Mediation/indirect effects in structural equation modeling
Oscar Gonzalez, Matthew J Valente, J Cheong, David P MacKinnon
2022-11-10
Abstract:—Sewall Wright, 1921, p. 568 the development of mediation models goes hand in hand with the history of SEM (see Matsueda, Chapter 2, this volume), and its applications in the social sciences (Duncan, 1966; Kendall & Lazarsfeld, 1950; Lazarsfeld, 1955). Modern approaches to quantifying indirect effects or mediation began with Sewall Wright’s work on path analysis (Wolfle, 1999; Wright, 1920, 1921). Investigating the relative influences of heredity and environment on the breeding of guinea pigs, he used a system of equations and a path diagram to display the causal relations among the variables. Wright showed that the contribution of the intermediate variables connecting the two variables in a complex system of causally related variables was the product of the path coefficients in a chain of paths connecting the two variables (Wright, 1923). He further argued that path analysis was not a method to infer causation but a method to quantify the supposed causal relations. Causal sequences of the variables were already assumed, based on available information including theory, prior correlations, and results of prior experiments; path analysis was used to gauge the strength of the causal relations. The specification of causal relations among variables is still one of the main issues in the use of SEM and we discuss this issue in relation to indirect or mediation effects later in the chapter (for a more general treatment of causal inference in SEM and structural causal models, see Pearl, Chapter 3, this volume). With the development of covariance structure modeling during the 1970s (Jöreskog, 1970, 1973; Wiley, 1973) and 1980s (Bentler, 1980), the …