The Impact of Confirmation Bias on the Release-based Defect Prediction of Developer Groups

Calikli, G. and Nemer, A. (2013) The Impact of Confirmation Bias on the Release-based Defect Prediction of Developer Groups. In: 25th International Conference on Software Engineering and Knowledge Engineering (SEKE 2013), Boston, MA, USA, 27-29 Jun 2013, pp. 461-466. ISBN 9781510841598

Abstract

During software development life cycle (SDLC), source codes are created and updated by groups of one or more developers. Information about the defect rates introduced by developer groups for the current release of a software product might guide project managers to form developer groups in order to manage defect rates for the next releases. In this research, we use partial least squares regression (PLSR) and principal component regression (PCR) to model the relation between defect rates and a specific cognitive aspect of developers, namely confirmation bias. In order to empirically estimate the performance of our model, we use datasets from three industrial software projects.

• Item Type: Conference Proceedings
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Calikli, Dr Handan Gul
• Authors: Calikli, G., and Nemer, A.
• College/School: College of Science and Engineering > School of Computing Science
• ISBN: 9781510841598