Further Experiences with Scenarios and Checklists

  • Authors:

  • J. Miller;M. Wood;M. Roper

  • Affiliations:

  • Dept. Computer Science, University of Strathclyde, Scotland. E-mail james@cs.strath.ac.uk;Dept. Computer Science, University of Strathclyde, Scotland;Dept. Computer Science, University of Strathclyde, Scotland

  • Venue:
  • Empirical Software Engineering
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

Software inspection is one of the best methods of verifyingsoftware documents. Software inspection is a complex process,with many possible variations, most of which have received littleor no evaluation. This paper reports on the evaluation of onecomponent of the inspection process, detection aids, specificallyusing Scenario or Checklist approaches. The evaluation is bysubject-based experimentation, and is currently one of threeindependent experiments on the same hypothesis. The paper describesthe experimental process, the resulting analysis of the experimentaldata, and attempts to compare the results in this experimentwith the other experiments. This replication is broadly supportiveof the results from the original experiment, namely, that theScenario approach is superior to the Checklist approach; andthat the meeting component of a software inspection is not aneffective defect detection mechanism. This experiment also tentativelyproposes additional relationships between general academic performanceand individual inspection performance; and between meeting lossand group inspection performance.