R Pawson

This is by far the sharpest book I've read on the methodology of social science. The author presents an admirably broad perspective on measurement, beginning with natural science experiments and moving onto their nearest equivalent in social science; interviews. But he also dismantles some classic problems in philosophy of science more convincingly than most philosophers manage to do. His basic idea is that natural science experiments don't test theoretical ideas one at a time: instead they link every new idea to a whole network of older theories in the measurement process. The author believes that this scheme, when modified in a suitable way, is applicable in social science as well. I don't know to what extent this is true, but the author certainly brings the discussion down to a very practical level at the end, which is quite admirable. All in all a very good book that all social science practitioners and theorists should study.… (meer)