Path Coefficients and Path Regressions: Alternative or Complementary Concepts?

(2022) - Sewall Wright

Journal:
Link::
DOI::
Links::
Tags:: #paper #Methods #PathCoefficients
Cite Key:: [@Wright2022]

Abstract

In a recent paper, Turner and Stevens [1959] develop a modification of the method of path coefficients (Wright [1921, 1934, 1954]). Following Tukey [1954] they advocate systematic replacement in path analysis of the dimensionless path coefficients by the corresponding concrete path regressions. The purpose of the present paper is to discuss this and other points which they raise.

Notes

“The authors concur with Tukey in treating the standardized and concrete forms of correlational statistics as if they were alternative conceptions between which it is necessary to make a choic” (Wright, 2022, p. 189)

“E1ven when the sole objectives of analysis are the concrete coefficients, actual path analysis takes a simpler and more homogeneous form in terms of the standardized one” (Wright, 2022, p. 189)

“It is of first importance in path analysis to make use of all of the available dat” (Wright, 2022, p. 189)

“The method is one for dealing with a system of interrelated variables. It is based on the construction of a qualitative diagram in which every included variable, measured or hypothetical, is represented (by arrows) either as completely determined by certain others (which may be represented as similarly determined) or as an ultimate factor. Each ultimate factor in the diagram must be connected (by lines with arrowheads at both ends) with each of the other ultimate factors to indicate possible correlation through still more remote unrepresented factors, except in cases in which it can safely be assumed that there is no correlation” (Wright, 2022, p. 190)

“It is assumed here that all relations are linear: non-linear relations may sometimes be transformed systematically throughout a diagram into linear ones” (Wright, 2022, p. 190)

“The validity of the system requires that variables that enter into two or more relations in the system (as a common factor of two or more, or as an intermediary in a chain) act as if point variable” (Wright, 2022, p. 190)

“The primary purpose was stated in the first general account (Wright [1921] as follows: "The present paper is an attempt to present a method of measuring the direct influence along each separate path in such a system and thus of finding the degree to which variation of a given effect is determined by each particular cause. The method depends on the combination of knowledge of the degrees of correlation among the variables in a system with such knowledge as may be possessed of the causal relations. In cases in which the causal relations are uncertain, the method can be used to find the logical consequences of any particular hypothesis in regard to them."” (Wright, 2022, p. 191)