It would be interesting to know if length of school day correlates with diagnosis rate (controlling for confounding covariates), or if, e.g., Montessori schools have fewer diagnoses than public schools (again controlling for confounding covariates like family income, etc.).
Controlling for confounding factors in a study like this would be very close to impossible. Income is one factor, but parental involvement, household stability, discipline methods (I don't have data to support this, but I have a feeling that the average discipline methods differ greatly between Montessori kids and public school kids), and many others come into play. It'd be interesting if someone could pull it off, but this would be an unbelievably difficult study.
