In a post about how Ohio is essentially targeting charter schools towards poor children, even though, technically, the criteria are ‘performance-based’, we come across this graph:
While the performance score is a somewhat odd way of calculating things*, the correlation is incredibly high, with an R2 of 0.83. As I’ve noted many times before, poorly performing schools are those with lots of poor students.
Might want to do something about that. Providing jobs would be a good start.
Quasi-off-topic: Have you tried running orthogonal regressions (where you minimize the perpendicular distance to the regression line rather than the vertical distance)? IIRC, it’s preferred if the Independent variable is something that varies naturally, rather than one that’s manipulated by the observer.
Pingback: The Economy is Depressing. – Bridget Magnus Shows the World as Seen from 4'11"