Empirical analyses of executable acceptance test driven development

  • Authors:
  • Grigori Igorovych Melnik

  • Affiliations:
  • University of Calgary (Canada)

  • Venue:
  • Empirical analyses of executable acceptance test driven development
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

This research investigates the process of Executable Acceptance Test-Driven Development (EATDD) in the context of specifying functional requirements using the FIT framework, when developing line-of-business applications. It is guided by three key research questions: (1) how business and technology experts utilize EATDD in the software development lifecycle; (2) what kind of benefits and limitations EATDD manifests, and (3) to what extent improvements in software quality (if any) are associated with EATDD? The research employs methods of quantitative and qualitative inquiries. Based on the findings of two academic observational studies, one academic quasiexperiment, and three industrial multi-case studies, the following main conclusions are drawn: (1) the use of Executable Acceptance Test-Driven Development is correlated with the enhanced communication in software teams; (2) executable acceptance tests are suitable for specifying functional requirements and are in fact unambiguous, consistent, verifiable, and usable (from both the business experts' and technology experts' perspectives); (3) EATDD provides sufficient evidence of requirements traceability in regulated environments; and (4) current state of tool support negatively impacts the maintainability and scalability of the artifacts produced in the course of EATDD. In addition, our contribution includes emerged conceptualizations and the sociotechnical model of the EATDD process. These not only help to explain the ways how EATDD is used in practice, but also form a base for future investigations.