NICHD Moving Research Into Practice in Schools and Classrooms

It’s not easy getting schoolteachers to put into classroom practice the lessons that scientific research says can improve their results, and a big part of the difficulty lies not with the teachers but at the doorstep of science itself. G. Reid Lyon is in a position to know.

A developmental neuropsychologist, Lyon heads the Child Development and Behavior Branch (CDB) at the National Institute of Child Health and Human Development. He says he has “firsthand experience of the devastating effect” that poor research can have on teaching practices and teachers’ trust in educational research. He learned that lesson when he taught reading to 28 third graders in the mid-1970s.

Lyon recounted that experience to the U.S. House Committee on Education and the Workforce in May 2000:

“At the beginning of the year, a third of my students could not read well enough to understand what they had read. Their reading was slow and labored and they mispronounced words constantly. Their spelling was lousy.”

He’d taught them just as he’d been taught to do in his education courses, had been guided by the school’s philosophy and curriculum, and had followed the suggested activities in his teachers’ manual. Still, “At the end of the year, the same third of my students could not read well and their spelling was still lousy. The only change that I could discern was that their motivation to learn to read had waned, and their self-esteem had suffered substantially. Likewise, I felt like a failure. I had let down the children I was responsible for.”

Only later did he discover the source of that failure: his school’s approach to teaching “was not only based upon research that was questionable at best, but the major assumptions upon which the instructional philosophy and recommended teaching interactions rested had never been adequately tested through well-designed studies.”

No wonder many teachers lose trust in research, he said. “Those that stay in the profession learn to simply ‘wait out’ the next ‘research-based’ instructional magic bullet.”

Lyon chose a different route. He left the classroom and set a course that eventually took him to the NICHD in 1991.

Under his direction, the CDB has heavily engaged in translating rigorous research, particularly how best to teach reading, into policies and practices in schools and classrooms.

About 38-40 percent of all fourth grade children can’t read and comprehend even a simple paragraph, and among those from low-income and minority populations, that rate soars to 70 percent. As early as the end of the first grade, Lyon says, children whose reading skills lag behind their classmates’ are already on the road to reading failure, and their inability to read and comprehend will mean continuing difficulty learning all other subjects as well. Lyon’s program at NICHD seeks to break that spiral of failure.

“For the past 12 years, we’ve been doing well designed clinical trials in public schools, looking at kids five years of age who are at risk for reading failure,” he explains. “Those kids are assigned to different instructional procedures, and we’ve been able to identify which kids respond to what kinds of approaches in what kinds of settings, and what they need to produce effectiveness.

“You can’t do this unless you bring the entire school and school system into the mix. We are supporting research that provides teachers with professional development and that allows us to interact minute by minute daily in the classrooms, monitoring the fidelity of interventions, having teachers and principals help us modify the protocols to correct deficiencies, and so on.”

Last year his branch completed a five-year program that put full-time research teams inside each of eight Washington, DC, schools. Now he has multi-disciplinary teams in 12 school systems nationwide, involving 12,200 students and 1,500 teachers in 960 schools. In addition to educators and psychologists, the teams include speech and language pathologists, pediatricians and even neuroscientists.

“What we’re doing is bringing to bear multiple protocols,” Lyons explains. “We are able to image the child’s brain both before and after the interventions to determine whether or not the interventions are not only promoting reading gains but also changes in neuro-physiology.”

The interventions all involve “highly at-risk children,” he says, and he’s achieving dramatic results. “When we get to the kids early, if we can get to them by the second half of kindergarten with that which we know is effective, we can reduce that failure rate down to 5 or 6 percent.”

Lyon traces the coming-of-age of translational research in his branch to 1996, when both state and federal governments sounded alarms that the education system was failing America’s children. Then-Gov. George W. Bush, he says, was “asking very critically and very openly for objective information that he could bring to bear to help make decisions about education in Texas, both for kids whose first language was English and for those whose first language was Spanish. He was extremely interested in providing kids with scientifically based educational programs. He was aware that in the past, education was guided more by beliefs and philosophy than by science.”

