Gary Rubinstein has been watching the Achievement School District in Tennessee since it started with lofty promises. Its leader Chris Barbic gathered the lowest scoring schools in the state and pledged that he would raise them from the bottom 5% to the top 25% within five years.


As Gary earlier documented, the first cohort of six schools were still very low performing after four years. Four were still in the bottom 5%, and two had advanced to the bottom 6%. Barbic resigned.


His replacement, Malika Anderson, arrived and began by boasting about the success of the ASD.


“In her letter introducing herself, Anderson shows that she has been briefed about how to spin their data. She writes:


“After only three years, we’re excited to say that half of all Priority Schools in the state are now receiving some form of significant intervention. The bar for the bottom 5% has increased nearly 10 points—over a 50% increase—and students in Priority Schools are growing 4 times faster than their statewide peers.”


Gary demonstrates the value of mathematics in analyzing and debunking spin:


“Since they cannot deny that their schools are still in the bottom 5%, they say that ‘the bar for the bottom 5% has increased by nearly 10 points — over a 50% increase.’ So what this means is that there is some metric, I think it is called the School Success Rate or SSR. It is calculated every year for every school, but only released every four years or so. So the bottom 5% school evidently used to score only 20 ‘points’ on this metric and now they’ve gone up by 10 ‘points’ to 30, which is a 50% increase. But apparently the other schools that were not in the bottom 5% have also increased by 10 points so that the bottom 5% schools have not overtaken anyone as they were supposed to. Also there’s this stat that students in priority schools are ‘growing 4 times faster than their statewide peers.’ Basically this means that there is some metric on which the statewide peers got a very low growth number, something very close to zero. And when you multiply something very close to zero by four you still get something very close to zero. It’s like if I go on a diet and lose one pound, I lost four times as much as a person who lost just a quarter of a pound. It is meaningless to talk about comparing growth rates this way when both numbers are so low.”