Some interesting stuff put together in this post by Ewan McIntosh. He first describes "a school inside an old people’s home [where] early-years students participating in the school’s reading programme frequently sit with the old people to read together." It's a great idea and probably helps both old and young. Yet, by contrast, politicians and administrators focus on curriculum and assessment, ignoring (says McIntosh) what research already tells us works in pedagogy. Such as, for example, "formative assessment–student-initiated, self, and peer assessment [which is] is far more effective at raising test scores than teaching to the test." Those making the decisions, says McIntosh, miss "the trap that is set for them" - "I haven’t heard one piece of discourse on formative assessment in the U.S. in 2011 that actually shows an understanding of what it is." If there is assessment, they expect to see a grade - they have been pushed into a frame of expecting these assessments to produce data or facts to prove that the pedagogy works. Except - what counts as data varies a lot, what counts as evidence depends on what you're trying to prove and what you already believe, and what you're likely to accept may depend as much on a recommendation from a friend as on anything else.