Improving the identification of functional system requirements when novice analysts create use case diagrams: the benefits of applying conceptual mental models

Abstract

Planning an information system (IS) is challenging, especially for inexperienced IS analysts, including students. One area that often causes particular difficulty is defining the requirements of the future system—a phase that is crucial to the success of any IS project. While serving as advisers for students carrying out their first IS projects, we observed that “imagining” the future system’s functionality and workflow from the user’s perspective makes it easier for students to identify the system’s functional requirements. This insight led us to suggest refining the requirements phase by structuring the “imagining” process into a formal Conceptual mental model (CMM)—that is, a tangible visual representation of the user’s beliefs and expectations (i.e., the user’s mental model) about the system to be developed. We expect that this approach, taken from the domains of cognitive psychology and human–computer interaction, will improve the ability of inexperienced IS analysts to identify the functional requirements of the future system, along with their associated relations and actors. In particular, we expect that this approach will yield better Use case diagrams (UCDs)—a key component of the Use case model for representing and then specifying system requirements, relations and actors. Here, we report on a controlled experiment testing whether creation of a CMM prior to the creation of a UCD improves the identification of functional system requirements, relations, and actors in terms of correctness, completeness, and irredundancy of the UCD. We found that participants who produced a UCD after producing a CMM performed significantly better in defining system requirements, relations and actors, as expressed in their UCDs, compared to participants who produced a UCD without first producing a CMM. We conclude that our suggested refinement improves the performance of novice IS analysts in capturing, analyzing and defining functional system requirements, relations and actors in a UCD, in terms of correctness, completeness, and irredundancy.