Interesting discussion over at the LinkedIn group Software Testing & Quality Assurance about metrics and whether or counting defects is useful. Personally I don’t think it is, at least not without context. There are a few responses that seem to take this to mean “Aha, so you don’t count defects so how do you know how many you have” but that’s not what’s being discussed.

Michael Bolton (no, not the singer) has weighed in on the argument, something that usually makes me take an interest in the otherwise disappointing dross of posts that make up the topics discussed. One of his posts pointed to a resource he maintains on metrics. Hitting a few of the links leads to some quite interesting information resources. Go educate yourself, starting with Software Engineering Metrics: What Do They Measure and How Do We Know? Written by Kaner and Bond, the pdf goes into construct validity, a term I hadn’t come across before. I think I need to do more reading.

Curtis Stuehrenberg (he’s presenting at CAST 2012) also pointed the group towards a paper that gives an overview of qualitative research methods. More reading material.