Related to yesterday’s post and to an entry I wrote two years ago on measurement, I just found an old (2003) post from Martin Fowler on why we can’t measure productivity in software development. He’s much clearer than I am, and his article is well worth reading. His conclusion:
I can see why measuring productivity is so seductive. If we could do it we could assess software much more easily and objectively than we can now. But false measures only make things worse. This is somewhere I think we have to admit to our ignorance.
(Thanks to Jon / Greg for sending it my way)
I admit I didn’t RTFA, but I cannot help bringing up the dichotomy between philosophical stances with regard to what is knowledge and how to get it, the Positivist stance vs Constructivist stance dichotomy (that I learned about at Steve’s course).
Now, that said, and knowing I’m no real philosopher or epistemologist, the very notion of building theories (as you mentioned at your previous post) and measuring the world to test them and so on, falls nicely into the Postpositivist camp. Of course, there is absolutely nothing “bad” with Positivism, if you are one (and most of us CS and engineering people, indeed are). But, what about a Constructivist approach to the matter?
What if there is no point at all in trying to deduce a general theory of productivity? What kind of research questions would we be posing then? Would then the task of the researcher be rather geared towards describing the processes that generate the need for assessing productivity, the ways it is indeed assessed in the various production contexts, and ultimately the very construction of the notion of productivity itself?
What do you think? Could such an approach be at all useful?
I think it would be useful, though constructivism doesn’t tie you to only these questions.
Just a clarification though — theory building, and even quantitative measurement, isn’t at odds with constructivism, nor with (almost) any other epistemological stance. But you’re right in pointing out that where one falls in the quantification debate depends mostly on their epistemology. For the record I don’t see how positivism would work currently in our field, given our vague, flexible, undefined constructs.