RVA Politics is a blog about politics in the state of Virginia And the capital city of Richmond.

The author is a political scientist. Please don't hold that against him.


RPS Needs Data, Transparancy

The state of Richmond Public Schools will play a key role in the city's elections this fall. The increasingly crowded field of mayoral candidates are already having to answer questions about school funding, with protesting crowds packing into City Council meetings to demand more for schools in next year's budget.  At least two sitting members of the School Board are running for Council in the fall as well.

The schools probably do need more money. But they might need good data even more. And there's another fight going on behind the scenes, with a legal battle between entrenched interests and an avowed crank, to force the schools to produce them.

John Butcher, a local activist and blogger who posts at the Cranky Taxpayer blog, has been complaining for years about mismanagement at RPS. He posts whatever data he can on his blog to try to analyze it for information about how the schools are performing.

Recently he stole an idea from a Loudon County activist: use a Freedom of Information Act request to demand SGP, or Student Growth Percentile scores. According to the Virginia Department of Ed, SGP scores try to measure a student's growth since the last time they were measured. Presumably, if a student has a bad SGP score this year, his or her teacher hasn't done very much to help them improve since last year.

Loudon's Brian Davison won a huge victory last month, when a Richmond Circuit Court compelled the county's schools district to release the SGP data, anonymized by teacher. Butcher has followed suit, but his request -- and Davison's -- is bringing new legal action from the teacher's union.

The Virginia Education Association has apparently filed a request for an injunction against Butcher, Davison, and the VA Dept. of Ed in an effort to block the SGP data. Their concern? According to the injunction petition, "Davison and Butcher have, and intend to, use VDOE's statistical data and reports to make prejudicial judgments about the evaluation, performance, success, or failure of individual teachers."

The union has a point. In an era where teachers have been unjustly blamed for failing schools -- and a whole lot more -- they're probably right to worry that teachers will take the fall for matters beyond their control. And test scores are surely an imperfect way to measure the impact of teachers, and creates all sorts of incentives to game the system. (Just ask Atlanta.)

Butcher is certainly guilty of placing too much emphasis on test scores. But he's right to ask whether the RPS is doing ANYTHING to evaluate its practices and teachers. At least Butcher is looking at whatever data is available and asking what's working and what's not. For instance, his data consistently shows that George Washington Carver Elementary outperforms its peers, by a lot. (See here for an example.) Why? What is Carver doing that's working? And why aren't other RPS schools stealing it?

My friend Craig Larson, a math professor at VCU, published an op-ed last year in the Richmond Times-Dispatch that called for better use of data in making decisions at RPS. Larson noted his frustration with RPS leadership:

I pointed out to a School Board member that Richmond Public Schools Superintendent Dana Bedden’s Academic Improvement Plan (AIP) did not appear to be based on any research. The superintendent’s office ultimately sent us both a list of bullet points and notes culled from various educational experts — but with no data or analysis.

There's a "big data" revolution going on in sports, finance, and just about everywhere else. Organizations are analyzing internal data to improve efficiency and best practices. Transparency and data sharing allows the public to crowdsource solutions for problems ranging from potholes to healthcare to art.

But RPS doesn't seem to be paying attention. If the schools and the teachers don't want to share the data they DO have, then they at least need to show that they're using data internally to make decisions, but there's no evidence of that.

Will next year's City Council, including newly empowered former School Board Members, demand data-driven practices? Will admired Superintendent Dana Bedden make it happen? Will the union stand in the way? The answers to these questions may be more important than any budget numbers in deciding the future of Richmond's schools.

Terry Mac's Good Year

VA GOP's felon rights move won't matter for November