Arkansas Schools Needing Improvement Listed -- Rural Policy Matters, Volume 4, Number 11 (November 2002)

The Arkansas Department of Education has listed 47 schools in 26 districts that it considers potential "School Improvement Schools" under the federal No Child Left Behind Act. The classification is based on the schools' students' performance on the Stanford Achievement Test in 1998-1999 and 1999-2000. They were listed if they receive federal Title I funds (based on poverty levels) and if 25% or fewer of their students scored "proficient" on the test. They also had to have failed to show "adequate yearly progress" on the test in both years. Students were tested in grades 5, 7, and 10.

The state had not intended to list any school until the end of this school year because it wanted to use results from its own Arkansas Benchmark Exam, which it has only started using recently. But no, said the federal government -- if you want federal money, you have to list schools now. So the state used the older SAT test scores.

Students in a "listed school" can transfer to another school in the district, or get tutoring from an approved tutor, both at the district's expense.

Data from the Department of Education's Web site is available for 45 of the 47 schools. It indicates that:

Those 45 had a total enrollment of 15,284, or about 3.5% of the state's total enrollment. They represent just fewer than 4% of the state's schools.
On average, two thirds of the students in these schools are African American and 75% are poor, compared to 25% and 50.5% statewide.
Fifteen of the 45 schools are located in Pulaski County, with Little Rock at its core. But the Delta region, which is largely rural, poor, and African American, also has 16 schools on the list, including five in Phillips County alone, among them the Lake View Elementary School in the district that filed Arkansas' school funding lawsuit.
Only four of the schools had a poverty rate less than the state average. Twenty-seven have poverty rates exceeding 75%.
Only seven of the 45 schools have African American enrollment below the state average of 23%; 35 of the 45 have African American enrollment at least double the state average.
Only eight of the 45 have more than 100 children per grade, the minimum number that testing research suggests are needed to make judgments about school performance. Twenty-six had half that number, and six had fewer than 20 students per grade.
Only 10 of the 45 are high schools, and only two of them are in the same county (Chicot).

Two comments: First, using student test scores to rate school performance based on a small number of students taking a test at a given grade level, especially measuring "progress" from year to year, is not a responsible use of test data. The pool of students taking the test can influence variation from year to year too heavily.

Second, while these schools may need improvement (a lot of others do, too), the remedy provided hardly seems likely to achieve results. For rural schools in particular, sending kids to another school is probably not practical. And using the school's limited resources to pay tutors seems likely to deprive it of the support needed to serve the remaining students. While schools, particularly rural schools, are scrambling to meet the demand for "highly qualified" teachers and paraprofessionals to teach Title I eligible students, no one is certain of the criteria for selecting tutors for the most needy children.

In this case, public policy needs improvement as much as the schools.