Weighting systems for linear functions of correlated variables when there is no dependent variable

  title={Weighting systems for linear functions of correlated variables when there is no dependent variable},
  author={Samuel Stanley Wilks},
  • S. S. Wilks
  • Published 1 March 1938
  • Mathematics
  • Psychometrika
When no criterion variable is available, the combination of tests or other variables by the use of multiple correlation is not possible. Three methods of combining variables are described mathematically, and discussed with reference to the linear combination of tests. Iterative computational schemes are outlined and illustrated. 
A unified treatment of the weighting problem
The general procedure is shown to yield certain desirable invariance properties, with respect to transformations of the variables, that are desirable in the context of weighted linear combinations of variables.
Regression analysis of linear composite variance
The problem of defining and determining the effective contribution of a component variable to the variance of a composite is briefly reviewed. Another method of dealing with this problem is proposed
The Effect of Unlike Distributions on the Weights of Variables
WHEN scores of several variables are available, on the basis of which a decision is to be made, some method must be used to derive a single score, which should represent the standing of individuals
The Relative Efficiency of four Test Weighting Methods in Multiple Prediction
APPLIED psychologists frequently have occasion to combine a number of performance measures for an individual in order to obtain a single composite estimate of behavior. Typical in school and industry
Best linear composites with a specified structure
Least squares linear composites of predictors for estimating several criteria are derived, satisfying the restriction that the composites have an arbitrary specified intercorrelation matrix. These
Fungible Weights in Multiple Regression
Every set of alternate weights (i.e., nonleast squares weights) in a multiple regression analysis with three or more predictors is associated with an infinite class of weights. All members of a given
Scoring under ordered constraints in contingency tables
Methods are developed for analyzing contingency tables which have ordered categoriesa priori. A correspondence analysis is extended to incorporate this a priori ordering. An exact represention is
A Constrained Linear Estimator for Multiple Regression
Abstract“Improper linear models” (see Dawes, Am. Psychol. 34:571–582, 1979), such as equal weighting, have garnered interest as alternatives to standard regression models. We analyze the general
Alternative Methods for Combining Several Test Scores
The author briefly reviews the existing methods and proposes some new ones for obtaining weights that are best in some sense, and illustrates the methods by using the data on certain city government employees.