Since our daughter will be kindergarten age in the fall, Elizabeth and I have been spending a lot of time researching public schools in the area. Our primary figure of merit has been Massachusetts Comprehensive Assessment System - MCAS - test distributions as adjusted for ethnic composition, not because we think that's a great way of grading schools, but because it's the only data we have and we think it's more accurate than nothing.
It turns out that all but one of the schools in our town are well below average for the state, and probably for the country. What was interesting was to contrast the approach of the one school that has good test scores with one of the schools with poor test scores.
I visited the West Somerville School, one of the schools with poor test scores, first. The school seemed actually quite focused on test scores and how to improve them; the principal talked about various testing techniques, and the teacher who showed us around gave us examples of how they planned to implement the advice that the test taking consultants were giving them. Evidently this was all part of a strategy on the part of the school district to improve test scores.
The district's message did not seem to have penetrated at the one school with good test scores, however. The principal was clueless about why they had good test scores - it sounded like she thought the school was just lucky, and she certainly wasn't doing anything to keep the test scores up. With the teachers, testing didn't come up at all.
A bit of classroom observation, however, made it clear that there was a real difference between the schools: at the poor school, the students sat at desks listening to a teacher trying to teach them, while at the good school, the students were taking much more of the initiative, learning on their own, with help from the teachers only when they needed it. The teachers' job at the good school seemed to be more to facilitate - or perhaps merely to refrain from interfering with - the students' learning.
In addition, at the good school, the parents were very involved - to the point where they, instead of staff, were running the tours - and the teachers did a lot to communicate with and solicit feedback from the parents. Meanwhile, at the low scoring school, the school's attitude towards help from parents was basically "don't call us, we'll call you"; it was clear that they wanted to be in charge, rather than letting the parents be in charge as at the good school.
So there you have it. Test scores do seem to reflect a good learning environment and the associated better education. Ironically, however, teaching to the test does not appear actually to succeed at getting better test scores. Helping the students learn seems to be the only thing that improves test scores, and there aren't any real shortcuts.