The Economics of Unit Testing

Authors: Michael Ellims James Bridges Darrel C Ince

Publish Date: 2006/02/18

Volume: 11, Issue: 1, Pages: 5-31

PDF Link

Abstract

Conventional wisdom and anecdote suggests that testing takes between 30 to 50 of a projects effort However testing is not a monolithic activity as it consists of a number of different phases such as unit testing integration testing and finally system and acceptance testUnit testing has received a lot of criticism in terms of the amount of time that it is perceived to take and its perceived costs However it still remains an important verification activity being an effective means to test individual software components for boundary value behavior and ensure that all code has been exercised adequately We examine the available data from three safetyrelated industrial software projects that have made use of unit testing Using this information we argue that the perceived costs of unit testing may be exaggerated and that the likely benefits in terms of defect detection are quite high in relation to those costsWe also discuss the different issues that have been found applying the technique at different phases of the development and using different methods to generate those tests We also compare results we have obtained with empirical results from the literature and highlight some possible weakness of research in this area

Keywords: