There is a very interesting book handling the topic agile software engineering and metrics, "Agile Management for Software Engineering". Its worth to be read.
At the moment there are not clearly defined metrics for evaluating the success of software projects using agile methods.
Why should the metrics for Agile methods be different from those for other development processes?
At the moment there are not clearly defined metrics for evaluating the success of software projects using agile methods. Therefore...
One possibility might be to study the spread, or speed of adoption, of agile methods, instead of their direct effectiveness. One can assume that speed of adoption is related to effectiveness (I know it is a stretch...). Word counts, article titles, regional surveys repeated over three years or five years. Number of conferences. Number of books. How often listed in resumes, etc...
Software is not like anything else, so it needs a whole new set of tools; trying to use the old ones is just holding us back.
I would take the initial statement and restate it as:
"At the moment there are not clearly defined metrics for evaluating the success of software projects."
THere is an interesting article about the RunningTestedFeaturesMetric? on http://www.xprogramming.com/xpmag/jatRtsMetric.htm -- MirkoBlueming
Hear, hear!