

Jones, Jeff A. ^{[1
]}

机构：

[1] Univ Minnesota, Dept Psychol, Minneapolis, MN 55455 USA

来源：

PSYCHOMETRIKA | 2010年 / 75卷 / 01期

关键词：

multiple regression; correlation weights; alternate weights; parameter sensitivity; LINEAR-MODELS; ESTIMATING COEFFICIENTS; EQUAL WEIGHTS; NO NEVERMIND; SENSITIVITY; ALTERNATIVES; PREDICTION; COMPOSITE;

D O I：

暂无

中图分类号：

O1 [数学];

学科分类号：

0701 ; 070101 ;

摘要：

A general theory on the use of correlation weights in linear prediction has yet to be proposed. In this paper we take initial steps in developing such a theory by describing the conditions under which correlation weights perform well in population regression models. Using OLS weights as a comparison, we define cases in which the two weight systems yield maximally correlated composites and when they yield minimally similar weights. We then derive the least squares weights (for any set of predictors) that yield the largest drop in R-2 (the coefficient of determination) when switching to correlation weights. Our findings suggest that two characteristics of a model/data combination are especially important in determining the effectiveness of correlation weights: (1) the condition number of the predictor correlation matrix, R-xx, and (2) the orientation of the correlation weights to the latent vectors of R-xx.

引用

页码：58 / 69

页数：12

共 50 条

[21] Multiple regression models [J].

Gillaizeau, Florence ;

Grabar, Sophie .

SANG THROMBOSE VAISSEAUX, 2011, 23 (07) :360-370

[22] The Prediction Model of Stroke on Climate Factors by Multiple Regression [J].

Yang, Jie ;

Ji, LuPing ;

Wang, Qiang ;

Lu, Xin .

2016 IEEE INFORMATION TECHNOLOGY, NETWORKING, ELECTRONIC AND AUTOMATION CONTROL CONFERENCE (ITNEC), 2016, :587-591

[23] Assessment of slope stability using multiple regression analysis [J].

Marrapu, Balendra M. ;

Jakka, Ravi S. .

GEOMECHANICS AND ENGINEERING, 2017, 13 (02) :237-254

[24] Determining attribute weights for multiple attribute decision analysis with discriminating power in belief distributions [J].

Fu, Chao ;

Xu, Dong-Ling ;

Xue, Min .

KNOWLEDGE-BASED SYSTEMS, 2018, 143 :127-141

[25] Robust extrapolation designs and weights for biased regression models with heteroscedastic errors [J].

Fang, ZD ;

Wiens, DP .

CANADIAN JOURNAL OF STATISTICS-REVUE CANADIENNE DE STATISTIQUE, 1999, 27 (04) :751-770

[26] Regionalization of Magnetic Susceptibility Measurements Based on a Multiple Regression Approach [J].

Fuerst, Christine ;

Zirlewagen, Dietmar ;

Lorz, Carsten .

WATER AIR AND SOIL POLLUTION, 2010, 208 (1-4) :129-151

[27] Investigating the Performance of Alternate Regression Weights by Studying All Possible Criteria in Regression Models with a Fixed Set of Predictors [J].

Niels Waller ;

Jeff Jones .

Psychometrika, 2011, 76

[28] The Geometry of Enhancement in Multiple Regression [J].

Niels G. Waller .

Psychometrika, 2011, 76 :634-649

[29] The Geometry of Enhancement in Multiple Regression [J].

Waller, Niels G. .

PSYCHOMETRIKA, 2011, 76 (04) :634-649

[30] Etemadi multiple linear regression [J].

Etemadi, Sepideh ;

Khashei, Mehdi .

MEASUREMENT, 2021, 186

← 1 2 3 4 5 →