PBR vs. checklist: a replication in the n-fold inspection context

  • Authors:

  • Lulu He;Jeffrey Carver

  • Affiliations:

  • Mississippi State University, MS;Mississippi State University, MS

  • Venue:
  • Proceedings of the 2006 ACM/IEEE international symposium on Empirical software engineering
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Inspection is considered a powerful method to check software documents for defects. Many published work shows that inspections in requirements specification phase are particularly effective and efficient. Perspective-Based Reading (PBR) is one of the systematic techniques to support defect detection in requirements documents. In this paper we describe an experiment to validate the effectiveness of PBR in a meeting-based N-fold inspection. Our goals were: (1) re-test the hypothesis of the original experiment that PBR helps to increase individual and team defect detection effectiveness compared to an checklist approach; (2) investigate the different impact of PBR and checklist on the effectiveness of N-fold team meeting; and (3) investigate some interesting characteristics of PBR (e.g. the relationship between background experiences and performance of the subjects). The results of the study showed that PBR was significantly more effective than checklist (supporting the original study). We also found that the team meeting is much more important for checklist teams, based on the number of meeting gains and the number of false defects eliminated. Finally, we found that teams using the PBR techniques have less overlap in their defect detection than those using checklist. The ultimate goal is to provide best practices (guidance) for applying PBR in software inspection and also some advice for PBR (or software inspections) process improvement.