writes with a link to his story in Wired, according to which "Test scores and student/teacher ratio are nearly meaningless. But three new numbers do describe school quality: 1. (Test Scores/Parent Education): How do scores outpace expectations? 2. Test Score Growth: Any single score can be socioeconomics, but growth is due to the school. 3. (Teacher Salary*%Highly Qualified/Teacher Age): The best teachers will become highly qualified early, and will gravitate toward the best paying jobs."
These factors seem to be at least interesting starting points; if you've shopped around for elementary schools, what else did you consider?