I’m re blogging this because: I agree that qualitative research is effective depending on your research question, and especially if relating to intangibles; I agree that absolute-hunting isn’t healthy; and, I think that looking at students’ books shouldn’t be to assess the marking of a teacher, but to use content analysis to examine the extent of transformation in student learning over time.
While I take and understand the point about confirmation bias, it is absolutely the case the Dylan Wiliam articulated in one hour everything I have been banging on about in tweets and blogs about educational research over the past couple of years. The only difference being that while Dylan is an internationally regarded educator, that I am a totally insignificant figure, even in my own school (or house – ask my wife).
It was the section about meta-analysis that really struck a cord. Dylan showed in ten minutes the major difficulties with this method of analysis. Put simply, you cannot compare like for like unless you drill down to the exact thing you are investigating. Thus, it is pointless comparing 100 papers on feedback if 33 refer to immediate verbal feedback, 33 to written feedback, or 34 to peer feedback (not sure if that last thing exists, but it sounds ok)…
View original post 819 more words