Posts Tagged ‘High Stakes Testing’

Virtual Twins

Tuesday, June 16th, 2009

At the risk of blogging this Edweek article by Leslie A. Maxwell into the ground…

national study released todayRequires Adobe Acrobat Reader casts doubt on whether the academic performance of students in charter schools is any better than that of their peers in regular public schools.

Looking at 2,403 charter schools in 15 states and the District of Columbia, researchers at Stanford University found that students in more than 80 percent of charter schools either performed the same as—or worse than—students in traditional public schools on mathematics tests.

Specifically, researchers at the Center for Research on Education Outcomes at Stanford found that:

• Thirty-seven percent of the taxpayer-funded but largely independent schools posted gains that were “significantly below” what their students would have realized if they had enrolled in their local traditional public schools instead.

• Forty-six percent of charters produced learning gains that were indistinguishable from their local public schools’.

• Seventeen percent of charters posted growth that exceeded that of their regular public school equivalents by a “significant amount.”

“If this study shows anything, it shows that we’ve got a two-to-one margin of bad charters to good charters,” said Margaret E. Raymond, the director of the center and the study’s lead author. “That’s a red flag.”

To produce the study, “Multiple Choice: Charter School Performance in 16 States,” researchers used student-level longitudinal data from each of the participating states and the District of Columbia. They created a “virtual twin” from local public schools that matched each charter school student’s profile according to race and ethnicity, eligibility for the federal subsidized-meals program, participation in special education programs, English-language proficiency, and starting test scores.

So if student A performs better than his virtual twin A-2 at two different public schools, and student B performs better than his virtual twin B-2 at two different charter schools does it say that:

the tougher questions that we still need to answer relate to why some teachers and some schools in both sectors do better than others in the same sector

as Jeffrey R. Henig, an education professor at Teachers College, Columbia University claims? Or does it mean that you can’t compare kids (or teachers or schools) accurately with standardized tests?


Tuesday, June 9th, 2009

I found this link in Diane Ravitch’s blog.  Thanks to the Daily News for its critical examination of New York’s inflated testing results:

It’s the state exam version of grade inflation.

Soaring scores on the state math test don’t necessarily add up to better schools or smarter kids.

That’s because it has gotten easier to teach to the test as the questions have gotten easier to predict, a Daily News analysis revealed.

And, the tests may also be easier.

“It’s the lesson of the financial crisis, and it’s the lesson here – you can’t just trust the numbers, you have to look at what the numbers mean,” said Columbia University sociology doctorate student Jennifer Jennings.

“If you can always make pretty good guesses about what’s going to be on the state tests, teachers aren’t stupid and we’re putting them under a whole lot of pressure, so basically they’re strategic about what they teach.”