Blogs About Reading
Shanahan on Literacy
Literacy expert Timothy Shanahan shares best practices for teaching reading and writing. Dr. Shanahan is an internationally recognized professor of urban education and reading researcher who has extensive experience with children in inner-city schools and children with special needs. All posts are reprinted with permission from Shanahan on Literacy.
I'm a Terrific Reading Teacher, Why Should I Follow the Research?
What does it mean that something has research support? I’ve been a teacher for years and I’ve taught hundreds of children to read. Now I’m being told that in our district we are expected to teach in some new way that has research behind it. I like how I teach reading and I don’t want to change. Why should I?
I suspect that there are a lot of teachers who agree with you. Someone like me claims that a particular approach is essential, but they see learning proceeding well without this supposedly indispensable element.
Why trust some researcher who doesn’t even know your kids, when you can trust your own eyes?
First, it’s important to know what we mean when we say that research shows that a teaching approach “works.” It does NOT mean this approach supports learning and nothing else does. It means that when such comparisons have taken place, this particular approach showed marginal benefits.
That there is some improvement at the margin doesn’t mean that all the kids in the experimental group prospered and that all the kids in the comparison group languished. It simply means that the average level of performance differed for the two groups.
You might have noticed that I used the term comparison group rather than control group. We don’t usually have “control groups” in reading studies as it would be unethical to withhold reading instruction from anyone. Any instructional approach will likely increase learning a bit, so learning will always be evident in both groups.
In this case, the kids getting phonics outperformed the kids who didn’t. Maybe many of the phonics-taught kids did a bit better than the comparison kids. Or, maybe some kids in each group failed to learn to read, but there were fewer of these outright failures when phonics was part of the instruction.
In any event, the outcome of these studies isn’t that the phonics kids learn, and the non-phonics kids don’t. It isn’t black and white like that, just whiter shades of gray. Essentially, following the research means trying to alter the students’ probabilities of success.
Second, in considering whether you should follow the research, it helps if you understand the concept of “opportunity cost.”
Let’s say you have been teaching reading for 10 years, and your kids are learning, their test scores are pretty similar to those obtained by your colleagues, the parents have seemed happy with your services, and your principals have given you positive evaluations.
Obviously, what you are doing is working. If it ain’t broke, why fix it?
That sounds so good, but it ignores what I referred to as opportunity cost. That you have been teaching reading as you have deprived you of the opportunity to see what might have happened if you’d taught reading in some other way.
That’s where research comes in. Researchers will try to arrange their studies in such a way that it becomes possible to evaluate how your business as usual approach does against some other approach. And, that’s where we might see that as well as your kids have been doing, they could have done better if they had been taught a bit differently.
In other words, your self-satisfaction with the way you are teaching may not be justified, at least when scrutinized against alternatives. No teacher can ever know what her kids are missing from staying with the current approach that seems so satisfying.
Third, the kind of research I’ve been alluding to focuses on particular outcomes, most often reading achievement, but other outcomes, too (e.g., reading interest, motivation). Teachers sometimes may sound like “science deniers” because they don’t share a commitment to the goals of the research.
The researcher worries about how many words correct per minute the kids can read, whether the kids can decode a list of nonsense words, or how many multiple-choice questions were answered on a standardized comprehension test. The teacher may have other criteria in mind.
I myself come across this one frequently. I’ll write about research proving that Method A obtains better learning results than Method B, and I’m inundated with missives from Method B advocates who are certain that if I visited their classroom, I’d change my mind.
These teachers aren’t idiots, they just have different goals than the ones espoused by the researchers.
I’m trying to get kids to the highest literacy levels possible. The teachers may want that, too, but they may value a feel-good classroom environment even more. That’s why they are so sure that if I saw how great their classroom is, then I’d prefer their approach to the purportedly more effective one.
Teachers often tell me that they need to do it their way, because they are teaching “love of reading.” That their state lacks a “love of reading” standard, or that their approach hasn’t been found to foster any especial fondness for reading, or that love of reading depends heavily on how well the kids can read doesn’t seem to faze them.
We can pile research study upon research study at these teachers’ feet, but no matter how high the stack gets they’ll never be persuaded since it ignores their personal goals.
Finally, educators usually know very little about research — its methods, its reasoning, its ethics. Even at Research I universities, Colleges of Education provide a dearth of research training for teachers and principals. It is hard to trust in something that you don’t understand.
Teachers often confide their sense that “research can prove anything,” both revealing their deep suspicions of the unreliability of research findings and the lack of trustworthiness of researchers.
Of course, researchers themselves are quite concerned about those reliability problems. That’s why, for instance, most of us don’t make recommendations for practice based upon single studies. We’ve learned to look for an average effect across bodies of research to ensure a scope and consistency of findings.
If a study says the XYZ Reading Program gets great results, that’s one thing; but if 38 or 51 studies say it does, then I’m on board. That’s the reason why reports like those of the National Reading Panel, the National Early Literacy Panel, or John Hattie’s compendium of meta-analyses get so much play… one more study is unlikely to disrupt their results since these results are already based on so many data. It’s that consistency that we trust in.
Another benefit of that meta-analytic approach to thinking about research findings is that it allows us to know not only that an approach worked — that is, had a positive effect on learning — but it can allow us to see who the approach worked with or under what circumstances it has worked.
The “it worked or didn’t” school of thought apparent in many school districts tends to obscure the level of detail teachers need to apply the research results successfully — and that detail, frankly, would give teachers greater confidence that the research was worth following. If, in the research studies, 30 minutes a day of fluency instruction improved achievement, then requiring fluency instruction, but with no specific guidance as to dosage or text type is likely to fail.
The teacher who has kids read aloud one easy page of text for fluency practice is honoring the research finding (fluency instruction is effective) but is not following the particular instructional practices that led to that finding.
The benefits of following the research should be marginally improved reading achievement. Research is the only tool we have that allows us to determine the kinds of teaching most likely to advantage our students’ learning; commonsense and past experience are useless before such questions. It is easy enough for a teacher to wave off higher achievement as a laudable goal — "these kids seem to read well to me” — but technology and changes in how we work and interact socially demand higher levels of literacy than in the past if our students are to fully participate in the benefits of our society. More knowledge about how research is done and how it is evaluated would go a long way towards helping teachers trust that research can really help them to do better.
I hope you’ll be able to successfully make the changes your district is asking for. If done well, it could lead to better achievement for your boys and girls.