and the montessori method is effective in lowering disruptive peer-behaviour. it's part of the point. it teaches children to not be disruptive by letting them focus on their activities.
These type of studies trot the line of cargo cult though. Incredibly small effects, weak causation, full of possible confounders.
I'm not going to say being in a class where you are trying to pay attention and others are being very disruptive, and interrupt the lesson is enjoyable, it's annoying, but if you take even the studies you link, say the second link, it finds a 2% correlated effect, that is peers had scores 0.02 times lower than the standard deviation.
So if we were to change and group kids based on disruptiveness, instead of a 80% test core, your kid would have a 79.7% test score...
Now before you respond to this, I want to reiterate the point of my argument, that none of these ideas focus on actual teaching method improvements. How do you take a child at any level, and more effectively teach them so they learn faster and improve their intelligence and knowledge.
These alternatives, grouped by disruptiveness, grouped by current abilities, etc. they don't really change the pedagogy, just the environment. It seems their known effects are really small, and the effect on the average are not known.
So I'm not against them, as just from a pure setup they seem more appropriate, but it seems unlilely to result in much improvement learning wise, the kind that I'd be interested in.
