Direct and Indirect Effects
This work addresses a conceptual and practical difficulty in causal inference for researchers and practitioners dealing with nonlinear models, offering a more natural and broadly applicable approach to assessing direct and indirect effects.
The paper tackles the problem of defining and measuring direct and indirect effects in nonlinear models, where traditional methods fail, by introducing a new way to define path-specific effects without blocking other paths, applicable in both linear and nonlinear models. It establishes conditions for consistent estimation from experimental and nonexperimental data, extending path-analytic techniques to nonlinear and nonparametric models.
The direct effect of one eventon another can be defined and measured byholding constant all intermediate variables between the two.Indirect effects present conceptual andpractical difficulties (in nonlinear models), because they cannot be isolated by holding certain variablesconstant. This paper shows a way of defining any path-specific effectthat does not invoke blocking the remainingpaths.This permits the assessment of a more naturaltype of direct and indirect effects, one thatis applicable in both linear and nonlinear models. The paper establishesconditions under which such assessments can be estimated consistentlyfrom experimental and nonexperimental data,and thus extends path-analytic techniques tononlinear and nonparametric models.