Object-Oriented Design Quality Adaptation: A System Dynamics Simulation

Abstract

Despite the increasing interest of the software engineering research community in assessing and improving quality at the early phases of the software development lifecycle such as the design phase, less attention is devoted to adapting software quality to changing environment conditions especially at the design stage. Design quality can be significantly impacted when design decisions are modified due to changes in requirements or design strategies. This research sheds light on possible adaptation mechanisms that can effectively mitigate any decrease in object-oriented design quality due to a particular design change. To forecast the impact of design changes and possible adaptations on design quality, a system dynamics simulation is developed in Powersim®. The simulation variables are grouped into sub-models and represented by the quality factors of the Quality Model for Object Oriented Design (QMOOD) developed by Bansiya and Davis. Each sub-model simulates the interactions between a specific QMOOD quality attribute and its corresponding design properties as well as design metrics. The simulation is developed by following Pfahl and Ruhe’s System Dynamics Model. If after a design change, the simulated design quality decreases below a defined reference value, designers can apply one of the suggested adaptation equations that are extracted from the QMOOD quality attributes equations. The simulation is validated by applying design changes and their adaptations on real academic designs and computing correlations between the simulated and the real values of each QMOOD quality attribute after adaptation. High correlations are obtained for all the quality attributes, which shows the effectiveness of the adaptation mechanisms in adjusting design quality.