Love the DC Schools Lotteries? Me Neither! You can Help!

[Anonymous]

The title of this thread refers to DC Charter School Lotteries.

Sure, the LOTTERY process could be improved upon. But we don't need to scrap the entire CHARTER system just because the LOTTERY needs a few tweaks.

[Anonymous]

All -

Just reaching out to give a quick thank you to those who supported the request for data. Wanted to let you know that OSSE responded and while they couldn't in that short time frame provide all of the data I asked for, they did share 12 datasets just in time to include them in the open data day project. Here's a bit more about it: http://greatergreatereducation.org/post/17870/open-data-can-help-families-with-confusing-school-choices/. I'm working on a post about the actual hack day that will hopefully be out soon. More importantly, the work on open data day got some momentum around this project, so hopefully, we'll have a rich dataset with details about schools that go beyond testing and more into what I consider to be practical things, like 'are there many families connected to that school from my neighborhood, so I can hope to set up carpools?'

In response to all of the Qs about my intentions, conspiracy theorists can breathe a sigh of relief . (sorry, couldn't resist). My personal goal is to make decisions around existing choices, and if I am able to access information, then should be able to, too. And why not? It's data about our kids and our families. Also, not that I wish the fighting over charter vs DCPS to continue, but if it's going to, let's have folks have informed debates based on actual data.

Thanks again to those who supported.

Sandra
(parent to two DCPS kids, one of which will likely be a charter kid, soon)

[Anonymous]

Anonymous wrote:

Anonymous wrote:The data you get would be completely useless. Family Educational Rights and Privacy Act protection would make each dataset completely anonymous and you would have no way to link data from one set to the other. It would be a tangle of generic IDs.

If they chose to, you could create a primary index that correlated IDs over time. It would still be anonymized. You could even introduce noise if you needed to for security.

I'd be interested if they did do that. The problem, as I see it, is this is going to be lots of data from unrelated sources so it would be hard to figure out exactly what records are related unless a single provider did the matching for you prior to compiling it.