"A child's learning is the function more of the characteristics of his classmates than those of the teacher." James Coleman, 1972

Wednesday, August 29, 2007

NCLB, the Sequel: OCAD (Obsessive Compulsive Accountability Disorder)

If any department of the federal government operated under the level of scrutiny and distrust envisioned by the Miller-McKeon draft version of NCLB, another whole department would be required just to keep track of the paper. Miller and McKeon have that one figured out, though--they just plan to pass their impossible continuation of a pie-in-the-sky policy to the states to figure out.

There is lot here to read for those expecting something different. I found this on page 47 and stopped reading:

[MILLER-MCKEON DISCUSSION DRAFT]
‘‘(I) TIMELINE.—Each State shall establish a timeline for adequate yearly progress. The timeline shall ensure that not later than the end of the 2013–2014 school year consistent with subparagraph (F), all students in each group described in subparagraph (C)(v) will meet or exceed the State’s proficient level of academic achievement on the State assessments under paragraph (3).
In short, the impossible targets and the accompanying assured failure remain unchanged. Email Congressman Miller to say SCRAP IT AND START OVER!!!!!!!

Ed Week's "Learning the Language" has a spot to react to the draft, and thanks to Monty Neill's note at ARN, I found this reaction by Jim Crawford to the ELL revisions in the Miller-McKeon draft. Nice analysis.

I'm still slogging through the 435-page draft of Title I -- a necessity, I'm afraid, because the summary isn't altogether accurate. There are provisions affecting ELLs throughout, but see especially pp. 67-87. Also note that a draft of Title III, which deals specifically with ELLs, has yet to be released.

To quote Samuel Johnson out of context, I'd say that this reauthorization proposal reflects the triumph of hope over experience. Over the past 5 years, experience has revealed fundamental flaws in No Child Left Behind. But instead of fundamentally rethinking the law, Reps. Miller and McKeon propose to add new layers of complexity, apparently hoping to vindicate their rash creation.

Especially where English language learners are concerned, I think the draft bill is unlikely to improve NCLB's garbage-in, garbage-out approach to accountability, which uses invalid, unreliable assessments for high-stakes purposes.

Here are just a few reasons:

1. The bill mandates valid and reliable assessments for ELLs, but so does current law. We know how that's worked out: the vast majority of content assessments now being administered to these kids are English-language tests that nobody even bothers to claim are valid or reliable. All that would change under the Miller-McKeon proposal is that states would face a two-year deadline to develop valid/reliable tests or face financial penalties.

The assumption seems to be that states have simply dragged their feet in meeting their obligations. But would it even be possible to develop English-language academic assessments that are valid for ELLs at many different levels of English proficiency? That's never been achieved, and I suspect that trying to do so is a fool's errand.

2. How about accommodations for ELLs taking tests designed for proficient English speakers? The draft bill would require states to use only those with "research-based" validity and reliability. None currently exist, and research in this area remains quite limited. Accommodations do not offer an easy solution, at least in the near term.

3. The bill would allow the use of portfolios and other alternate assessments -- potentially an improvement. But again, these types of tests are rarely available and would take time to develop. Meanwhile, the "inclusion" of ELLs, using inaccurate assessments for measuring "adequate yearly progress," would continue until most schools with an ELL subgroup would find themselves in "corrective action" status or worse.

4. The bill would allow states to use English language proficiency scores for beginning ELLs for up to 2 years, while valid/reliable assessments are supposedly being developed. It would allow them to exempt ELLs from language-arts assessments for their first 12 months in U.S. schools. And it would allow them to count the scores of former ELLs for AYP purposes for up to 3 years.

Yet all of these provisions are arbitrary, unscientific, and unlikely to do much to mitigate the unfairness of the current system. Nor do they address the well publicized abuses that NCLB has created: emphasizing test prep to the exclusion of real teaching, stressing basic skills over critical thinking, and narrowing the curriculum to the 2 high-stakes subjects.

5. As a way out of this mess, Miller and McKeon rely heavily on expanding the use of native-language assessments. But they don't seem to recognize that only about 15 percent of ELLs now receive native-language instruction in core subjects. For those who lack literacy in, say, Spanish, native-language tests are inappropriate.

6. Yet the draft bill requires states to develop native-language assessments for any language group that makes up at least 10 percent of the state's ELLs. This would be a huge change. It would probably have some positive effects (e.g., for kids now in bilingual classrooms), along with some very perverse consequences.

First, the good news. Ensuring that native-language assessments are universally available would relieve some of the pressure to push ELLs into English as rapidly as possible (which contradicts research evidence on best practices). It also might slow down the trend of dismantling bilingual programs because of anxiety about ELLs making AYP on English-language tests. Which would be good thing, in my view.

On the other hand, the mandate could impose a crushing burden on many states. Virtually all of them would have to develop native language assessments in Spanish (or, e.g., in Texas, redesign these tests to make them valid/reliable). Some would probably have to do so in Tagalog, Hmong, Vietnamese, Chinese, Korean, Yup'ik, and possibly other Native American languages.

Developing such assessments would be quite expensive and time-consuming, especially in states with new or diverse immigrant populations. It would be a foolish diversion of resources away from what they most need: professional development and technical assistance to develop schools' capacity in serving ELLs. This is where the feds could play a major positive role -- if only Congress could get over its assessment obsession.

7. Finally, there's a loophole in this provision requiring that native-language assessments for ELLs be "consistent with state law." The perverse effect here might be that many states would choose to outlaw such assessments, denying them to ELLs for whom they would be meaningful and appropriate.

Looking at this draft -- a Rube Goldberg contraption if there ever was one -- I have to wonder whether Reps. Miller and McKeon heeded any expert advice on ELL needs whatsoever. Ignoring the views of most educators and researchers, while listening primarily to Washington "think tanks," proved disastrous in the first authorization of NCLB. Let's hope Congress can learn from its mistakes.

1 comment:

  1. you know we win here...i mean...the llonger this stays in place, the closer we'll come to 1 million signatures.

    sad huh?

    ReplyDelete