A 43 page paper that examines the philosophy, need and objectives for software metrics. This paper contains a detailed description of the definition and classification of software metrics. Although there are several complicated formulas in the definition of the various metrics, including reliability calculations, the subject is presented in a logical and clear manner.

The definition of defect removal efficiency is the percentage of the defect potentials that will be removed before the software application is delivered to its users or customers. This article goes a little deeper into this subject by discussing several approaches to defect removal including the personal Software Process (PSP), from the SEI as well as JAD sessions, design and code inspections. A readable and comprehensive coverage of the subject.

This, 7 page, white paper presents a good overview of metrics management for software. This paper looks at Requirements, Design, Coding, Testing and maintenance phases in order to put the appropriate software metrics in context of the whole SDLC. A good introduction to the definition and use of software metrics.

This, easy to read, paper, examines different quality models including McCall’s quality model (1977), Boehm’s quality model (1978), the FURPS quality model (Robert Grady), Doremey’s quality model as well as the ISO 9126 standard. Although there is no single right way of breaking down software quality characteristics the presentation and comparison on the various quality models is insightful to the model’s intended use and value.

A list of heuristics for software testability, including controllability (for automated testing) for which the ability to independently test modules or objects is listed. Although this paper is presented as a list of bullet points under the main testability headings, the points are easy to understand and the list is useful as a checklist for inspections that could be undertaken early in the SDLC.

There is a useful breakdown of software metrics against the management functions of planning, organizing, controlling and improving. This breakdown of the management functions that can use metrics is a useful perspective as several other introductory articles focus mainly on the quality aspect of metrics, with regard to process improvement.

There have been many attempts to break down software into characteristics that can be measured in terms of overall software quality. The ISO has looked into the problem of classification of quality characteristics and the ISO 9126 standard is one of their earlier solutions to this problem.

In the defect density calculation, see articles in this directory, the size of the software needs to be known. One method of calculation size is Lines of Code (LOC), there are no reference to LOC  as a measurement technique, in this directory, because it is just the lines of code. Another measure of the overall size of software is Function Point Analysis (FPA) and that is what is presented in this article from the Q\P management Group Inc.

This paper was taken from a (2006) CIO symposium on software best practices and earns a place in the software testing directory based on its straight forward presentation style that conveys a technical subject to a broad audience. The presentation style of this paper is conversational and begins with a brief history of software metrics then delves into the pro and cons of using traditional metrics to predict future trends.

This white paper, 28 pages, is more on the software engineering technical side but has sections on interface testing which includes GUI, J2EE, SOAP and other main stream application development components. There are interesting discussions on automation, including the problem of ‘custom controls’ which is an issue for most of the GUI automation toll kits. 

We could define defect density here, OK let’s do it; “total known defects divided by the size of the software measured”. That said, there are a number of questions, such as how do you measure size and what is this measure used for. Those type of questions are answered in this small informative article.

This paper presents an overview of measuring program (software) complexity. This is a useful presentation of a complicated subject and the paper not only presents the popular McCabe Metric but discusses potential problems with using this metric. In a similar fashion Halstead’s approach is also presented, with potential problems of using that approach. Overall a good and insightful introduction and discussion on the technical subject of software complexity.

This is pretty much a complete book on a structured methodology that seeks to systematically test software based on analysis of the cyclomatic complexity of the software. Although the presentation is a little too technical, for the average applications tester, the relationship between software complexity and software testing is comprehensively presented.

A presentation at the International Function Point User Group (IFPUG) conference in the spring of '94. This paper does a good job of presenting FPA, although I don’t know why the title has ‘forgotten aspects’ in it as it stands out as a good primer on the subject for anyone. I think the intention of the title was to present an old subject to a knowledgeable audience, anyway this is one of the best presentations on the subject we could find.