Conservatives are often accused of being “science deniers” when it comes to issues like climate change. But recent events reveal that those on the left suffer from significant confirmation bias when it comes to a stance that is increasingly central to the liberal education agenda — opposition to charter schools.
In December, the Network for Public Education put out a report entitled, “Still Asleep at the Wheel.” The study purports to examine the prevalence of waste in the Department of Education’s Charter School Program, which provides resources for new charter schools to get started, or for existing charter schools to expand and improve.
Recommended Stories
The NPE report claims that a significant number of charter schools that received grants under the program never opened, or had already closed. But even a cursory look at the data revealed to me that at least three of the schools in my home state described as “closed” are very much open — indeed, they are among the top-performing schools in the city of Milwaukee. If they got that wrong, what else might there be?
Further examination of the report’s findings led myself and the president of School Choice Wisconsin, Jim Bender, to write a post for the Fordham Institute that brought to light some additional errors. Indeed, at least 10 Wisconsin schools that the report claims are closed remain open. Christy Wolfe, vice president of the National Alliance for Public Charter Schools, identified similar errors in the identification of closed charter schools in California. Additionally, Nina Rees, president of NAPCS, pointed out that the total amount of money received by schools that never opened is likely to be far less than what is reported in the study because such schools are likely only to receive the planning portion of multi-year grants.
Identifying closed schools is difficult work — we even made one error in our corrective piece. Schools merge, change names, and sometimes change authorizers. That makes it difficult to tell exactly what has happened without a thorough analysis of the data on the ground in each state. What appears to have happened here is that the author made the assumption that when a National Center for Education Statistics school number disappears, it means that the school has gone away. But this is not always the case. The scenarios mentioned above can also lead to a change in a school’s NCES number.
To her credit, the author of the study has been willing to make changes where problems are identified. But these sort of basic verifications should have been conducted prior to the release of the study, not afterward. The “fix errors when they come to us” approach of NPE is simply not an appropriate way to conduct research if the goal is to influence public policy. The die is cast at the point of publication — most of the eyes that will view the report will do so in the immediate days following its release. Indeed, if NPE had published the report without errors or corrected them sooner, they wouldn’t have appeared in Valerie Strauss’s Washington Post article this month. NPE apparently knows this and is willing to concede on individual data points long after perceptions have been formed.
Despite the extensive errors, the study still has its defenders. The National Education Policy Center, housed at the University of Colorado, has continued to push the results on its Twitter feed even after these revelations. This is especially ironic when one considers that this organization receives significant funding to “fact check” research in support of school choice. Because they are housed within a university, they are sometimes treated as an objective source by the news media on voucher studies. But their continual pushing of such shoddy research should give reporters pause before they rely on this NEA-funded group for objective analysis.
Did the authors at NPE willfully manipulate the results? We don’t believe so. But it is likely that their strong, consistent opposition to charters led them to be far less careful about verifying results that were consistent with their preconceived notions. This sort of confirmation bias is problematic whenever it appears, but is especially insidious when it’s hidden behind the veneer of objective analysis from honest brokers.
Dr. William Flanders is the research director at the Wisconsin Institute for Law & Liberty.
