Orwell Award Announcement SusanOhanian.Org Home


NCLB Outrages

Reading First, Science Last

This is from a talk by Robert Slavin at the Education Writer's Association annual meeting in New Orleans, June 2, 2006. The talk included handouts and PowerPoint material that is not reproduced here.

In 2002, the U.S. Congress passed the No Child Left
Behind Act. It contained a radical idea, long overdue: That
schools receiving federal dollars should use programs and
practices that have been proven to be effective in
scientifically based research, defined as studies in which
programs were compared to matched or randomly
assigned control groups.

Nowhere in NCLB was the concept of scientifically
based research expressed more specifically and
repeatedly than in Reading First.
(Handout: RF law "programs, learning systems, and
strategies to implement methods that have been proven
to prevent or remediate reading failure")

The Reading First legislation could not have been
clearer. Use what has been proven to work it said. Not
"use what is based on principles that have been found to
be promising." Use what has been proven to work.

I work at Johns Hopkins University, and am Chairman
of the non-profit Success for All Foundation. Our Success
for All reading program is one of two reading programs
that have been extensively evaluated and found to be
effective. The other is Direct Instruction. Both have been
successfully evaluated in studies with random assignment
to conditions, and the majority of the dozens of studies of
both programs were done by third-party researchers.
(You have summaries of research on Success for All in
your packet.) Hardly anyone denies that Success for All
and Direct Instruction have exceptionally strong evidence
of effectiveness.

(Handout: Ed Week story quoting Russ Whitehurst
on SFA: "It's a sophisticated study that uses everything the
evaluation field has come to recognize as high-quality."

(Handout from Reading First technical assistance
web site: "In Sum, SFA is well-supported by research in
two ways...")

Naturally, advocates for scientifically proven practice
assumed that a significant portion of the $1 billion a year
in Reading First would go to help schools adopt programs
with proven benefits for children.
We were idiots. Nothing of the kind took place.

What did happen is that at least 95% of federal
money was spent to implement the same commercial
basal textbooks that most schools would have used had
Reading First never existed. Worse, Reading First has
actively pushed out programs with strong evidence of
effectiveness, in favor of commercial programs and
professional development programs lacking even a shred
of evidence.

Reading First has turned out to be a giant step
backward in reading reform, not just because it has nearly
destroyed research-proven programs, but because it has
made a mockery of the idea that scientific evidence
should guide educational decisions for vulnerable
children.

Reading First: The Early Years
When Reading First began, I believe that everyone
involved in administering it was truly committed to the
idea of funding schools to use research-proven programs.

However, when the legislation passed, they realized that
they faced a dilemma. Only two programs, Success for All
and Direct Instruction, had extensive evidence of
effectiveness. One of the five top-selling commercial
basals, Open Court, also had a limited research base. The
others had nothing.

Reid Lyon, the primary architect of Reading First,
expressed the dilemma he faced in a 2005 interview.
(Handout: Reid Lyon Interview: "There weren't enough
programs that went through that level of rigor....the
Department of Education made the decision to make the
criteria more general.")

What Reid Lyon was essentially saying is that once he
and other Reading First leaders realized that insisting on
proven programs would limit schools to just a few
programs, they had to change the rules to emphasize a
much lower standard, programs based on good principles
of practice. This is a key distinction.

The problem is that just about any serious reading
program can claim that it is "based on" scientifically based
research. The operational definition of scientifically based
research was an emphasis on phonemic awareness,
phonics, comprehension, fluency, and vocabulary. Except
for the most extreme whole language basals, what
reading program doesn't include those elements?

Now, I'd be the first to recognize that politically,
Reading First could never have ruled out commercial
basals. They are made by multi-billion dollar companies
that make a lot of campaign contributions. However,
Reading First leaders could have at least let applicants for
Reading First funding know that programs with evidence
of effectiveness were preferred, or even given a few
competitive preference points if applicants included such
programs. At bare minimum, they could have made sure
that such programs were permitted in state and local
applications.

Instead, Reading First leaders did their best to make
certain that very few schools would use research-proven
programs. From the earliest Reading First academies, for
instance, top-selling basal texts were given by name as
exemplars of what Reading First was all about. Programs
with strong evidence of effectiveness were never
mentioned.

Why would they do this? To understand why, put
yourself in the position of a major textbook publisher. To
them, the very idea that schools should choose programs
with strong evidence is anathema. It took Direct
Instruction 40 years to build up its impressive research
base. It took Success for All 19 years. The commercial
basals would be starting from zero. Frankly, their
programs would probably not succeed in well-designed
studies. Remember, standard commercial basals were the
control groups in all studies of Success for All and Direct
Instruction.

The large publishers could not possibly allow
research to have a serious role in textbook selection. So
they put enormous pressure on federal officials,
congressional staffers, and key leaders in academia. They
gave major publishing and consulting contracts to many
of the individuals they knew would be influential. I'll come
back to the issue of conflicts of interest in a moment.

I can't tell you all the ways the big publishers
ensured that evidence of effectiveness would play no role
in Reading First, but I can share with you unequivocal
evidence that this is indeed what happened.

Back in 2003, when the first states were qualifying
for Reading First funding, we were shocked to learn that
the earliest-funded states, such as Michigan, restricted
Reading First funding to the five top-selling basals:
Houghton-Mifflin, Harcourt, Scott-Foresman, Macmillan,
and Open Court. The two programs with strong evidence
of effectiveness, Success for All and Direct Instruction,
were not allowed.

In 2003, we met with Chris Doherty, then federal
director of Reading First. In a one-hour meeting, he spent
55 minutes repeating over and over again that there was
no federal list of approved basals. If those darn states
chose the five top-selling basals, well, it's a shame, but
there's no list, no list, no list. But there was a list.
(Handout: Reading First (Florida TA center) web page, "a
list of core reading programs that were aligned with the
basic RF standards and were suitable for use by RF
schools in Florida...[is] Harcourt, Houghton Mifflin, Open
Court, Reading Mastery Plus, and Scott Foresman")


Here's a page from the web site of Joe Torgesen's Florida
Center for Reading Research, which later became one of the Reading First
technical assistance centers. To his credit, Torgesen,
unique among Reading First leaders, did review evidence
of effectiveness of some reading programs, and gave a
very positive rating to Success for All and Direct
Instruction.

However, note that Torgesen's review did not
present evidence of effectiveness for the six top-selling
basals. They were given a free pass.
What this said to states desperate to get their share
of $1 billion per year was this:

Picking a standard basal is safe
Picking anything else is risky

The University of Oregon, another Reading First center,
also published a list of
basals "based on scientifically based research." Six of the
top seven were the top-selling commercial basals.

The state proposal writers were no fools. They knew
an approved list when they saw one.
(Handout: Michigan list - Houghton Mifflin, Harcourt,
Open Court, MacMillan, Scott Foresman)

Here's Michigan's list. By amazing coincidence, it
consists of five of the six programs "excused from having
to have evidence" on the FCRR web site, down to the
requirement to supplement Scott Foresman. The only
missing program on the FCRR list but not the Michigan
list was Reading Mastery, which, along with extensive
training, is used in the Direct Instruction program.
(Handout: Wisconsin list - Houghton Mifflin, Harcourt,
Open Court, MacMillan, Scott Foresman)

Here's Wisconsin's list. Not just the same list, but the
same chart, the same order of programs. Go with a
winner, as they say in Wisconsin.
(Handout: Rhode Island list- Houghton Mifflin, Harcourt,
Open Court, MacMillan, Scott Foresman)

Here's Rhode Island's list. Look familiar? Rhode
Island had to revise its proposal six times. Not until they
settled on this list, and forbade their districts to choose
anything else, did their federal reviewers finally accept
their proposal.
(Handout: initial Oklahoma proposal: "[Districts] must
choose a program that provides documented evidence of
a minimum of three years of longitudinal data from a
rigorous study.")

Some states actually read the words in the legislation
and guidance and thought they meant what they said.

Here's Oklahoma's rejected language. Imagine! Proposing
to ask Reading First schools to have scientific evidence of
effectiveness! Reviewers twice rejected this language.
(Handout: Final OK language: "Any reading program
adopted by a Reading First school from this approved
[textbook adoption] list may be used...")

