Modularity in Process Models: Review and Effects

The use of subprocesses in large process models is an important step in modeling practice to handle complexity. While there are several advantages attributed to such a modular design, including ease of reuse, scalability, and enhanced understanding, the lack of precise guidelines turns out to be a major impediment for applying modularity in a systematic way. In this paper we approach this area of research from a critical perspective. Our first contribution is a review of existing approaches to process model modularity. This review shows that aside from some limited insights, a systematic and grounded approach to finding the optimal modularization of a process model is missing. Therefore, we turned to modular process models from practice to study their merits. In particular, we set up an experiment involving professional process modelers and tested the effect of modularization on understanding. Our second contribution, stemming from this experiment, is that modularity appears to pay off. We discuss some of the limitations of our research and implications for future design-oriented approaches.