Reading for Profit
Ohanian Comment: The author seems to accept as given the definition of 'scientific' reading and suggests that if Reading First had operated according to that science, Success for All and Direct Instruction would be in everyone of its schools.
The ore Rep. George Miller talks, the more hope we abandon:
Reading First "is an important part of meeting the No Child Left Behind Act's goals of closing the achievement gap. The problem is with how the Department of Education has implemented the program, not with the program itself."
--Rep. George Miller
Oh yeah? You don't read your mail, Rep. Miller. I personally have mailed you multiple copies of Ken Goodman's book on DIBELS. All unacknowledged.
Whistle-blowers allege that U. of Oregon scholars steered bounty from the No Child Left Behind Act to themselves and their colleagues
By David Glenn
When President Bush signed the No Child Left Behind Act in January 2002, Robert E. Slavin, a researcher at the Johns Hopkins University's Center for Social Organization of Schools, quietly cheered.
During two decades of work at Johns Hopkins, Mr. Slavin and his colleagues had developed a comprehensive school-reform model known as Success for All. He believed that the program offered powerfully effective techniques — including explicit phonics instruction — to reach young children from low-income families.
A number of experiments by outside evaluators seemed to back up Mr. Slavin's assertions. And as No Child Left Behind worked its way through Congress in 2001, colleagues and policy experts repeatedly told him that Success for All would be a perfect match for one of the new law's keystone programs: Reading First, a $900-million-a-year project that offers states grants to improve reading instruction for children in kindergarten through third grade.
After all, the new law required states to spend the federal money only on products that were grounded in "scientifically based reading research." Mr. Slavin's model had as strong a scientific pedigree as any program in the country, and it was already used in approximately 1,500 schools.
With the advent of Reading First, Mr. Slavin hoped to add another 200 schools, and perhaps even more. Even before the law was signed, the Success for All Foundation, a nonprofit organization of which he is chairman, took on extra staff and rented new office space.
But when Reading First actually got under way, Success for All was almost entirely shut out. Fewer than 5 percent of Reading First schools use Success for All — and most of those had used it even before Reading First began. Worse, some schools that had long used the method have abandoned it because, Mr. Slavin says, state officials and federal consultants have urged them to conform to the typical practices of Reading First schools. Instead, most states spent their grants on commercial textbooks and other products that Mr. Slavin regards as unproved, at best. He worries about what Diane McGuinness, an emeritus professor of psychology at the University of South Florida, has called "junk phonics" — that is, textbooks that take on the trappings of phonics instruction but do not actually present it effectively.
After the expected contracts failed to materialize, the foundation was forced to lay off more than half of its staff. Downstairs from Mr. Slavin's office lies a large suite of empty cubicles.
Frustration spurred Mr. Slavin and his colleagues to action. In early 2004, they began to make phone calls, trying to learn why they had been excluded. Later that year, they filed a formal complaint with the U.S. Department of Education.
That complaint has snowballed in a way that even Mr. Slavin might not have imagined. Reading First is now awash in allegations of mismanagement and conflicts of interest. Last September the Education Department's inspector general issued a report that declared that the program's managers had violated a long list of rules and procedures. In February the Government Accountability Office is expected to weigh in with a report of its own. And Rep. George Miller, Democrat of California, the new chairman of the House Committee on Education and Labor, plans to hold hearings soon on Reading First, a program he championed in 2001.
"The law laid out specific guidelines that the Department of Education chose to ignore or to interpret in such a way as to serve its own agenda," Representative Miller said in an e-mail message to The Chronicle.
Some of the most serious allegations of malfeasance, all of which remain in dispute, center on a close-knit group of educational psychologists at the University of Oregon.
Three of those scholars have been accused of personally profiting from their roles as federal contractors who offered technical advice to state governments about how to design their Reading First programs. In many cases, those states purchased textbooks or other materials designed or written by the Oregon scholars themselves. Those purchases are now generating tens of thousands of dollars in annual royalty income.
