The Application of Structural Equation Modelling in Information Systems Research

Recent years, structural equation modelling (SEM) has become one of the most used multivariate data analysis techniques in information systems (IS) research. Although textbooks and guidelines on the application of SEM have been published, the increasing use of SEM in the domain calls for an updated review of its use within IS research. With this article, we aim to contribute to a more consistent practice of SEM. This will make it easier to compare SEM studies in the future. A framework developed by Shook et al. (2004) for categorizing the application of SEM is applied. In this article, we modify the framework, apply the modified framework to a sample of SEM-studies in IS research, present our results and compare these results with the results from Shook et al. (2004). Our study reveals that there are still elements in both the reporting and application practices of SEM that may be improved. Among these elements are the use and reporting of measurements of reliability and validity, weaknesses in reporting items such as input matrices enabling reproduction and replication of study results, validity of measurement and structural model, and the issue of cross-sectional versus longitudinal designs.