Shazam! Check the data!!!

As I reflected on this to figure out the gain from 2011 to 2012, I realized that even though my instruction didn't change, there was a big difference that is still consistent with flipped learning:

In 2011, the kids had only the online book as review material outside of class (at at that point, they were unfamiliar with it).

In 2012, I curated some video/study material and posted it here and in Haiku. So, even though I wasn't on the flipped class model at that time, that could be enough of a difference to explain the gain in the average scores.