<$BlogRSDUrl$>

My personal musings about anything that gets on my radar screen--heavily dominated by politics.

2004-12-14

A Useful Tool For Measuring Education

From this morning's WaPo:

Most days, whenever he has a free moment, Mark Twain Elementary School Principal Scott Ebbrecht can be found peering at the screen of the gray computer on his desk, trying to see exactly how well each student at his Westerville, Ohio, school is doing.

He is using something called value-added assessment, the hottest new tool in the national effort to improve public schools. At 485-student Mark Twain Elementary, a one-story modern brick building on East Walnut Street, the method has already brought results. Ebbrecht recently discovered that his top students, despite their high scores, were not improving as much as the value-added equations predicted, and he quickly made changes.

Value-added, which uses test scores to compare each child's progress to predictions based on past performance, has its critics. Some experts say that the tests are too narrow and that the analysis ignores differences in subject matter. But 16 state school chiefs have asked U.S. Education Secretary Roderick R. Paige to let them use the system or similar measures to meet federal requirements, and districts in at least 35 states have shown interest in using it.


If you ask me, this is the only way to make standardized testing a useful tool for measuring student performance and, by extension, school performance. In fact, Gov. Owens just signed into law a few months ago a bill that would change how Colorado schools get graded on test performance. They are no longer evaluated strictly on raw numbers but rather on the improvement within cohort groups from year to year (that is, same-student performances).

To illustrate the difference, the normal way tests are evaluated is to have the students take the test in, say, fourth grade. You then compare those tests to the performance of last years' fourth graders and make a conclusion about whether the school is getting better or not. See a problem here? In this model, you are comparing apples with green apples; that is, different sets of students at the same school. How do you compensate for strange performace factors in one group that weren't present in the other group? In other words, the scientific design of this model is very weak, and would be tossed out by most Ph.D. boards on the basis of poor baseline comparison value.

In the value-added equation, you evaluate those fourth graders' tests in comparison to LAST YEAR'S third grade tests--that is, with the same student group--and see if they are performing any better as fourth graders than they were as third graders. This allows you to draw some reasonable conclusions about the quality of instruction, and allows you to pinpoint key aspects of education because the controls on that experiment are much tighter. Tennessee has had a model in place for years that works like this, and it is often singled out as the most useful measure of student achievement in the country on a state-wide basis.

Of course, the usual suspects are going to fight this:

Value-added assessment has also become a political irritant because some school boards and superintendents want to pay teachers based on how much value they are adding, as measured by individual student test scores, for students in their classes. In Ohio and most other states, the system is being used only to diagnose student needs, leaving the question of teacher pay for later.

"We use it to improve instruction, not to evaluate teachers," Ebbrecht said. Among his teachers, however, its potential for affecting salaries "is a big fear," he said.


Once again, the teacher's union demonstrates how its first, second, and third priorities are NOT the students and the quality of education, but rather the protection of weak teachers and the assurance of perpetual existence. If this President really wanted to draw a line in the sand with a union, in much the same way that Reagan did with the Air Traffic Controllers, he should try with all his might to push this small reform through.

Value-added assessments do a lot more to pinpoint the accountability that is lacking in public education. Of course, that's why teachers oppose it, and why it must be a pretty smart idea.
Posted by: Michael / 12/14/2004 01:39:00 AM

Weblog Commenting by HaloScan.com

This page is powered by Blogger. Isn't yours?