Jaga: It's hard to really get a clear snapshot of something as complex as teaching and learning in any given study. So one of the questions is how do you decide which ones to look at? How do you go about deciding your inclusion criteria for these many big large-scale multiple analyses? At the level of meta analysis, this is where someone takes many studies and then asks what's the average effect? And remember, when Jing Glass invented this procedure, his obsession was not just about the average. His obsession was about the variability around the average. Jaga: What works best tends to work best with special needs kids with gifted kids with five year olds with 15 year olds with kids