- 19.12 (released) ...
Attributes, metrics and measures
The most general notion is "attribute":
Examples of attributes include reliability (a product attribute, non-quantitative) and total project cost (process, quantitative).
A metric is simply a quantitative attribute:
Examples of metrics include the number of source lines of a program (product) and the total cost of a project (process). The metric tool provides by default a set of metrics, they are available in the Metric Evaluation tab of the metric tool
Attributes other than metrics will be called "qualitative":
An example of qualitative attribute is the reliability of a software product.
The "process" vs. "product" distinction carries over to metrics:
"Relevance", as defined in the previous section, suggests that the purpose of metrics is to help us gain information about attributes that are of direct interest to us. Often these will be qualitative; for example we may want to estimate the reliability of our software. Metrics provide us with numerical values that can serve to assess or predict such attributes.
Applying a metric will give us measures: