October 1, 2006

The Washington Post on Reading First

We're finally starting to get some MSM analysis on the Reading First scandal. As usual, it's superficial and gets it mostly wrong. Let's look at the Washington Post's article By Michael Grunwald:
President Bush's No Child Left Behind Act was premised on three revolutionary goals. The first was to focus on low-performing schools and students; hence, No Child Left Behind. The second was to beef up the federal role in education, enforcing national standards through testing. The third was to bring facts and evidence to the notoriously squishy world of education policy, promoting teaching methods backed by "scientifically based research" instead of instinct and fad. This was the least-publicized goal, but arguably the most vital; the phrase "scientifically based research" appeared more than 100 times in the landmark 2001 law.
SBRR is indeed one key term defined in section 1208(6) of the NCLB act. But, there's another even more critical term we need to understand that gives meaning to SBRR. That term is the essential components of reading instruction which is defined in 1208 (5) as:
explicit and systematic instruction in—
‘‘(A) phonemic awareness;
‘‘(B) phonics;
‘‘(C) vocabulary development;
‘‘(D) reading fluency, including oral reading skills; and
‘‘(E) reading comprehension strategies."
These are the five reading components that the National Reading Panel (NRP) determined were essential to reading instruction. They are now enshrined in the Reading First statute along with the critical requirement that instruction in these essential components be "explicit and systematic."

The NRP performed a meta-analysis on all the extant SBRR and distilled these essential components. Much of the SBRR involved two instructional programs, Direct Instruction (DI) and Success for All (SfA). So, when we look for guidance as to what explicit and systematic instruction in these five essential compnents is going to actually look like we are going to be looking at SBRR validated programs like DI and SfA since those are the programs that the NRP distilled their conclusions from. Keep this in mind as we continue analyzing Grunwald's article.
The centerpiece of the new research-based approach was Reading First, a $1 billion-a-year effort to help low-income schools adopt strategies "that have been proven to prevent or remediate reading failure" through rigorous peer-reviewed studies." Quite simply, Reading First focuses on what works, and will support proven methods of early reading instruction," the Education Department promised.

Five years later, an accumulating mound of evidence from reports, interviews and program documents suggests that Reading First has had little to do with science or rigor. Instead, the billions have gone to what is effectively a pilot project for untested programs with friends in high places.
This would be true if Reading actually insisted that the reading programs be validated by SBRR, but Reading First has no such requirement. Reading First permits programs that are "based on" or "are consistent with" SBRR-- a much less onerous requirement. All this requires is that reading programs look like the reading programs having actual SBRR, i.e., DI and SfA, with respect to the five essential components of reading and the "explicit and systematic instruction" thereof.

So, the entire premise that Grunvald's article is built upon is completely wrong. Now watch the masterful job Grunwald does piling on innuendo and half-truths on his faulty premise to complete his hatchet job.
Department officials and a small group of influential contractors have strong-armed states and local districts into adopting a small group of unproved textbooks and reading programs with almost no peer-reviewed research behind them. The commercial interests behind those textbooks and programs have paid royalties and consulting fees to the key Reading First contractors, who also served as consultants for states seeking grants and chaired the panels approving the grants. Both the architect of Reading First and former education secretary Roderick R. Paige have gone to work for the owner of one of those programs, who is also a top Bush fundraiser.
That "small group of unproved textbooks and reading programs with almost no peer-reviewed research behind them" are the reading programs that the review panelists determined were sufficiently "based upon" or "consistent with" the programs having actual SBRR as permitted by law. Sure, it's possible that these programs did not meet this requirement but there is nothing in the OiG report or Grunwald's article to suggest that there was.

All Grunsvald provides is the same innuendo that the OiG report provided with respect to the "apparent conflicts" of the panelists. But, as I've written in previous posts, almost every panelist had a connection to one or more reading programs. That's why they were qualified to be panelists--they were experts in reading research and there is a high correlation between reading researchers and creators of reading programs. The DoE, which was under no obligation to check for any conflicts, looked for financial conflicts and found none. Neither OiG nor Grunsvald have provided any hard evidence that such financial ties exist. The "significant professional contact" standard is a non-standard since it would have ensnared almost every panelist and would have effectively precluded every expert reading researcher from serving on the panel. Not exactly a desirable result.

Then Grunsvald recites the infamous Doherty emails and then plays up both of his faulty premises.
But the report barely scratched the surface of the incestuous process that dominated the formation of Reading First. The initiative didn't promote scientifically based reading instruction, the third goal of No Child Left Behind. And it's providing ammunition to critics of the second goal, strong national standards. The billion-dollar question is whether it may imperil the first goal: Will some children get left behind?
Oh boy.
Bush administration officials frequently say that Reading First does not play favorites or intrude on local control, that states and districts are free to choose their own textbooks and programs -- as long as they're backed by sound science. But aggressive muckraking by the newsletter Title 1 Monitor and reading advocates at the Success for All Foundation have eviscerated those claims, and the inspector general's report officially contradicted them, accusing the department of breaking the law by promoting its pet programs and squelching others.
My definition of eviscerate does not include arguments based on unproven claims and a wrong reading of the reading First statute. To reiterate, these would be good arguments if the Reading First statute were limited to programs validated by SBRR and if the panelists had real conflicts. The OIG report fails to show that any panelist actually was on a panel in which a program he had ties to was actually being considered. And, the OIG report failed to actually show that "pet programs" were benefited or that any programs except ones lacking SBRR were excluded. Slavin, the creator of SfA, himself has actually stated that both SfA and DI combined have received less than 3% of Reading First funding. Not exactly a boon for DI.
In his internal e-mails, Doherty frequently admitted using "extralegal" tactics to force states and local districts to do the department's bidding. A report by Success for All documented how state applications for Reading First grants that promoted the preferred programs were the only ones approved.
As long as these programs were consistent with or based upon SBRR, there is no violation of the law. In fact, it was the intention of the law to exclude any and all programs which were not based on SBRR. SfA in fact was approved for use by about 28 states. But, to be approved for use, states had to actually include SfA in their applications. Many states failed to include either of the programs with validated SBRR, SfA or DI, in their applications in favor of programs without SBRR (i.e., whole language programs). DOE could rightly exclude the non-SBRR programs, but it could not approve, endorse, or force any state to include any program, including DI or SfA, in its application. If there is any scandal here at all, this would be it. Why did so many states exclude either or both of the only validated SBRR programs off its Reading First funding list?

In fact, the vast majority of the 4,800 Reading First schools have now adopted one of the five or six top-selling commercial textbooks, even though none of them has been evaluated in a peer-reviewed study against a control group. Most of the schools also use the same assessment program, the same instructional model, and one of three training programs developed by Reading First insiders -- with little research backing.

"They kept denying it, but everybody knew the department had a list," said Jady Johnson, director of the Reading Recovery Council of North America. "They're forcing schools to spend millions on ineffective programs."

There's nothing wrong with DOE keeping "a list" of reading programs which they deemed were based on SBRR. DOE was not permitted, however, to tell a state what reading programs met this criteria, so, apparently, they referred states to applications of other states that were approved. Frequently, these lists excluded SfA. And, don't think any state was gung ho to go out of their way to include SfA.
To some extent, the controversy over Reading First reflects an older controversy over reading, pitting "phonics" advocates such as Doherty against "whole language" practitioners such as Johnson.
The only reason why the "controversy" persists is because the "whole language" cultists refuse to give up the ghost despite being the losers in the SBRR race. There exists no valid scientifically based reading research which supports the use of whole language to teach children how to read in K-3.

The administration believes in phonics, which emphasizes repetitive drills that teach children to sound out words.

The administration believes in phonics because the SBRR unquestionably shows that phonics is more effective than whole language. And, a program that "emphasizes repetitive drills that teach children to sound out words" is what the Reading First statute calls "explicit and systematic instruction" in the five essential components of reading.

Johnson and other phonics skeptics try to teach the meaning and context of words as well.
Being a phonics skeptic in 2006 is equivalent to being a evolution skeptic. Teaching "meaning and context of words as well" means not teaching phonics in a way that is systematic and explicit according to the SBRR. This is why these non-SBRR programs were excluded from reading First.
Reading First money has been steered toward states and local districts that go the phonics route, largely because the Reading First panels that oversaw state applications were stacked with department officials and other phonics fans.
Wrong. States had to go the phonics route because they had to pick programs that were based on SBRR. if they didn't pick phonics programs, they would not have been in compliance with the Reading First statute. The panels were stacked with phonics experts because the statute required that the panels be comprised of reading research experts. Having non-phonics experts on the panel would be like having witch doctors, faith healers, and shamans on a panel of doctors.
"Stack the panel?" Doherty joked in one e-mail. "I have never *heard* of such a thing . . . ." When Reid Lyon, who designed Reading First, complained that a whole-language proponent had received an invitation to participate on an evaluation panel, a top department official replied: "We can't un-invite her. Just make sure she is on a panel with one of our barracuda types."
That's a good example of an instance in which a witchdoctor scammed her way on the panel. I'd argue that not kicking her off was itself a scandal.
Doherty bragged to Lyon about pressuring Maine, Mississippi and New Jersey to reverse decisions to allow whole-language programs in their schools: "This is for your FYI, as I think this program-bashing is best done off or under the major radar screens." Massachusetts and North Dakota were also told to drop whole-language programs such as Rigby Literacy, and districts that didn't do so lost funding. "Ha, ha--Rigby as a CORE program?" Doherty wrote in one internal e-mail. "When pigs fly!"
Under the statute, DOE was required to force these states to exclude these whole language programs for failing to be based on or consistent with SBRR. Good job.
Said Bruce Hunter, a lobbyist for the American Association of School Administrators: "It's been obvious all along that the administration knew exactly what it wanted."
And what it wanted was for states to include in its Reading First applications only reading programs having SBRR as they were required to by law. The question remains as to why so many states were including in their applications reading programs that did not have SBRR.

But it wasn't just about phonics.

Success for All is the phonics program with the strongest record of scientifically proved results, backed by 31 studies rated "conclusive" by the American Institutes for Research. And it has been shut out of Reading First. The nonprofit Success for All Foundation has shed 60 percent of its staff since Reading First began; the program had been growing rapidly, but now 300 schools have dropped it. Betsy Ammons, a principal in North Carolina, watched Success for All improve reading scores at her school, but state officials made her switch to traditional textbooks to qualify for the new grants.

Apparently, shut out means included for funding by 28 states.

And, SfA does not have the "strongest record of scientifically proved results" it has the second strongest. We measure effectiveness by a combination of number of studies and effect size. DI is twice as effective as SfA in teaching children at-risk for reading failure.

And, remember, DoE could not force any state to include any particular program in it's Reading First application. It could only make them exclude non-SBRR programs. This it did.
"You can't afford to turn down the federal money," Ammons said. "But why should we have to give up on something that works?"
Because your state failed to include SfA in it's Reading First application, jackass.
The answer lies in the Reading First grant process, which was almost comically skewed.
Not nearly as comically skewed as this article.
Michigan was the first state approved, after it simply proposed to adopt the five best-selling textbooks. But when Rhode Island officials proposed to require "high-quality reading programs that meet the test of having a scientific research base," they were rejected.
Har de har har. Every whole language reading program, including reading Recovery, claims to be scientifically based. All you need is an ideologically skewed State DOE that was willing to allow such non-research to pass for research, which many would have, and it would have made a travesty out of Reading First. DOE was doing its job by not permitting states to play this game and subvert the Reading First statute.
Doherty told them to check out Michigan's list, so they cut and pasted it into their application, while suggesting that districts could still adopt other programs justified by research.
This is exactly what he was supposed to be doing -- not permitting programs lacking SBRR while not endorsing any program in particular. It was up to each state to decide what programs to include in its Reading First application.

They were rejected again. So they limited their program to the textbooks. Only then were they approved.
Perhaps because they once again tried to sneak in some non-SBRR programs.
Similarly, Oklahoma unsuccessfully proposed to require reading programs backed by three years of longitudinal data before it got the hint and proposed the Michigan list.
Another transparent ruse. It's easy to cook the books if you have a biased researcher. Ken Goodman, the creator of whole language, famously conducted the only experiment in which whole language actually beat out a phonics program. It has never been successfully replicated. Reading Recover cooks the books by dropping out the students who do not make progress in their program and excludes them from the results. And, their assessment test includes predictable text which allows students to guess the text based on pictures. This is what DOE was up against.

Oklahoma got the hint that DOE wasn't going to allow these shenanigans again and cause another generation of kids to fail to read.
So instead of advocating scientifically based reading programs, Reading First has promoted programs with "key elements" endorsed by a national reading panel, which could describe almost any program. It may not be a coincidence that the initiative was essentially outsourced to a few experts with a dizzying array of apparent conflicts of interest.
Those "key elements" are, in fact, the "essential components of reading instruction" and are specifically defined in section 1208(5) of the statute and are present in almost every substantive requirement of the statute. Moreover, these essential components could not "describe almost any program" as long as one confines himself to the SBRR underlying these essential components. As long as the program teaches the essential components systematically and explicitly like they do in DI and SFA this would effectively preclude many programs. There is no indication in the OIG report that DOE did anything but follow the statutory requirements.

And, those few experts, as I've pointed out, were, for the most part, THE experts. It's a select bunch. None supposedly had financial ties or were on panels in which the programs they had contacts with were up for consideration.

I'll skip the assessment and professional services portion since it follows the same analysis.
But Johnson said states are ultimately responsible for making sure their programs are scientifically based, which is small comfort for applicants pressured into adopting programs they didn't want. "It's been very frustrating for those of us who really believe in evidence-based programs," said Richard Long, a lobbyist for the International Reading Association, which represents 90,000 reading teachers and specialists nationwide.
The states were pressured into only including programs that were based on SBRR in accordance with the statute. The IRA is a whole language lovin' association who mocks those very same "evidence-based standards." It should have been frustrating for them since there is little doubt that none of the programs they lobbied for were based on SBRR. The OIG report did not find otherwise.

then we get a few paragraphs that show that Reading First appears to be actually working. Grunwald tries to cast doubt on these results.
Of course, $5 billion over five years ought to help states; the question is whether it's helping as much as it should. Without the kind of rigorous studies the law promised but the implementers failed to deliver, it's not clear.
The implementers did what the law permitted. If anything, they were over-aggressive. The fact of the matter is, that if DOE only permitted validated reading programs to get Reading First funding, then only three programs would have been funded at all-- DI, SfA, and, possibly, Open Court. The sad fact is that few reading programs have actual SBRR behind them and states hate the ones that do. The law says what it says and DOE could only go so far. Could you have imagined the squealing there'd be now if two of the three funded programs were McGraw-Hill programs?

Then Grunvald concludes by pointing that some textbook companies made out under Reading First at the expense of others (that was by design) and that, OMG, the Feds are actually demanding some accountability for their funding by excluding bad reading programs. I'm not sure what Grunvald was trying to acomplish with this article, but he's failed miserably at what whatever it was.

No comments: