Factors Affecting Comprehension of Contribution Links in Goal Models: An Experiment

• Published in: Conceptual Modeling pp. 525 - 539
• Main Authors: Liaskos, Sotirios, Tambosi, Wisal
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Conceptual modelling; Experimental study; Goal models; Model comprehension
• ISBN: 9783030332228; 3030332225
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-33223-5_43

Goal models have long been regarded to be useful instruments for visualizing and analysing decision problems. Key to using goal models for the purpose is the concept of satisfaction contribution between goals. Several proposals have been offered in the literature for representing contributions and performing inferences therewith. Theoretical arguments and demonstrative examples are typically used to support the usefulness and soundness of such proposals. However, the degree to which users of goal models intuitively understand the meaning of a specific contribution representation and use it for making valid inferences constitutes an additional measure of the appropriateness of the representation. We report on an experimental study to compare the intuitiveness of two alternative contribution representation approaches via measuring the degree to which untrained users perform inferences compliant with the semantics defined by the language designers. We further explore the role of individual differences such as cognitive style and attitude and ability with arithmetic in establishing and applying the right semantics. We find significant differences between the representations under comparison as well as effects of various qualities and levels with regards to individual factors. The results inspire further research on the specific matter of contribution links and support the overall soundness and operationalizability of the intuitiveness construct.