The Bruce Randolph Rorschach Test
Poor Bruce Randolph School. First, President Obama praises the school in his 2011 State of the Union address. Then Diane Ravitch, in a New York Times op-ed, cited the school as an example of “statistical legerdemain.” And now, Paul Tough, in a New York Times magazine piece, uses Randolph as an example of excuse-making and says students “deserve better.”
Each has a slice of the truth. This display from the Colorado Department of Education, included in Kevin Carey and Rob Manwaring’s excellent report on growth models, helps to illustrate the dilemma:
Looking at the y-axis, very few Bruce Randolph students scored as proficient on state tests. Yet, at the same time, 65 percent of students at Bruce Randolph had growth, or year-to-year gains in scores, above the median for similar students. Only one other high school in the city, the Denver School of Science and Technology, had better growth scores. And, in May 2010, 97% of Randolph students graduated.
Diane Ravitch is partially right: There are no miracles — this is the story of a school with students that are very, very far behind. But, so is Barack Obama. If you factor in where students start, Bruce Randolph is doing an exceptional job. And, while Tough is right that students deserve better, Bruce Randolph appears to be part of the solution, not the problem.
– Bill Tucker
[If you explore the data yourself on the Colorado Growth Model site, you’ll see that it’s even more complicated — Bruce Randolph’s high school outperforms the middle school. FYI, Bruce Randolph is located in Denver.]
UPDATED on July 11: I’ve updated my characterization of what Bruce Randolph’s growth percentile means in the chart above. I originally said, “65 percent of students at Bruce Randolph had growth above the median for similar students.” The correct interpretation is that there was a median student growth rate at the 65th percentile. In other words, half the students grew above the 65th percentile. The 50th percentile is Colorado’s median growth rate percentile. Thanks to Richard Wenning, former Associate Commissioner at the Colorado Department of Education and one of the persons behind the Colorado growth model, who called to offer the clarification.