Path Coefficients and Path Regressions: Alternative or Complementary Concepts?

Abstract

The method of path coefficients and its more important pitfalls are reviewed briefly with reference to recent misunderstandings. Reasons are discussed for looking upon standardized coefficients (correlations, path coefficients) and concrete ones (total and path regressions) as aspects of a single theory rather than as alternatives between which a choice should be made. They correspond to different modes of interpretation which taken together give a deeper understanding of a situation than either can give by itself. It is brought out that even where the sole objectives of analysis are the concrete coefficients, actual path analysis takes a simpler and more homogeneous form in terms of the standardized ones, which can easily be converted into the concrete forms as the final step.