lovely analysis, thank you..  

the same nonsense of misunderstanding, misinterpreting and misleading the data goes on at the national level too.  

It's astonishing how the single theme that emerges from a careful look at the data, is that we don't have an education problem, we have a poverty problem. However it's much easier to punish teachers and schools with NCLB and other whips, than to think about poverty.

Yeah, I'm not discovering anything new to science here, but I was pretty astonished how precise the correlation is in this particular dataset. I expected poverty to be one factor among many. Certainly most annecdotal discussion of individual cases makes it seem like there's plenty of room for schools to make a difference, no matter what their situation. But the numbers really don't support this view.

Another friend contributes this link to more iterations of the same discovery:  

School Performance Articles (Walter R. Tschinkel)

This is really great. There's probably enough analysis here for a paper of your own. Did you send a link to Boston Magazine?

This brings to mind a discussion I recently had with a high school friend who spent a few years teaching middle school in a low income school in East Boston.  

(She also has three Harvard degrees, to her husband's two, which I mention more as a comment on their obvious commitment to Education with a capital E rather than as an indication of expertise.)  

Her belief is that, for the individual parent making a school decision, the most important thing in a system is the presence of a large enough cohort of children/families that are invested in the idea of education and college.  

This normalizes the idea of doing well in school for at least a subset of students (you do have to hope that your child identifies with that subset, I suppose, but thats your job, isn't it, parent?)  

Look for a few students every year that go to good colleges...it shows that there are teachers, students, and parents there that will support your child. As the individual parent, averages just aren't that important.  

Not a prescription for breaking the link between poverty and test scores, I admit, but if college is your bag, I'm betting Cambridge will do just fine...

Yeah, averages elide all the interesting differences between individuals, so none of this answers the real questions, at least parentally speaking, which are things like "Do we need to move? Or pay for private school? Or is any decent school with OK teachers going to do well enough with a well-prepared and well-supported kid?" You kinda want the data for every single kid. Actually, you kinda want the future data for your own kid, snuck back through a wormhole, because the prospect of guessing wrong is too scary.

A superb and extremely useful analysis. Have you extracted Charter School data to look at how well they are doing in addressing the achievement gap?

In a quick check, many of the Boston charter schools don't have all the same scores or data available, but the few that did are right in line with the predictions. When I have a chance I might try to do a complete state-wide dataset to see if there's more variation...

Nope, charter schools do no better.  

Throwing Charter Schools at Home Poverty

Maybe poverty isn't one factor among many. Thanks to Stephen Licht, maybe household resources tend to serve us in a proxy less primarily for risk, full stop, than for an embodied history of knowledge-work-to-mimic built into a family.
If we could find a suburb made entirely of pastors and teachers, I suppose this would be at its most simply falsifiable. I'm sure they will compare just fine to legaled and stockbroke spots on the map, at fractions of the rate of revenue. I'll see where this takes us before I chime in with the more general speculations that would follow if I keep writing.

Without trying to condense words to stay as short as the commenting convention:
Most of the people with their family names on companies possess a long history of investment by society in their family. Even the few that are self-made seldom came from untouchables.
The same is true of your membership in the middle class, and mine. We benefited indirectly from the attention that caregivers paid to our forebears. On top of that, we had knowledge work to mime and to model our identities on. These two dividends don't always lead to job opportunities (I think of people who go into philosophy, or who really care about why they do what they're doing for other reasons) and the people without high incomes are often the ones investing the gradual attention in future generations; big rewards go to less nurturing, more measurable tasks that score bigger right now instead: But it's fair to suggest that children at high risk for failing are children not of the philosophers, not of the poor as such, but of those for whom the stakes have not been lowered by our care. Families without a rich embodied history of caring-capital *ugh* - to put it in less nurturing terms - have less education but also less of other things to pass on in their time at home.