Improving efficiency of change impact assessment using graphical requirement specifications: An experiment

Abstract

Objective: Graphical requirements representation is often considered needed to advance model-driven development. Dedicated modelling languages include formalisms for graphically representing requirements, and together with new methods for structuring requirements, graphical modelling promises improvements such as more efficient change management. This paper examines whether the use of a graphical notation of a requirements affects the task of assessing the impact of a proposed change to a requirements specification. Method: The efficiency of using a graphical requirements representation was examined through an experiment - using 18 student subjects. Time, perceived confidence and accuracy were measured as dependent variables. Result: The results showed that using a graphical representation decreased the time required and increased the perceived confidence, but the accuracy decreased. However, the statistical analysis of the results showed that only the difference in time was significant. Furthermore, there was a large difference in variance within the dependent variables between the groups. © 2010 Springer-Verlag.