Do Declarative Process Models Help to Reduce Cognitive Biases Related to Business Rules?

Abstract

Declarative process modeling languages, such as Declare, represent processes by means of temporal rules, namely constraints. Those languages typically come endowed with a graphical notation to draw such models diagrammatically. In this paper, we explore the effects of diagrammatic representation on humans’ deductive reasoning involved in the analysis and compliance checking of declarative process models. In an experiment, we compared textual descriptions of business rules against textual descriptions that were supplemented with declarative models. Results based on a sample of 75 subjects indicate that the declarative process models did not improve but rather lowered reasoning performance. Thus, for novice users, using the graphical notation of Declare may not help readers properly understand business rules: they may confuse them in comparison to textual descriptions. A likely explanation of the negative effect of graphical declarative models on human reasoning is that readers interpret edges wrongly. This has implications for the practical use of business rules on the one hand and the design of declarative process modeling languages on the other.