Letter to the Editor: Report on Cowen Institute points out fatal flaw

Thomas Agnew

Last issue’s article on the resignation of John Ayers, director of this school’s pro-charter Cowen Institute, left out an important detail. The fact that the “erroneous methodology” employed in the Institute’s report on students’ progress was so basic that any introductory statistics student could have spotted it.

The study combined achievement metrics from New Orleans high schools and schools from throughout Louisiana, and then used the data to create a linear regression model, which plots a line through a cloud of data points.

Any New Orleans high schools above the line were then declared to have “beaten the odds.” Unfortunately, the “odds” beaten by the schools did not derive from poverty or other social factors, but from the model itself. One “beats the odds” in the same way by calling a coin flip six out of ten times.

The error was so fundamental that it calls into question the intellectual integrity of the Institute itself. The Hullabaloo’s article quotes Will Griffin as speculating that many pro-charter groups that praised the study lost credibility as a result.

Rightly so: pro-charter think tanks such as the Cowen Institute have long made their bones producing dubious studies meant to prove that pro-corporate education reform helps impoverished students, when the exact opposite is the case. By promoting this study, the Institute’s hands were caught in the proverbial cookie jar.

Sincerely 

Thomas Agnew