But the Reading First controversy has much higher stakes than professional pique or profit. Representative Miller's hearings may become a staging ground for a grand battle between proponents and opponents of the entire No Child Left Behind law. The central question is whether the federal government can — or should — use its leverage to promote scientifically grounded classroom instruction.
Mr. Slavin demands that the government clean up its act before attempting any new such programs. But other scholars insist that Reading First is highly effective, and they fear that it will be destroyed because of misplaced anger over a few procedural violations. The complaints raised by Mr. Slavin and his allies, they say, have been tendentious and contradictory.
"I worry that Reading First is going to be a political casualty," says Roland H. Good III, an associate professor of school psychology at Oregon whom Mr. Slavin has accused of conflicts of interest. "Five years ago, if you'd told me that there was going to be an initiative implemented on this scale, and that it was going to have this kind of impact, I would have told you that there's no way that that is going to happen. I have been astounded at what has been accomplished."
The federal government had gathered interesting data about effective classroom instruction since at least 1965, but in Mr. Slavin's view it had generally let that advice sit on a shelf. He believed that in Reading First, Congress had finally devised a mechanism that would encourage local school districts to choose effective programs.
Just weeks before the No Child Left Behind law was completed, a significant change was made in the statutory language that established Reading First. In early drafts, states would have been required to spend their grants only on programs with "strong evidence of effectiveness," based on randomized experiments published in peer-reviewed journals. But toward the end of the process, lawmakers became concerned that such a standard would restrict states to a very few programs, according to Robert Sweet, a former Republican Congressional staff member. Mr. Sweet, who is now president of the National Right to Read Foundation, a pro-phonics organization, says that only two programs seemed to qualify: Success for All and Direct Instruction, a nonprofit school-reform model that was developed almost 40 years ago.
Such a narrow menu would have seemed too much like a federal curricular diktat, Mr. Sweet says, so the bill was softened. States were required to spend their grants only on products "in alignment with" the five elements of effective reading instruction that were identified in a major report issued in 2000 by a federally sponsored committee known as the National Reading Panel (see box).
That change did not bother Mr. Slavin. Indeed, he says that it would have been absurd to expect Success for All and Direct Instruction to expand quickly enough to serve all 5,500 schools targeted by the program. Of course, he says, states would inevitably spend a large fraction of their Reading First grants on certain commercial textbooks that had recently beefed up their phonics content. But he did expect the federal government to energetically remind states that Success for All and Direct Instruction had (at least at that moment) stronger evidence of effectiveness than those recently revised textbooks.
That did not happen. At the "Reading First Academies" — federally financed conferences that introduced the program to state officials in 2002 — Success for All and Direct Instruction went essentially unmentioned.
That silence seemed especially peculiar to Mr. Slavin because several of the speakers at the Reading First Academies were past or present faculty members at Oregon who had begun their careers immersed in the culture of Direct Instruction. (Direct Instruction was developed at the University of Illinois at Urbana-Champaign in the mid-1960s, but its creators decamped for Eugene in 1970.) Why, he wondered, did they not tout the virtues of a nonprofit program they obviously knew well? Could it be because some of them had relationships as authors or consultants with for-profit publishers?
The presiding spirit at Oregon is the famously blunt scholar Siegfried E. Engelmann, who was one of Direct Instruction's two primary creators. Mr. Engelmann has retired from his position as a professor of education but continues to work as a consultant for Direct Instruction schools. He shares Mr. Slavin's annoyance that their nonprofit programs have failed to thrive under Reading First. "When Slavin filed his complaint," Mr. Engelmann says, "he had a point. Neither Success for All nor Direct Instruction benefited from the program. The point was not just to benefit commercial programs but to use programs with evidence of effectiveness."
But he adds that Mr. Slavin might be too pessimistic about the quality of commercial textbooks, which were "absolutely pathetic in the mid-1990s, but some of them are pretty decent programs now."
