Structural Equation Modeling In User Experience Research Two Case Studies

The intent of this paper is to advance the use of structural equation modeling (SEM) in user experience research, adding another tool to the user experience toolkit. The resulting models help communicate the value of user experience research to business stakeholders and decision makers. SEM is a confirmatory modeling technique that allows practitioners to specify and test relationships of latent (unobserved) and manifest (observed) variables. To this end, two SEM case studies are described, where the observed variables were users’ responses and experiences with product features hypothesized to impact system usability, customer satisfaction, and customer loyalty. The models were used to identify actionable items that can lead to measurable improvements in customer experience. SEM is recommended over other modeling methods for its ability to condense large datasets into a few, theory-driven dimensions, its direct modeling of variable covariances, and its missing data management.