Influence Of Team Size And Defect Detection Technique On Inspection Effectiveness

  • Authors:

  • Stefan Biffl;Walter Gutjahr

  • Affiliations:

  • -;-

  • Venue:
  • METRICS '01 Proceedings of the 7th International Symposium on Software Metrics
  • Year:
  • 2001

Quantified Score

Hi-index 0.00

Visualization

Abstract

Inspection team size and the set of defect detection techniques used by the team are major characteristics of the inspection design, which influences inspection effectiveness, benefit and cost. In this work we focus on the inspection performance of a nominal, that is non-communicating, team, similar to the situation of an inspection team after independent individual preparation. We propose a statistical model based on empirical data to calculate the expected values for the inspection effectiveness and effort of synthetic nominal teams. Further we introduce an economic model to compute the inspection benefits, net gain, and return on investment. With these models we determine a) the best mix of reading techniques (RTs) to maximize the average inspection performance for a given team size, b) the optimal team size and RT mix for a given inspection time budget, and c) the benefit of an additional inspector for a given team size. Main results of the investigation with data from a controlled experiment are: a) Benefits of an additional inspector for a given RT diminished quickly with growing team size. Thus, above a given team size a mix of different RTs is more effective and has a higher net gain than using only one RT; b) The cost-benefit model limits team size, since the diminishing gain of an additional inspector at some point is more than offset by his additional cost.