Evaluating what is evaluate-able

Quirkiness is good (see below) - especially after the readings that I subjected myself to at the tailend of last week. With education policy and its evaluation being the general theme for which I'm trying to find something sensible to say or a question to ask, I turned to that body of literature - or at least what's available in the LSE library.

Can I just say that it is just about the driest thing I've ever read? But at least it makes some interesting points, including that evaluation of schools, classrooms, students, etc is difficult to do since there are so many variables at work, all interacting with each other than it can be almost impossible to achieve an input-output result. It's all about the process within the school, classroom, by the teacher, etc.

Which has just made my area of research that much harder. How on earth can I assess whether left-wing education policy has made any difference when there are so many variables, factors and different ways of doing evaluation available?