The influence of cyclomatic complexity distribution on the understandability of xtUML models

Executable software models formalize functional software requirements. This means that the understandability of software models is of paramount importance. In this paper, we investigate the influence of cyclomatic complexity on the understandability of executable translatable unified modeling language (xtUML) models. We adapt traditional cyclomatic complexity metrics to different xtUML sub-models and present two different ways of measuring complexity distribution: horizontal, among elements of the same type, and vertical, among elements of different types. In order to test our hypothesis that cyclomatic complexity distribution influences the understandability of xtUML models, we have performed an experiment with student participants in which we have evaluated the understandability of three semantically equivalent xtUML models with different complexity distributions. Results indicate that a better distribution of cyclomatic complexity has a positive influence on model understandability.