A Hint of Trouble
Mr. Slavin points to the first state application approved under the Reading First program as an early, vivid hint of trouble. In late 2002, Michigan's application was endorsed with the proviso that its districts choose from among the materials of five commercial programs that the state determined to be in alignment with the five elements. Success for All and Direct Instruction were not included on the list.
"There it was," he says. "Five basal textbooks. And their best friends wouldn't say that any of them, with the exception of Open Court [a series published by McGraw-Hill], had anything like evidence of effectiveness at that time. And yet there was Michigan saying that these were the five programs that you have to use, and no others."
Three thousand miles away, scholars in Eugene were hard at work getting Reading First off the ground. Because the Education Department had hired just two staff members to direct the program from Washington, almost every element of its administration has been outsourced to various entities.
Several University of Oregon scholars accepted contracts to act as advisers to states that were struggling to draft their Reading First grants. At least seven Oregonians also took roles on committees that were charged with reviewing various textbooks and other products for alignment with the five elements.
(In 2004, after the states' grants had been set in stone, Oregon won a contract to house one of three major technical-assistance centers for Reading First. The others are located at Florida State University and the University of Texas at Austin.)
Among the early Oregon efforts was a committee that later attracted the attention of Mr. Slavin and investigative reporters for the newsletter Title I Monitor. In the summer of 2002, the Education Department was eager to make an official statement about which assessment tools — that is, the tests used to monitor students' week-by-week progress — were scientifically valid and in alignment with the five elements.
In a collaboration with the National Institute for Literacy, a federal office, the department financed an "assessment review committee" that was composed of eight scholars, four of whom were on Oregon's faculty. That committee wound up endorsing, among other products, the Dynamic Indicators of Basic Early Literacy Skills, or Dibels — a product developed by Mr. Good and his colleague Ruth A. Kaminski.
Reading First has been a great windfall for Dibels, whose materials are published by Sopris West, a company based in Longmont, Colo. Thus far, Dibels has been adopted by at least 35 states' Reading First programs, producing hundreds of thousands of dollars in royalty income.
Back-Scratch or Best Practice?
Mr. Good himself was one of the eight scholars on the assessment-review committee. He says, and others confirm, that he recused himself from evaluating his own product. But critics say that the committee was generally a back-scratching session.
"I think these conflicts are going to bring Reading First down," says Kenneth S. Goodman, an emeritus professor of education at the University of Arizona and a past president of the International Reading Association, a professional organization for reading teachers. "And the Dibels committee is the most glaring example." (Mr. Goodman, it is worth noting, is skeptical of the entire phonics-based turn in reading instruction. He and Mr. Slavin do not often find themselves on the same side of an argument.)
The committee was marred by several instances of procedural oddness. First, its report was publicly released without having been vetted by the National Institute for Literacy. And, according to the September report issued by the Education Department's inspector general, the original director of Reading First, Christopher J. Doherty, and two Oregon scholars exchanged e-mail messages in which they discussed the possibility of being deceptive about the committee's federal financing.
But even if Washington bureaucrats played odd games with the committee's work, that does not necessarily invalidate the committee's conclusions.
In his office in Eugene, Mr. Good speaks passionately in defense of his work. He has already been interviewed at length by representatives of the inspector general, and some of his answers have a well-rehearsed quality.
As a school psychologist in Pennsylvania in the early 1980s, Mr. Good says, "I practiced in the best way that I could with the training that I had, but I felt very — well, there was a tension in that. I felt like I was working as hard as I could to make a positive difference in children's lives. But I didn't think I was making enough difference."
Then he discovered the emerging technique known as "curriculum-based measurement," which was developed by scholars at the University of Minnesota-Twin Cities. Curriculum-based measurement encourages teachers to evaluate how children in special-education programs are progressing on specific skills that are tied to the ordinary classroom curriculum. (Previously, school psychologists had primarily given broad, IQ-style tests that were not so linked to what went on in school.)
