Evaluating Software Modelling Recommendations : Towards Systematic Guidelines for Modelling

Abstract

Background: Despite having several advantages, software modelling remains unpopular for developers. Similarly, university students do not see the benefits of software modelling in the university curriculum. Prior research show the lack of guidance for students to do so. Aims: We aim to evaluate the effectiveness of four modelling recommendations made in related work to improve student modelling knowledge. Additionally, we aim to discover students' perceptions of software modelling after taking a course with the recommendations included. Method: We conducted a mixed method study, including interviews with teaching assistants, student surveys, and a focus group study involving students, teaching assistants, and experts from both modelling and education. Results: We find that the four recommendations overall have a positive impact as they help students better understand the modelling knowledge from the course. Students express that specific recommendations help them grasp the concept of software modelling well. We also extend the recommendations by adding more details specific to software modelling and solidifying the recommendations into systematic guidelines. Conclusions: The guidelines can potentially enhance education and training in software modelling, catering to both academic settings and industrial environments. Additionally, the guidelines contribute to improved communication between students and the course itself by outlining what students can expect from modelling assignments and the value inherent in each of these assignments.