Finally, Oklahoma got with the program. Here's the
language that replaced the earlier insistence on evidence.
Just pick any basal from the state adoption list and you're
A-OK. Want to guess what texts were on their list? One
immediate result when Oklahoma was funded is that three
very successful Success for All schools were forced to
drop their evidence-based programs and implement Scott
Foresman, which has zero studies showing its
effectiveness.

I could show you proposals from state after state,
and you'll see excerpts in the thick document in your
folders. However, I know you're waiting impatiently for the
conflict of interest part.

As I said earlier, the publishers early on identified
people who they knew would be leaders in Reading First
because they had been leaders in the earlier Texas
reading reforms carried out when George W. Bush was
governor.
(Handout: Tech center leaders - Torgeson, Kame'enui
(with Simmons), Vaughn)

Here's a list of the leaders of the three Reading First
technical assistance centers.
(Handout: Reading Street ad - authors are Kame'enui,
Simmons, Vaughn)

Here's a list of authors of the 2007 Scott Foresman
reading series. By coincidence, the technical assistance
center leaders and their colleagues are also authors of
America's largest basal series. True, they are authors of
the 2007 edition, so they were not making royalties on
the earlier textbooks schools were pressed to use. But I
was told by Pearson's president that Ed Kame'enui,
Deborah Simmons, and Sharon Vaughn were to be Scott
Foresman authors in 2003, early in Reading First, and
they knew this too. Ed Kame'enui and Deborah Simmons
of the University of Oregon center were already authors of
a Scott Foresman supplemental program used in Reading
First (in fact, a link to enable readers to add this text to
their shopping cart was conveniently placed on the
Oregon Technical Assistance Center's web site until
recently). Were these individuals beholden to a publisher
of America's most widely used reading text who was
going to give them hundreds of thousands of dollars in
royalties for very little work? You decide.
(Handout: Voyager design team - Kame'enui, Vaughn,
Simmons, Good)

Here's the "design team" of consultants who helped
create Voyager, another commercial basal, and Voyager
Passport, a supplemental program designed specifically to
take advantage of Reading First. Do you see the pattern?
Every person on this list (except a Voyager employee) was
a leader of the Oregon or Texas Reading First technical
assistance centers, and was enormously influential in
Reading First even before their centers were funded.

Voyager, founded by Randy Best, a Dallas friend of
George W. Bush, is a particularly interesting story. Before
Reading First, Voyager had lost money for eight years in a
row, according to press reports. Last year, after profiting
massively from Reading First, Voyager was sold for $361
million.

It may be a coincidence, but after Voyager was sold,
Reid Lyon, friend of George W. Bush and architect of
Reading First, left government to join Randy Best's
company. Rod Paige, Secretary of Education when Reading
First was beginning, has also joined Randy Best's
company.

I can't tell you that there was any corruption involved
in any of this. The Department of Education's Inspector
General and the GAO are doing major investigations of
the whole sordid business. However, this I can tell you.
Somehow, in a program intended to help at-risk children
gain access to proven reading programs, the opposite
happened.

What happened is that rather than receiving proven
programs, millions of vulnerable children were used as
guinea pigs in a massive experiment, receiving methods
and instructional strategies that have never been
evaluated anywhere. When Reading First comes up for
renewal, it will have wasted $6 billion intended to improve
the lives of at risk children. Reading First has only
enriched large publishers and academics who, at best,
were too nave to know they were being used.
(Handout: List of reforms)
The last page of your packet is a set of
recommendations intended to bring Reading First back to
its original purpose. I'd suggest you read it at your
leisure.

I continue to believe that American education will
only advance when we all make a commitment to use the
process of research and development that has made
America great in medicine, agriculture, engineering, and
other fields. Reading First, which started with the same
vision, has instead demonstrated once again that politics
and greed trump research and benefits to at-risk children
every time. It is not too late to correct this well-
intentioned program, but we must do so very soon. Our
children deserve no less.

— Robert Slavin
a presentation at the Education Writer's annual meeting
June 2, 2006


INDEX OF NCLB OUTRAGES


FAIR USE NOTICE
This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available in our efforts to advance understanding of education issues vital to a democracy. We believe this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. For more information click here. If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner.