Effect on Internal Quality Indicators

Abstract

According to Bansiya [18], internal quality indicators influence external quality attributes and, therefore, evaluating a product’s internal characteristics is reasonable. As a result, some useful conclusions can be drawn about the product’s external quality attributes on the basis of its internal characteristics [18]. Relying on Briand et al. [31], measures of structural design properties are considered to be indicators of external system quality attributes, such as reliability or maintainability. BØegh [30] mentioned class-level metrics proposed by Chidamber and Kemerer (CK metrics) [43] as typical examples of internal measures. This chapter concentrates on the CBO , WMC and (RFC) metrics (from the CK metrics suite) as their suitability for assessing fault proneness and fault content has already been empirically confirmed (see Sect. 3.3.2.2). The question is whether those metrics, often called design complexity metrics [238], are influenced by the TF practice.

It is important to note that average values of class-level code metrics (i.e. $\mathrm{CBO}_{\mathrm{Mean}}$ , $\mathrm{WMC}_{\mathrm{Mean}}$ , $\mathrm{RFC}_{\mathrm{Mean}}$) have been calculated for each project and analysed in this section. Another approach would be a class level analysis violating the assumption of independent observations. Furthermore, to keep the book concise and, simultaneously, present the most essential results for the meta-analysis conducted in Chap. 9, this chapter is focused on the TLSP vs. TFSP selective analysis (see Sect. 4.7.7). It is also worth mentioning that none of the collected pre-test results seemed to be a good candidate to include in the model as a covariate because none of them was related to internal code quality concepts like, for example, coupling.

Anything you need to quantify can be measured in some way that is superior to not measuring at all.

Tom DeMarco and Timothy Lister [61]