Give Me Attitude: Making Smart Use of Structural Equation Modeling and Other Tools When Analyzing Survey Data

Many researchers were trained primarily in the tools of either econometric-style regression or psychology-inspired structural equation modeling (SEM). Given that quantitative research in public policy and administration regularly draws on both traditions of modeling, researchers frequently find themselves in the position of reading and evaluating studies that utilize modeling tools they don’t understand well. This study aims to address this problem through a detailed comparison of each set of tools, with examples offered through reanalysis of two datasets used in prior studies. I argue that in many cases, econometric and SEM tools yield similar substantive conclusions, though they also tend to be used to answer slightly different questions. Econometric tools fit more naturally with research designs intended to probe issues of causality, while SEM is better suited to generating detailed descriptions of patterns of associations among several variables. SEM also encourages researchers to think carefully about both mediation and measurement, and it offers greater flexibility in correcting for measurement error. At the same time, measurement error in the real world is typically even more complicated than what SEM can account for. Econometrics tools offer more options for dealing with measurement error than many researchers realize, and econometric models offer more flexibility in terms of the number and type of variables that can be realistically analyzed. The study also provides a practical set of guidelines for analyzing hard-to-measure constructs—guidelines which can be used by researchers regardless of which disciplinary tradition of statistical modeling they draw on.