Why use a Structural Equation Model?
In this edition of CBA Office Hours, Dan discusses some of the principal advantages of the structural equation model (SEM) relative to more traditional data analytic approaches like the linear regression model. Advantages include the ability to account for measurement error when estimating effects, test the fit of the model to the data, and specify statistical models that more closely align with theory. Dan describes these advantages with an example on factors that relate to children’s popularity with peers. We consider these issues and various extensions of the SEM (such as longitudinal applications, ways of formally testing mediation and moderation, and evaluating invariance of effects across groups both known and unobserved) in greater detail in other posts on SEM and in our summer training workshops.