Fitting covariance models for theory generation

Covariance-based structural equation modeling (CB-SEM) is an increasingly popular technique for analyzing quantitative data in Information Systems research. As such, it is traditionally viewed as a method to test theory, rather than build it. However, many of the theoretical models tested with this technique in IS research show significant differences between the model and the data. This paper shows that as part of the pursuit of model fit, researchers using CB-SEM can provide deeper insights into a phenomenon, allowing us to build theories based on quantitative data