Mr. Good says Dibels is an extension of the curriculum-based-measurement concept. With his tool, teachers can do weekly assessments of children's progress in phonemic awareness and oral reading fluency. One advantage of Dibels, if it is used properly, is that it presents children with texts that they have not seen before, so there is little danger of teaching to the test. The test is designed as a sort of weekly progress report, not as a high-stakes measure that should affect students' grades or teachers' salaries, Mr. Good says.
Mr. Good also touts Dibels's efficiency. The assessment can be administered in less than two minutes, and teachers can record and graph the scores with a hand-held digital device. "We don't want teachers to feel as if they're spending all of their time testing," he says, "and no time actually teaching."
Mr. Good says he does not believe his role on the assessment-review committee involved a serious conflict of interest, for three reasons: He recused himself from evaluating his own product; Dibels is available in a free online version, so it isn't a purely commercial product; and he and Ms. Kaminski had already agreed to donate all of their royalty income through 2005 to a research fund administered by the University of Oregon Foundation. (Despite relinquishing those royalties, Mr. Good concedes that he earns a substantial income giving seminars to states and local districts about how to use Dibels.)
Blowing the Whistle
In early 2004, Mr. Slavin abandoned his hopes that Success for All would be broadly included in Reading First.
He and his colleagues began their own investigation, filing open-records requests and working the telephone. Mr. Slavin fixed his gaze on what he saw as evidence of his model's exclusion, including a PowerPoint presentation given at federally sponsored conferences in 2002 by Marcy L. Stein, a professor of education at the University of Washington at Tacoma.
The presentation was designed to teach state officials how to choose a core instructional program for Reading First. While it included examples taken from at least three commercial textbooks, the presentation never mentioned Success for All or Direct Instruction.
Ms. Stein responds that she had no intention of plumping for any commercial products. She says that she would gladly have used examples from Success for All if she had had any at hand, and adds that she would never consciously slight Direct Instruction, with whose creators she has often collaborated. (She earned her master's degree at Oregon.)
The point of her presentation, Ms. Stein says, was to teach state officials how to use "A Consumer's Guide to Evaluating a Core Reading Program, Grades K-3: A Critical Elements Analysis," a document posted on Oregon's Web site in early 2003. The "Consumer's Guide" explains, in 58 pages of elaborate detail, how to tell if a particular textbook is in alignment with the five elements required under Reading First. For example, in the realm of "phonemic awareness," does the textbook focus on "segmentation or the combination of blending or segmenting for greatest transfer"?
Ms. Stein says that she originally drafted her presentation with no examples, but discovered that it seemed impossible to explain the "Consumer's Guide's" arcane formulations — especially to state officials who often had little or no experience with phonics — without showing examples of actual textbook pages. So, after vetting the idea with the department's lawyers, she added a few samples of real-world textbooks.
Mr. Slavin says he accepts that Ms. Stein's intentions were benign, but he also suggests that she and others were kidding themselves if they believed that many state officials were actually going to check to see if various products met the "Consumer's Guide's" long list of specifications.
The audiences listening to Ms. Stein's presentations, he says, were really just desperately hoping to hear the bottom line: Which products do we need to include in our application to get our state grant approved? And in that context, the fact that Ms. Stein favorably mentioned three commercial textbooks carried enormous weight.
"The original state proposals were almost all turned down in the first round," Mr. Slavin says. "And some of them were turned down multiple times. Rhode Island was rejected six or seven times, depending on how you count ... When you're the grant writer for a $30-million proposal to get money for K-3 for your state, it's very embarrassing to be rejected."
'Look at the Oregon Web Site'
Mr. Slavin concedes that in certain instances, those state applications may have been rejected for good reason, perhaps because they proposed to use "whole language" programs or other products that clearly lacked the five elements prescribed in the law. In any case, after states were rejected, they would almost always call Mr. Doherty, then the director of Reading First, and ask what they should do differently when they revised their applications.
