Introducing Simultaneous Equation Modeling for Social Work Research

摘要

Social science researchers generally use simple, single equation descriptions of complex social problems. More accurate descriptions would portray the joint dependency among a model's variables by using a separate equation for each dependent variable. When the parameters of these more accurately described models are estimated with ordinary least squares (OLS), the resulting estimates do not have good statistical properties. A two-stage least squares procedure to estimate the parameters of simultaneous equation models is described using an example about heavy workplace drinking and the social climate at work. In addition, this article defines key concepts related to simultaneous equation modeling and provides an interpretation of the results for practitioners.