Skip to main content

Linear causal models are conveniently developed by the method of path coefficients proposed by Sewall Wright. Path analysis is useful in making explicit the rationale of conventional regression calculations. It may also have special usefulness in sociology problems involving the decomposition of a dependent variable or those in which successive experiences of a cohort are measured. Path analysis focuses on the problem of interpretation and does not purpot to be a method for discovering causes. It may, nevertheless, be invaluable in rendering interpretations explicit, self-consistent, and susceptible to rejection by subsequent research.