Empirical analysis of factors affecting confirmation bias levels of software engineers

Calikli, G. and Bener, A. (2015) Empirical analysis of factors affecting confirmation bias levels of software engineers. Software Quality Journal, 23(4), pp. 695-722. (doi: 10.1007/s11219-014-9250-6)

Full text not currently available from Enlighten.

Abstract

Confirmation bias is defined as the tendency of people to seek evidence that verifies a hypothesis rather than seeking evidence to falsify it. Due to the confirmation bias, defects may be introduced in a software product during requirements analysis, design, implementation and/or testing phases. For instance, testers may exhibit confirmatory behavior in the form of a tendency to make the code run rather than employing a strategic approach to make it fail. As a result, most of the defects that have been introduced in the earlier phases of software development may be overlooked leading to an increase in software defect density. In this paper, we quantify confirmation bias levels in terms of a single derived metric. However, the main focus of this paper is the analysis of factors affecting confirmation bias levels of software engineers. Identification of these factors can guide project managers to circumvent negative effects of confirmation bias, as well as providing guidance for the recruitment and effective allocation of software engineers. In this empirical study, we observed low confirmation bias levels among participants with logical reasoning and hypothesis testing skills.

Item Type:Articles
Status:Published
Refereed:Yes
Glasgow Author(s) Enlighten ID:Calikli, Dr Handan Gul
Authors: Calikli, G., and Bener, A.
College/School:College of Science and Engineering > School of Computing Science
Journal Name:Software Quality Journal
Publisher:Springer
ISSN:0963-9314
ISSN (Online):1573-1367
Published Online:06 September 2014

University Staff: Request a correction | Enlighten Editors: Update this record