Wednesday, February 01, 2006

EDITORIAL >> School report badly skewed

Lots of teachers and parents in more than a few of our school districts must have been dismayed to see that some of the poorest instruction in Arkansas was occurring in their schools. Only 17 of Arkansas’ 1,116 public schools, for example, were judged to be worse than the Jacksonville Middle School in a performance index crafted by the research team at the new University of Arkansas Department of Education Reform.

Teachers, don’t surrender your certificates. Parents, don’t yank your kids from the schools just yet. You can invest about as much confidence in this academic study as you could in President Bush’s promises of jobs from tax cuts.

We have no idea how any school in the area measures up against the rest of the state, much less the nation, but the methodology and results of this celebrated study are so counter-intuitive — goofy, actually — that no one can take it seriously.

If the “School Performance Index in Arkansas” is accurate, the most abysmally performing schools in Arkansas by traditional standards actually are the best schools, starting with little Altheimer in Jefferson County. So awful were the financial conditions and test scores in Altheimer that the state Department of Education took over the district and ran it until recently. Yet no school district in Arkansas comes close to matching the effective teaching force at Altheimer in the performance index.

Its rankings suggest that teachers who are paid less and have weaker traditional qualifications than teachers generally do, provide the most effective teaching. A goal of the study was to prove that the amount of money spent on education has absolutely nothing to do with its quality, which has long been the thesis of the Dr. Jay P. Greene, the chair of the school-reform department and the director of the research team. He declared that the study proved it, although many of his results supported the opposite. Pulaski County’s magnet and incentive schools, which get extra infusions of money, ranked much higher than other schools in the area.

The premise of the study is undeniable. Standardized test scores alone cannot measure the relative effectiveness of a teacher, although that soon may be how teachers are paid. Gov. Huckabee plans to ask the legislature at a special session to enact a law requiring each teacher to be paid on the basis of standardized test scores of her students. An eighth-grade math teacher who inherits a class in which 60 percent cannot read at a third-grade level or do simple subtractions is not going to make as much progress as a teacher of lesser ability with a class of affluent and motivated kids.

The case against the high-stakes testing required by the No Child Left Behind law is that they do not account for the disabilities that children bring to class. So Greene’s team set out to do that. They factored in the percentage of black and Hispanic kids and those on subsidized lunches because family poverty is a huge factor in student progress. Also, they attempted to factor in the education and income levels of parents and the percentage of single parents, which numerous studies have shown to be big factors in student achievement. They assigned an arbitrary value to each of those factors and arrived at an expected performance level for students in each school. If scores fell below the expected level for that school, the teaching was judged to be failing, and if they were above, it was effective.

But the researchers really could not get all those figures, so they seem to have made up some based on communitywide census data. For example, in Pulaski County, where a high percentage of affluent and middle-class white families send their children to private and religious schools, all the public schools were assigned those characteristics — a high quotient of white, relatively affluent, well-educated and stable parents. Public school enrollment often is quite different. Greene’s men, new to the state, might not have known that. Accurate demographics for schools are not available for much of the state. Arkansas school districts do not follow city, county or even census-tract boundaries.

Conservative foundations, including the Wal-Mart family foundation, settled $20 million on the university to start the school-reform department at Fayetteville in hopes of shaping education policy in Arkansas. Corporate influence on scientific research creates great peril for universities. This is the first product of the big foundation grants, and it is not heartening.