And those conversations often had a Marx Brothers quality, according to Mr. Slavin's accounts and the inspector general's September report. If, say, a state had proposed to use a program that obviously fell outside Reading First's statutory requirements, Mr. Doherty would not say so clearly, for fear of violating the law that forbids federal officials from dictating state and local curricula. Instead, he would give vague suggestions like "Look at the Oregon Web site" or "Look at the Michigan list."
State officials gradually learned through trial and error, Mr. Slavin says, that their applications would be approved if they proposed to use any or all of five particular commercial textbooks as their core instructional material. And insofar as that trial-and-error process involved "looking at the Oregon Web site," Mr. Slavin believes that there was at least one serious conflict of interest in play.
The Web site featured a list of core instructional products that a committee of Oregon scholars deemed to be in alignment with the five elements, according to a formal analysis of the "Consumer's Guide." (Here, in other words, was the "bottom line" that Mr. Slavin believes state officials were searching for.) Among the products on the list was a textbook series published by Pearson Scott Foresman, which is one of the five commercial series that have been widely purchased under Reading First.
In 2003, Pearson Scott Foresman hired as authors of the next edition of its textbook series two of the Oregon scholars who created the list: Edward J. Kame'enui (who is now on a two-year leave as commissioner of special-education research at the U.S. Department of Education) and Deborah C. Simmons (who has subsequently moved to Texas A&M University at College Station). Mr. Kame'enui and Ms. Simmons had already written a supplemental product for Pearson Scott Foresman.
When Mr. Kame'enui took up his government post in 2005, he filed a conflict-of-interest disclosure that indicated that he earned between $100,000 and $150,000 annually in royalties from that earlier product.
Shortly thereafter, the Oregon list was altered so that Success for All was moved from the fifth position to the seventh, and Pearson Scott Foresman was moved from seventh to fifth.
In an e-mail message to The Chronicle, Ms. Simmons says that the list was altered after the committee removed the "vocabulary" and "comprehension" elements from its analysis of the products. That was done, she said, because the 2000 National Reading Panel did not establish definitive ways to teach vocabulary and comprehension.
Mr. Slavin says that he finds that explanation "ridiculous. Imagine Henry Ford sitting on a government panel to evaluate cars, and saying, well, let's remove the ratings for safety and repair records. Oh, gosh, removing those ratings turns out to hurt Toyota's ranking and help Ford's. What a surprise."
In any case, various records on Oregon's Web site back up Ms. Simmons's account. In 2003 the scholars removed vocabulary and comprehension from their product analyses — and they did so consistently, not just on the "core program" list where Success for All's ranking suffered.
More than any of Mr. Slavin's other accusations, this one inspires eye rolling among his targets. Whether in fifth position or seventh, they say, Success for All was prominently featured on the Oregon list — so the list seems to be an unlikely villain in Mr. Slavin's tale. Some other factor must explain Success for All's failure to thrive in Reading First. And isn't Pearson Scott Foresman's hiring of Mr. Kame'enui and Ms. Simmons — two scholars with long track records in reading research — exactly the sort of thing that Reading First was designed to encourage? The new textbooks might not have Success for All's high evidence of effectiveness, they say, but with careful studies it should be clear in five or 10 years whether they are reasonable substitutes for Success for All and Direct Instruction.
"Textbook publishers have told us for a long time that if we want to include more scientifically based material, we'll have to change the market and create a demand for that material," says Ms. Stein. "Reading First has allowed us to change the market."
Mr. Slavin agrees that Pearson Scott Foresman's hiring of Mr. Kame'enui and Ms. Simmons is welcome, but he says that they should immediately have recused themselves from any role in advising states that were still in the process of building their Reading First packages. "If they'd been government employees, this could never have happened," he says. "But because they were government contractors, they somehow felt that these things didn't apply to them."