Meanwhile in Washington, U.S. Rep. William Goodling (R-PA), then chair of the House Education and Workforce Committee, was concerned that past and proposed educational programs were not based on objective evidence and most likely would not lead to improved results. He asked Lyon for a one-on-one briefing. “We spent a good part of the day going through the data – what we knew, what we didn’t know, and what we would need to look at in the future.”

Since then, Lyon has testified every year before Congress about scientific information on reading development and instruction. He now also serves as the President’s advisor on child development and education research and policies.

NICHD was also the major supporter of the congressionally established National Reading Panel, which reviewed the scientific literature to identify gaps in the educational research. The panel issued a landmark report in 2000 that found fewer than one-third of the studies over the past 30 years that had tried to test the effectiveness of reading instructional approaches had met basic scientific criteria.

“Many studies,” Lyon told Congress, “did not have even the most rudimentary elements of scientific methodology, such as adequate control or contrast groups; many did not define study participants or instructional approaches sufficiently to permit application in the classroom or replication by other studies, and many did not measure student achievement outcomes appropriately. Thus, it is not surprising that many teachers and researchers have lost faith in the ability of quantitative research to inform instructional practices over the years.”

“The effects on educational policy-making and teaching practices are insidious and harmful. Information derived from poorly designed and conducted studies will inevitably produce recommendations that are doomed to failure at the system level, the school level, and the classroom level. When teachers turn to research to inform their teaching, they expect and deserve information that is trustworthy. When the information is not, which it typically is, teachers fail, students fail, schools fail, and our nation fails.”

The panel’s report laid the groundwork for the president’s legislation, the “No Child Left Behind” Act, and its “Reading First” program now being implemented nationally. “This is the first time, I think, that a President has asked that a scientific arm of the federal government be involved in translating the findings of research into actual legislation,” says Lyon.

Unfortunately, the intensive team interventions Lyon’s program has been testing would be far too costly to replicate in every school in the nation. Instead, NICHD is searching for a more cost-effective arsenal of interventions by attacking learning difficulties from multiple directions – pre-school readiness to learn, parental and adolescent illiteracy, learning to read English as a second language, learning mathematics and science, childhood experiences of abuse, neglect and violence, even mapping the development of children’s brains.

Each program informs the others. “We are very aware that you can’t move research in this particular arena forward through developing silos of information,” Lyon says. “The programs are highly integrated and correlated. A great deal of collaboration goes on between the programs within the branch and with other agencies,” notably the U.S. Department of Education (DOE).

On his second day in the White House, President Bush asked the CDB to determine what interactions children need from birth to age five in order to develop social competency, emotional health and the cognitive and language capabilities to be prepared for school. That led to the Early Learning and School Readiness program, directed by Kyle Snow, a developmental psychologist.

Snow now has two initiatives to fund research on experimental interventions, curricula and programs that might one day be translated into real-world settings. “We have a broad research base that has told us a great deal about how kids develop, both typically and not typically,” Snow says. “Given what we know about how kids develop, how can we put that knowledge into play to positively affect the lives of kids and prepare them for school?”

For example, he says, “There’s a pretty well known correlation between level of parental education, particularly maternal education, and child outcomes. It’s a reasonably robust positive correlation across a broad range of academic measures. What you’d like to do is develop some program that would break that cycle so you’d have kids doing well in school regardless of parental education.”

Unfortunately, before basic knowledge can be translated into real-world interventions, a two-headed problem that besets the field of early childhood learning must be resolved.

“Given how much is going on in service delivery, there hasn’t been much subjected to rigorous scientific evaluation,” Snow says, and compounding that, “they don’t keep very good records of what they’re doing, what the program looks like, what’s actually happening in the classroom or in the child care room. When you put those two together, you have lots of anecdotal, low-quality qualitative work but an inability to describe – or evaluate – the program very well.”

There are exceptions, of course, notably “The Chicago Child-Parent Centers” run by Arthur Reynolds, of the University of Wisconsin, Madison, which are demonstrating that intensive daily intervention continues to have positive effects 14 and 15 years downstream. “What we’re looking to fund is a larger number of that kind of quality work,” Snow says.

Of immediate concern is a national push to apply rigorous evaluation measures to Head Start, the nation’s federally-supported pre-school program that dates from the “Great Society” of the 1960s. Operators of local Head Start centers worry that the demand for evaluation is a political gambit to terminate or sharply curtail their funding.

