An Approach for Empirical Evaluation of Model-Driven Engineering in Multiple Dimensions

Evaluating methodologies and tools for software development has been subject of research for a while and various quantitative techniques have emerged. However, a common problem in any evaluation method is to select appropriate evaluation criteria. This paper describes the evaluation method and some results of work on empirical evaluation of model-driven engineering and its infusion in industry in the context of industrial participants in the EU research project MODELPLEX. The evaluation method combines qualitative and quantitative assessment, and observations with perceptions. It takes advantage of a combination of a) context-dependent research questions based on the goals of each industrial partner in adopting model-driven engineering and how it is applied, b) the Technology Acceptance Model to summarize the perceptions regarding model-driven engineering, and c) a questionnaire to evaluate opinions regarding specific tools and techniques. The method has been useful in identifying criteria for evaluation and has produced interesting results.