The Effect of Developers’ General Intelligence on the Understandability of Domain Models: An Empirical Study

Abstract

Software engineering has traditionally focused on the semantic and notational aspects of domain model understandability, overlooking the cognitive factors involved in this process. This study addresses this knowledge gap by exploring the role of General Intelligence, a cognitive factor associated with comprehensibility and problem-solving abilities, in the understandability of domain models. Existing literature has shown that understandability is not a one-dimensional concept, but rather involves multiple levels of comprehension, from surface-level understanding to deeper-level problem-solving abilities. However, these studies have often relied on subjective measures of comprehension, highlighting the need for more objective, quantifiable measures. In response to this, we conducted an empirical study with 102 participants from the University of Alicante, measuring their General Intelligence using the D48 test as proposed by Spearman’s Bifactorial Theory. Participants were then tasked with performing a series of understandability tasks on UML domain models. We also examined the impact of model understandability performance on the intention to adopt the model, using the UMAM-Q test. Our research methodology involved a two-way analysis of variance. The results confirmed that higher intelligence leads to better model understandability performance and that those with higher model understandability have a greater intention to adopt the model. This study underscores the need to consider cognitive factors in software engineering and provides a new perspective on improving domain model understandability.

Keywords

• Model understandability
• general intelligence
• human factors
• intention of adoption
• domain modeling
• empirical software engineering