Software Metrics
Software metrics are measures of the quality of software. They typically include measures of the complexity, understandability, testability, description and intricacy of code. The main function of software metrics is to improve the quality of computer software. In the past software metrics has focused on reliability and quality of the final software product but now the emphasis is moving more towards management of the software creation process as well as evaluation of intermediate software products. This is probably because software is becoming more and more complex and the human hours needed to write such complex software is ever increasing. In order to make the most effective use of time and resources many developers are seeking successful metrics that can be employed to facilitate such activities and make the most of the resources available to them.
One of the first software metrics was the number of lines of source code or the very similar delivered source instructions. It is not surprising that this was one of the first metrics since there are automated tools to measure the number of lines of code and the value of the result is inherently obvious to experienced programmers who are usually able to give reasonably accurate estimates of the number of lines of code necessary for a specified task before the code is written.
Measures in general are categorized by measurement theory into direct and indirect measures. Direct measures are those whose outcome is independent of the measurement outcome of any other attribute whereas an indirect measure is one whose outcome is inherently linked to the measurement of one or more other attributes. The number of lines of code is a direct measure. One of the main criticisms of the metrics extracted by measuring the number of lines of source code was that these became available only at a late stage of the project's development hence are of limited value when planning the crucial early phases of a project.
In an attempt to gain a better evaluation handle on software creation another approach utilizing software metrics divided into product and process metrics has been developed. They are metrics that are involved in the measurements of products created in the process and those resources used to create the products. Product metrics are often called design metrics as well and focus on the complexity, size and robustness of the software design being used for a specific application. They measure characteristics related to the source code, final executable program as well as analysis or design documentation. Process metrics, or project metrics as they are often called, are those metrics designed to quantify issues relating to management such as application size, staffing size and human effort expended over the process, scheduling of deliverables, number of defects, and the costs associated with fixing those defects. From these metrics it is clear that the measure of software quality is complex in that it is multidimensional, hierarchical and sometimes abstract. Because of this it is likely that no single metric is ever likely to be completely and accurately descriptive as a single expression of software quality.
This is the complete article, containing 512 words
(approx. 2 pages at 300 words per page).
