Software metrics have a reputation in industry of not being very useful. I believe the reason for this is that the metrics that have been proposed do not actually measure what their proponents claim they do, that is, they have been presented as "measuring" something that in fact they do not measure.
This situation has come about primarily due to a lack of understanding as to what constitutes "measurement". If we are to provide useful metrics, we must adhere to the principles of measurement, including what constitutes "measurement", criteria for useful metrics, and how to apply metrics properly. In this talk I present arguments to support my view, discuss an approach to one aspect of the problem, namely the use of standard software corpora, and present my experience and results using this approach.
Last modified: Thursday, 28-Jul-2005 17:23:30 NZST
This page is maintained by the seminar list administrator.