Ms. Simmons, however, strongly denies any wrongdoing. "I maintained an objective, unbiased, and uncompromising role whenever I assisted states in developing a review process," she says.
Baby and Bathwater
What does all of this portend for the future of Reading First — and for scientifically based education reform more generally? The inspector general's report recommended a number of narrow procedural reforms, most of which the department says it has embraced. Among other things, the Washington staff has expanded from two to seven.
Mr. Slavin says that, despite everything he despises about the program, he expects that it will turn out to have had some significant positive effects. (A large-scale federally financed study of student achievement under Reading First is expected to be released in early 2008.) "With a billion dollars a year, and with new reading coaches in every school, and with an average of 20 minutes more reading instruction a day, they've almost certainly managed to do something positive," Mr. Slavin says.
"The program has clearly been run unethically," says Jack F. Jennings, director of the Center on Education Policy, a nonpartisan Washington think tank that has issued two long reports on Reading First. "They wanted to get what they thought were the right programs in place, but they trampled on ethics and procedures in the process."
At the same time, Mr. Jennings says that the program has important virtues that he hopes are not lost. "It would be a shame to lose sight of the good effects," he says. "It's striking to me how well the program seems to be raising students' test scores. And principals consistently say that they like it."
Representative Miller agrees. Reading First "is an important part of meeting the No Child Left Behind Act's goals of closing the achievement gap," he says. "The problem is with how the Department of Education has implemented the program, not with the program itself."
One veteran of the Education Department, meanwhile, says that he feels badly burned by the controversy, and wonders if Reading First ought to be radically revised. Michael J. Petrilli, a vice president of the Thomas B. Fordham Foundation, a Washington think tank, and a former official in the department's Office of Innovation and Improvement, says that the inspector general's September report "was an incredibly unfair depiction" and "completely missed the policy context."
A muddy fight was almost inevitable, Mr. Petrilli says, because "Congress was talking out of both sides of its mouth. At the same time that it was telling the department not to fund nonscientific programs, Congress was highly squeamish about the department appearing to dictate local curriculum."
"From the evidence that I can see," Mr. Petrilli continues, "that prescriptive approach has worked very well at improving students' learning. On the other hand, all of the political blowback that we've seen suggests to me that this kind of effort is just not sustainable. We might be better off if the federal government just stepped back and focused on clarifying the results that schools are supposed to achieve in terms of student learning, and then allowed those schools to pursue those targets with whatever techniques they deem best."
Reading First has had one important unappreciated virtue, says Ms. Stein. Although Success for All and Direct Instruction have been largely excluded — an exclusion that Ms. Stein says is regrettable — their spirits have been incorporated throughout the program. Participating schools are required to intensively coach their teachers and to monitor their students' progress each week, and in each case, the techniques have been pilfered to a large extent from Success for All and Direct Instruction. "Every successful school reform model has had those elements," Ms. Stein says. "We were able to incorporate those essential components throughout Reading First."
Mr. Slavin, for his part, hopes that the program is structured very differently if it is renewed. "Common sense will tell you," he says, "that it was absurd to subcontract this entire billion-dollar program to universities, with only two people on staff in Washington to monitor the work. That was bound to fail."
THE 5 ELEMENTS
The No Child Left Behind Act, which established the Reading First program, names five essential elements of reading instruction and includes an elaborate definition of "reading" itself.
The five essential components of explicit and systematic instruction in reading:
1. phonemic awareness
3. vocabulary development
4. reading fluency, including oral reading skills
5. reading-comprehension strategies
The term "reading" means a complex system of deriving meaning from print that requires all of the following:
1. The skills and knowledge to understand how phonemes, or speech sounds, are connected to print
2. The ability to decode unfamiliar words
3. The ability to read fluently
4. Sufficient background information and vocabulary to foster reading comprehension
5. The development of appropriate active strategies to construct meaning from print
6. The development and maintenance of a motivation to read
Chronicle of Higher Education
INDEX OF NCLB OUTRAGES