Thursday, May 05, 2011

Big Data, Visualizations of Texas Student Achievement

The Texas Tribune:

Gotta get this guy on the O'Reilly Radar.

Also, he's interested in high schools, which really shows the limits of current charters.

This is what happens when physicists look at the numbers instead of douchebag economists.


Unknown said...

It's been years since edBlogger 2003 in SF. I still stop by here pretty regularly to read what's on your mind. This was a delightful pointer! I've forwarded to all my "slowly getting more political" math and stats colleagues. Thanks.

Stuart Buck said...

Marder's analysis of charter schools is amateurish and simplistic, as much as if I were to make a video about physics.

He is simply comparing passing scores and the like. But as anyone familiar with public policy and social science would know, you can't thereby tell anything whatsoever about the performance of charter schools. In order to make such claims, you'd have to have information on a student-by-student basis about how they were doing before entering the charter school and how they did afterwards. Perhaps charter schools in Texas are serving kids who would have done even worse somewhere else.

Someone more familiar with charter schools would know that Texas law historically favored charter schools for "at risk" children. Right now, charter schools that serve predominantly students identified as “at risk” can be rated under an alternative accountability system. In 2007-08, 43.3% of charter schools in Texas qualified to be rated under that system, compared to a mere 3.3% of public school district campuses in Texas.

As the New York Times recently pointed out, "Recovery charter schools in Texas serve about 18,000 students who have performed poorly at traditional public schools, according to state data from the 2009-10 school year. Many have skipped too many classes or used too many drugs to graduate on time. Others have gotten pregnant or have emotional problems or learning disabilities. Recovery charters offer a second chance."

Any social scientist would know that you can't simply plot school-level passing rates by poverty and think that this is a fair comparison when so many charter schools are specifically designated for potential dropouts who are pregnant or on drugs, etc.

Any legitimate analysis would also take into account the fact that Texas charter schools receive thousands of dollars less per pupil than other public schools (see

Tom Hoffman said...


Marder's analysis is no more simplistic than that used to close schools, fire teachers and make other high stakes decisions about high schools here in Rhode Island and around the country.

Anyone familiar with public policy can tell you that, and a social scientist can tell you we don't have the longitudinal data you call for.

I'm certainly sympathetic to the view that alternative high schools can improve the lives of at-risk students without necessarily getting them ready to take college level math. Perhaps you'd like to encourage the charter community to change their rhetoric to promote that feature.

Tom Hoffman said...

Exciting additional comments lost in the Great Blogger Outage of '11.