Reading by the numbers
Kenneth S. Goodman, Author, The Truth About DIBELS, 2006, Heinemann, comment: [This] story on the reading program at JFK Elementary in Brockton left out
some important information. The program is part of Reading FIrst, the core
component of NCLB. DIBELS is a series of 1 minute tests on such things as
the ability to "read" two or three letter nonsense syllables in three
seconds At JFK it is administered weekly and in between the kids are
drilled for the test. No wonder they improve on the test. But there is no
test of comprehension. One subtest involves reading a short text but it is
scored by counting the number of words read correctly in one minute.
On May 1, 2008 the US Department of Education released a report that showed
that after six years and $6 billion dollars in funding this Reading First
program showed no benefit in improving comprehension. Rep. George Miller,
chair of the House education Committee said: "Because of the corruption in
the Reading First program, districts and schools were steered towards
certain reading programs and products that may not have provided the most
effective instruction for students. That may explain why we are seeing these
DIBELS is at the center of a major conflict of interest scandal. According
to the DOE Inspector General, a number of states including Massachusetts
were forced to include it in their Reading First proposal by authors of the
program who were on Reading First review panels.. Roland Good, the chief
author made more than a million dollars on the test, developed with fedeeral
funds, while serving as a government consultant. The Justice Department is
considering criminal charges. Meanwhile five-year-olds in kindergarten are
flunking DIBELS the first week of school.
WITH ITS walls covered in progress-monitoring graphs and charts showing the distribution of test scores, the principal's office at the John F. Kennedy Elementary School in Brockton has the air of a command center. At the school, where 58 percent of students are from low-income households, hard data determine what each student's day looks like. And a tight focus on reading early on is intended to help students avoid academic trouble in the future.
more stories like this
Principal Virginia Lynch says she wasn't enthusiastic two years ago when the district selected the 550-student Kennedy School to be a pilot project for a data-driven instruction system to improve literacy. She was even offended when she learned that a "principal coach" would be arriving, along with specialists to train her staff in testing techniques and data collection. But as her students have become better readers, Lynch has become a devotee of methods that test students in key skills such as phonics, vocabulary, and comprehension.
"If you want to change achievement, then you need to change instruction," says Edward Moscovitch, chairman of the Bay State Reading Institute, a nonprofit group with a $1.2 million annual state contract to bring data-driven instruction to 18 Massachusetts schools. Moscovitch, one of the authors of the Education Reform Act of 1993, argues that the keys to proficiency include testing students often and teaching them in small groups tailored to their skills.
The Kennedy sets aside extra reading time each day for struggling students who take weekly one-minute tests. Teachers trained by two of the school's permanent reading specialists intervene quickly, allowing students to advance to groups with higher levels of reading fluency and comprehension.
The results are impressive. According to a widely used assessment called Dynamic Indicators of Basic Early Literacy Skills, Kennedy students improve significantly in reading over the school year. But the benefits are broader. One dramatic effect, says Lynch, has been a reduction in the number of children recommended for special education services. This year, staffers found it necessary to screen just 10 students for special ed, as opposed to about 75 before the introduction of the new literacy program.
Justice in Brockton
Both local history and poetic justice would be served should the Kennedy School become a model for the next stage of reform in Massachusetts. Jami McDuffy is a first-grade teacher at the Kennedy, where her mother, Janet, is a curriculum specialist working on a data-driven instruction plan for math.
But back in 1993, Jami was a 13-year-old Brockton student and the lead plaintiff in the landmark McDuffy v. Robertson case that highlighted gross disparities in educational opportunities in Massachusetts. The state Supreme Judicial Court ruled in her favor, finding that Massachusetts had failed to meet its constitutional duty to provide an adequate education for all children. Billions of dollars in state aid to equalize spending would follow from the Education Reform Act of 1993.
Now the question is focused more on how to make the most of available funds - and especially on how to close the achievement gap between white and minority students. Lynch says that useful comparative data by race for Kennedy students should be available next year. But schools in some states, including New Mexico and Alabama, are already reporting that similar data-driven literacy-instruction methods implemented under the national Reading First effort are helping to close the achievement gap.
Critics of these methods charge that they overemphasize the ability to read words aloud fluently and do not fully measure children's comprehension or lack thereof. At the Kennedy, though, students at least seemed to understand the message in what they were reading.
Part of state policy?
It would be a major disappointment if the Patrick administration fails to include data-driven instruction in its 10-year "readiness" plan for education due out later this month. Right now, the administration appears focused largely on longer school days and new governance models as ways to improve underperforming schools. These are valuable reforms. But data-driven instruction in some cases could prove as effective and less expensive. And if the early Kennedy School results can be replicated, it holds enormous potential for moving special ed students into mainstream classrooms.
The general acceptance of data-driven instruction programs by teachers is also significant. A clash could be brewing between teachers unions and Governor Patrick over the administration's proposal for so-called in-district Readiness Schools, where the influence of unions on curriculum and staffing would be limited. Moscovitch says, however, that teachers unions have not opposed any of the 18 schools now using his data-driven literacy program.
The Kennedy School still has much to prove in the next few years, including how its literacy program matches up with performance on the state's standardized MCAS test and how well it serves low-income and minority children. But for now, the line graph appears to be rising.
Editorial, with rebuttal by Ken Goodman
INDEX OF NCLB OUTRAGES