Structural Equation Modeling and Regression: Guidelines for Research Practice

TLDR

Growing interest in Structured Equation Modeling (SEM) in information systems research highlights the need to compare and contrast SEM techniques to guide appropriate research design selection. The study aims to offer guidelines on when to employ SEM versus linear regression models. The authors assess current SEM usage, illustrate the same dataset with three distinct statistical techniques, compare covariance‑based and partial‑least‑squares SEM, and discuss regression modeling. They present heuristics and rule‑of‑thumb thresholds for practice and evaluate how well current practice aligns with these guidelines.

Abstract

The growing interest in Structured Equation Modeling (SEM) techniques and recognition of their importance in IS research raises the need to compare and contrast the different types of SEM techniques so that research designs can be selected appropriately. After assessing the extent to which these techniques are currently being used in IS research, the article presents a running example which analyzes the same dataset via three very different statistical techniques. It then compares two classes of SEM: covariance-based SEM and partial-least-squares-based SEM. Finally, the article discusses linear regression models and suggests guidelines as to when SEM techniques and when regression techniques should be used. The article concludes with heuristics and rule of thumb thresholds to guide practice, and a discussion of the extent to which practice is in accord with these guidelines.