Hacker News Reader

Comment by qq66 | Hacker News Reader

You can't control for "bright parents." You can control for parents' income, parents' education, parents' height and weight, but you can't control for the kind of attitude towards childrearing and towards life that would cause a couple to send their kids to Montessori school.

This is why so many social science results that are "controlled" for various variables actually don't make any sense.

View on HN

Wouldn't accepting 1000 applications, and teaching Montessori-style to 500 of them and traditional-style to the other 500 pretty accurately control for all of those self-selection biases?

Only if you put the 500 non-Montessori children in a separate school, with the same level of funding as the Montessori school. From the little that I know about these studies, at least some of the differences could be explained by the mixing of the 'control group' into an environment that is less ambitious or focused as the Montessori group.

(I walk pass 2 schools on my way to work each day - one Montessori, one not, sometimes around the time where children are dropped off by their parents. Even not knowing anything about the relative merits of the systems, I can easily tell which school I'd want to send my daughter to in a few years time; at the risk of sounding elitist, it's like standing next to the exit of an opera performance vs a soccer stadium).

To add to this, The Steiner system places a lot of value in surroundings and buildings and a lot of time and effort (or is it money?) goes into making the "learning environment" inspiring and stimulating.... Etc etc. Basically the ones I've been to look fantastic.

Reply on news.ycombinator.com