Jared Polis is a Democratic Colorado Congressman. He recently had this interchange with Randi Weingarten on Twitter about Diane Ravitch.
Polis is a charter school supporter and has opened a few of his own charter schools, three in Colorado. Of those three schools, called the New America Schools, two are located in counties just outside of Denver and one is near Vail. Colorado is one of the states that has been most aggressive about tying standardized test scores to teacher evaluations and to school rankings. They have developed something called The Colorado Growth Model, which is a way of comparing how schools with similar achievement levels have progressed from one year to the next. So a school can have high test scores, but low growth and, conversely, there can be a school with low test scores but high growth. The Growth model, as the idea goes, is the great equalizer.
Now I don’t put much stake in these growth models. Like the New York City ‘progress’ score, a ‘growth’ number like 60 means that the students at that school generally scored better than 60% of the students in the state who had similar scores the previous year. Though they are not supposed to be biased, I think they are biased against low performing schools, and the graphs below support this belief. But people who fancy themselves ‘reformers’ like Polis do take these measures very seriously. So I thought I’d look at the excellent Colorado public data system called SchoolVIEW to see where The New America Schools stand.
The New America Schools, according to their website “empowers new immigrants, English language learners, and academically underserved students with the educational tools and support to maximize their potential and live the American dream.” Whereas most charters try to avoid English language learners, I appreciate that their mission is to serve such a population. I will permit them to use this as an ‘excuse’ for low proficiency rates.
But ‘reformers,’ now focus on ‘growth’ so I checked to see how good his schools do on that metric. From the data I was able to find on the two schools near Denver, they had some of the least ‘growth’ in their districts.
This first set of graphs shows a ‘bubble’ plot (like a scatterplot but the size of the dots depend on how many students are represented by each) is for all the schools in Jefferson County (where I used to teach, actually) for Math, Reading, and Writing. The schools with lower ‘growth’ will be to the left of schools with higher ‘growth’ on the graphs. The issue, once again, is not that The New America School has the lowest percent proficiency, but that they also have the lowest ‘growth’ score in the district on each of the tests.
The other New America School is in Adams County, and they suffer from the same ‘growth’ issue:
So for the data I was able to get on these schools, they, according to the ‘growth’ models are doing a very poor job getting their students to progress. One day these low growth numbers could cause these schools to get shut down. I wonder if Polis will still consider Ravitch ‘evil’ when he has to quote her arguments against these sorts of metrics to save his own schools.