Declarative process models offer flexibility in representing complex processes, but can be difficult to understand due to their implicit control flow. Improving their representation, particularly by investigating and enhancing their semantic transparency, can help address this challenge. This paper proposes an empirical research model to examine how varying levels of semantic transparency, encoded in different visual notations of declarative models, affect users’ cognitive load and visual behavior in the presence and absence of contextual information. The findings will lead to a better understanding of the effects of semantic transparency and will support the design of understandable declarative models.