Rebuttal of study

The Center for Education Reform rebuts the front page article in a recent NY Times that conventional public schools outperform private and charter schools.

Prominent researchers have argued that the variables used by the Illinois study were inadequate. Says Harvard Economist Caroline Hoxby, "The study by the Lubienskis uses Hierarchical Linear Modeling, which is nothing more than a way of comparing public schools to charter schools to private schools, controlling for some crude variables. This type of analysis has not been used by serious researchers for some time because it is grossly inadequate for making causal statements such as 'public schools do better' or 'Catholic schools do worse.'

------------------

REFUTABLE RESEARCH. Data and research on public charter schools keeps accumulating, but how and what to conclude remains the subject of much disagreement among a wide variety of researchers. Last week, a study of public, public charter and private school achievement in 4th and 8th grade was released. After allegedly using what the New York Times called "advanced sophisticated techniques" to level out demographic differences, Christopher Lubienski and Sarah Theule Lubienski of the University of Illinois at Champaign-Urbana concluded that conventional public school students outperform all others. Public charter students are overwhelming represented in urban areas while conventional public school students are more likely to be represented in suburbs. Prominent researchers have argued that the variables used by the Illinois study were inadequate. Says Harvard Economist Caroline Hoxby, "The study by the Lubienskis uses Hierarchical Linear Modeling, which is nothing more than a way of comparing public schools to charter schools to private schools, controlling for some crude variables. This type of analysis has not been used by serious researchers for some time because it is grossly inadequate for making causal statements such as 'public schools do better' or 'Catholic schools do worse.' Only methods that guarantee apples-to-apples comparisons, such as randomization, produce results that can be taken seriously."

A BETTER METHOD. What is the point of comparing apples and oranges? The Lubienski and Lubienski study does just that, forcing charter schools and private schools into the same category as public schools by using subjective variables. When using more accurate comparisons like matching or randomized studies, charter schools and private schools are shown to have far greater achievement than public schools. In a matching study, the achievement of a charter school or private school is matched with that of a nearby public school. An equally accurate method, randomization, takes a random sampling of charter schools and public schools for comparison. This kind of scientific education research, which has been done by Harvard researcher Caroline Hoxby, consistently produces the conclusion that charter schools and private schools out perform public schools.

REAL DATA. While the Illinois study was making the media rounds, new data was released that contradicts their conclusions. A new 2005 survey from the National Center for Education Statistics shows that charter schools open longer have higher math and reading scores than the national average. While the Lubienski and Lubienski study is just a snapshot of achievement and doesn't take progress into account, this survey reveals that students in the nation's oldest charter schools are a full 12 points ahead of their conventional school counterparts. Those results are also echoed in schools that are given more autonomy from the educational bureaucracy. The data also contradicts the study's assessment of demographics, revealing that, despite being eligible, 45 percent of non-participating charter schools choose not to participate in the federal Free & Reduced Lunch program, due mainly to administrative factors that make the program difficult to maneuver and obtain funds for. That data supports what dozens of previous studies have shown, that charter schools serve more disadvantaged students.