Videos vs. Use Cases: Can Videos Capture More Requirements under Time Pressure?

摘要

[Context and motivation] Many customers and stakeholders of real-world embedded systems are difficult to reach with traditional requirements elicitation and validation techniques. Traditional requirements engineering methods do not deliver concrete results for validation fast enough; stakeholders get impatient or misunderstand abstract requirements. [Question/problem] The problem is to achieve a mutual understanding between customers and the requirements engineer quickly and easily, and to get stakeholders involved actively. [Principal ideas/results] We propose to use ad-hoc videos as a concrete representation of early requirements. Videos have been used before in requirements engineering: Sophisticated videos were created at high effort. We show, however, that even low-effort ad-hoc videos can work comparably or better than use cases for avoiding misunderstandings in the early phases of a project. [Contribution] We replicated and refined an experiment designed using the Goal-Question-Metric paradigm to compare videos with use cases as a widely used textual representation of requirements. During the experiment, even inexperienced subjects were able to create useful videos in only half an hour. Videos helped to clarify more requirements than use cases did under the same conditions (i.e. time pressure).