Defect Density – Automated Software Testing Metric
Metrics are an important factor when evaluating the effectiveness of an automated software testing effort. Defect density is another well known metric. (See previous blogs on other testing metrics). Defect density is a measure of the total known defects divided by the size of the software entity being measured. This software entity could be a subset or component of the overall system. For example, if there is a high defect density in a specific component, it is important to conduct a causal analysis. The following questions can be asked: Is this functionality very complex and could be the reason for the defect density to be high? Is there a problem with the design/implementation of the functionality? If yes, were the wrong (or not enough) resources assigned to the functionality? The list of questions goes on. Note: There are various ways to measure the size of a software component, such as using function points, SLOCs, features, etc.
D # of known defects
DD = ———- = ( ————————————- )
SS total size of system
- DD = Defect Density
- D = # of known defects
- SS = Total Size of system
When evaluating defect density, the priority and severity of the defect should also be considered. Higher priority requirements are generally weighted heavier. For more details on this metric and other metrics, please see previous blogs or our book Implementing Automated Software Testing.