Abstract
Context
Eliciting user stories is a major challenge for agile development approaches. Conceptual models are used to support the identification of user stories and increase their understanding. In many companies, existing model documentation stored as either use cases or BPMN models is available. However, these two types of business process models might not be equally effective for elicitation tasks due to their formats.
Objective
We address the effectiveness of different elicitation tasks when supported either with visual or textual conceptual model. Since the agile literature shows little attention to reusing existing BPMN documentation, we propose several hypotheses to compare it to the use of textual use case models.
Method
We conducted an experiment to compare the effectiveness of the two business process formats: textual use cases and visual BPMN models. We studied their effects on three elicitation tasks: identifying user stories and understanding their execution-order and integration dependencies.
Results
The subjects better understood execution-order dependencies when visual input in the form of BPMN models was provided. The performance of the other two tasks showed no statistical differences.
Conclusion
We addressed an important problem of user story elicitation: which informationally equivalent model (visual BPMN or textual use case) is more effective when identifying and understanding user stories.
Eliciting user stories is a major challenge for agile development approaches. Conceptual models are used to support the identification of user stories and increase their understanding. In many companies, existing model documentation stored as either use cases or BPMN models is available. However, these two types of business process models might not be equally effective for elicitation tasks due to their formats.
Objective
We address the effectiveness of different elicitation tasks when supported either with visual or textual conceptual model. Since the agile literature shows little attention to reusing existing BPMN documentation, we propose several hypotheses to compare it to the use of textual use case models.
Method
We conducted an experiment to compare the effectiveness of the two business process formats: textual use cases and visual BPMN models. We studied their effects on three elicitation tasks: identifying user stories and understanding their execution-order and integration dependencies.
Results
The subjects better understood execution-order dependencies when visual input in the form of BPMN models was provided. The performance of the other two tasks showed no statistical differences.
Conclusion
We addressed an important problem of user story elicitation: which informationally equivalent model (visual BPMN or textual use case) is more effective when identifying and understanding user stories.
| Originalsprache | Englisch |
|---|---|
| Aufsatznummer | 106169 |
| Fachzeitschrift | Information and Software Technology |
| Jahrgang | 116 |
| DOIs | |
| Publikationsstatus | Veröffentlicht - 2019 |
Österreichische Systematik der Wissenschaftszweige (ÖFOS)
- 502050 Wirtschaftsinformatik