Not necessarily so, says Snow. “The research community has to answer these questions: ‘How well can you measure kids, and how is it important?'” The researchers, however, must do so in concert with the policy and practice communities – in effect erecting a trilateral translational bridge that has to keep communication flowing in multiple directions.

This puts “all three balls in the air, a trifecta” of research, policy and practice, Snow says. “You have a perfect excuse to get them all together in the same room” to begin the process of developing not only good measurement tools, but good dialog. To jump-start that process, NICHD, in collaboration with the Administration for Children and Families, the DOE and the National Institute of Mental Health (NIMH), funded a workshop in June 2002 that brought together researchers, practitioners and policy makers to discuss what pre-school outcomes should be measured and how.

The attendees raised a cautionary flag about some frequently used measures. “One of the ways people were looking very critically at the current measurements was their ability to actually be used by practitioners. How much training do you have to have to use them reliably? How complicated are they? And how informative are they to someone on the ground? The generalized conclusion was that there really are not many tools that look really good to the research community that are going to be very useful to the practice community.” The search continues.

If a child’s parents can’t read, however, that child is going to have a hard time being prepared for school, with or without Head Start. Associate Branch Chief Peggy McCardle directs the Adult, Family, and Adolescent Literacy program that is just beginning to build an evidence base to address such problems. Partnering with the DOE and the National Institute for Literacy, her program last year provided $18.5 million over five years for six multi-site studies that are developing and testing instructional interventions at more than 80 sites in 16 states.

That agenda itself was developed in a pair of workshops that brought together researchers and educators to start constructing the translational bridge that will be essential once interventions are ready for real-world testing. The investigators, McCardle says, are “taking what we know from the basic reading research, and saying: ‘We know what works for kids; now how much of that will be applicable for adults and in what way?'”

The same workshop approach was used to develop a research agenda for adolescent literacy as well, she says. “The first workshop was predominantly for researchers, but we intentionally seeded it with practitioners. [From that workshop] we generated a major document and presented that at a second workshop of about 250 people, predominantly practitioners, but this time seeded with key researchers. We had them talk about it, mix it up. What were we missing in this research agenda? What did we need to consider in implementing it? When people design their studies, what do they need to look out for?”

McCardle is also inviting research into the best way to teach reading English to children whose home language is not English. “They are being immersed in English from the day they walk in the school door,” she explains. “We don’t know that that’s the best way to handle it. Can you teach them to read in both languages simultaneously? How much of their reading depends on their oral language skills? Those are questions we are looking at.”

The $30 million program, half of it funded by the DOE, will eventually study 5,400 children across 13 or 14 projects in eight states plus the District of Columbia and Puerto Rico. “It starts out with the development of instruments,” she says, “measures to look at their skills in various areas of language and reading. We are studying the kids, the teachers, the classroom context, instruction in the classroom.”

By the end of the five-year funding period, some of the investigators will be developing, implementing and testing interventions, McCardle says. “I expect them to compete to continue the research with this cohort.”

Abuse, neglect and violence place yet other obstacles in the path of many children’s achieving their educational potential. That research domain is directed by Margaret Feerick, another former teacher.

Feerick taught second-, third- and fourth-graders during two semesters in New York City public schools. She was “amazed that there were no paper towels and no soap in the bathrooms, basic things that these schools clearly didn’t have money for,” she recalls. “How do you teach kids to wash their hands after they go to the bathroom if there’s no soap and towels? That’s a basic thing, right?”

She also taught two years in an independent school, introducing Chaucer to gifted children from tough, low-income neighborhoods like Bedford-Stuyvesant and South Bronx. “I was interested in research as an undergraduate,” she says, “but when I was actually out there working with kids, I became even more interested. I wanted to do research that was relevant to these kids’ lives.”

For her dissertation at Cornell University, Feerick investigated the long-term effects of exposure to childhood abuse and neglect among low income women in New York City, comparing the exposures of drug users and non-users. That research brought her to the CDB. “I wanted to come to Washington to work on issues related to child abuse,” she says, “and NICHD offered me the job. They were looking for someone to come in and start programs in child neglect, abuse and violence.” She launched that initiative her first year, 1998. (The pre-existing Cognitive, Social, and Affective Development program – largely basic research – was added to her portfolio the following year.)

Neglect, abuse and violence research was, of necessity, largely translational almost from the start, at least in terms of who was posing the questions. The nation was clamoring for answers to school violence, particularly after the April 1999 massacre at Columbine High School in Denver.

“People were calling up and asking what do you know about this? Are there things we can do, interventions we can use to prevent it? We had funded a lot of risk-factor research,” Feerick recalls, “but we needed more information on interventions. The goal was to plan innovative research on interventions that would lead to larger scale studies of the interventions. Several of those that had three-year planning grants (when her program began) are now coming back with full scale tests of their plans.”

Applied research is now developing interventions to address child neglect as well, Feerick says, but moving them from controlled conditions into real world settings remains in the future. “We’re still not there yet. We’re still trying to get the research done that can inform practice down the road. The field is still developing.”

The same is also true for childhood exposure to violence, she says, whether domestic or on the scale of the September 11, 2001 terrorist attacks. “We don’t have a good research base on this. What happens to kids that are exposed to domestic violence? What kinds of intervention programs are useful? We’re looking at everything from shelter programs to in-school programs.

“And community violence, too. We had not been funding much research on children in war and terrorism. After 9/11, we received a lot of phone calls asking what do we do about these kids? What do we know about what we should be doing? We really didn’t know enough, I think, to inform practice. There’s just so much research we need on this. Not all kids are going to have a traumatic reaction. For some the effects may be subtle, and a lot of it will be mediated by family variables, parenting and things like that.”

For her program, as for many in the CDB, the translational research bridge begins at the planning stage. For example, Feerick’s program announcement inviting research on children exposed to violence was developed at a workshop last July.

“We did bring in people engaged on the front lines,” she says. “In most of the areas I’ve been working on the past couple years, it’s always important to bring in some of the people out there doing the work because they often have the best ideas about what they need to know. You don’t want to be doing research that isn’t going to be useful.”

“Do we have good research on the best methods for dealing with abused kids?” Feerick asks. “We don’t. So good research is needed to inform what’s going on. On the other hand, I can see how it must be frustrating for the people out there doing the work that they don’t see the immediate application of the research. So we have to take the next step: how do we get what we know from those trials out into the real world?

“But in the real world, the kids are going to get programs based on other considerations. There are so many other things going on. A child protective services worker is going into a home that’s filthy, that’s not a safe environment, the families have multiple problems, abuse may be only one of many problems – poverty, domestic and community violence. They are faced with realities that are very complex.

“It is important to bring together these two worlds. It’s important for the research to inform practice, and I think it is just as important for the people out there doing the work to talk to the researchers about what they think is important and what they need to know.”

Feerick is now planning a book based on that workshop, one that will explore the implications of research for practice at all levels – in clinics, schools and communities.

Two other learning-focused programs at the CDB are currently almost exclusively concerned with basic research but are also vital to future translational research as well.

One is directed by APS Member Dan Berch, who joined the branch last year to launch a formal, concerted program of research on cognition and learning in mathematics and science – in essence seeking to build in those fields a body of basic knowledge similar to what was developed in reading research over the past few decades. What he is looking for, Berch says, is “use-inspired” basic research, but in mathematics there’s not yet even a “clear-cut accepted definition of mathematical learning disability,” and research into scientific learning is in an even earlier stage of development.

The other program, directed by APS Fellow Lisa Freund, is mapping brain development. It’s an ambitious five-year initiative within the Developmental Psychobiology and Cognitive Neuroscience program, in collaboration with the NIMH and the National Institute of Neurological Disorders and Stroke. Investigators at six sites will use functional MRI to chart normal brain development in 500 to 600 children from birth to age 18.

Mapping brain development will have important clinical applications, Freund says. “If from this we do get reasonable brain growth curves, those could be used by clinicians looking at children with various developmental disabilities, or children with any question about neurological disorders, to see if there are any major abnormalities or deviancies going on. It won’t be diagnostically definitive, but it would be a nice clinical comparison that’s not been available to this point.”