Experience from Replicating Empirical Studies on Prediction Models

Authors:
Magnus C. Ohlsson;Per Runeson
Affiliations:
-;-
Venue:
METRICS '02 Proceedings of the 8th International Symposium on Software Metrics
Year:
2002

Citing 0
Cited 11

Fault Prediction Modeling for Software Quality Estimation: Comparing Commonly Used Techniques

Empirical Software Engineering
Investigation into the exploitation of Object-Oriented features

ACM SIGSOFT Software Engineering Notes
Comparative Assessment of Software Quality Classification Techniques: An Empirical Case Study

Empirical Software Engineering
Assessment of a New Three-Group Software Quality Classification Technique: An Empirical Case Study

Empirical Software Engineering
Resource-oriented software quality classification models

Journal of Systems and Software
Experiences and results from initiating field defect prediction and product test prioritization efforts at ABB Inc.

Proceedings of the 28th international conference on Software engineering
Object-oriented software fault prediction using neural networks

Information and Software Technology
Software quality estimation with limited fault data: a semi-supervised learning perspective

Software Quality Control
Aggregating performance metrics for classifier evaluation

IRI'09 Proceedings of the 10th IEEE international conference on Information Reuse & Integration
Approximating deployment metrics to predict field defects and plan corrective maintenance activities

ISSRE'09 Proceedings of the 20th IEEE international conference on software reliability engineering
Unsupervised learning for expert-based software quality estimation

HASE'04 Proceedings of the Eighth IEEE international conference on High assurance systems engineering

Quantified Score

Hi-index	0.00

Visualization

Abstract

When conducting empirical studies, replications areimportant contributors to investigate the generality of thestudies. By replicating a study in another context, it isinvestigated which impact the specific environment has,related to the effect of the studied object. In this paper, wedefine different levels of replication to characterise the similaritiesand differences between an original study and areplication with particular focus on prediction models foridentification of fault-prone components. Further, wederive a set of issues and concerns which are important inorder to enable replication of an empirical study and toenable practitioners to use the results. To illustrate theimportance of the raised issues, a replication case study ispresented in the domain of prediction models for fault-pronesoftware components. It is concluded that the resultsare very divergent depending on how different parametersare chosen, which demonstrates the need for well documentedempirical studies to enable replication and use.