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.
More on Hanford: Phonics Reform and Literacy Levels
Recently, I posted commentary on Emily Hanford’s reporting and the critical response it received from some in the literacy community. I defended the major thrust of her work and called out criticisms I thought to be illogical, ill conceived, or ill intended — criticism more aimed at maintaining status quo than promoting literacy.
I admitted that my endorsement of that journalism was not without limits. I had concerns and said I may write about them in the future. The future has arrived.
I expressed two concerns, one substantive and one more stylistic. Let’s get the less important one out of the way first. Style.
In “Sold a Story” there a heavy focus on the financial gains of the creators of the curricula that were critiqued. I admit to being as titillated as the next guy about such juicy details, but I also am aware of its shortcoming of that approach. While I don’t know those authors well, I don’t doubt their seriousness of purpose or staunch beliefs. I agree with Ms. Hanford’s critiques of those programs and have no doubt the authors gladly accepted generous financial rewards.
But I don’t believe they “did it for the money” per se. I’ve offered this same defense against the same charges that were leveled against certain reading publishers during the early 2000s. Critics charged the only reason anyone would endorse phonics programs was to get rich. Hanford even went to the trouble of digging up such a quote from Lucy Calkins herself in 2002 — a pitiful example of “what goes around, comes around.”
In those days, I – and other National Reading Panel members — were accused of summarizing the research as we did “just for the money.” That we were neither paid for the work nor allowed to have any financial conflicts, while the critics were reaping financial gains from their criticism, was an irony missed by many.
Some will try to distinguish the events by concluding, the phonics authors are right, and the 3-cueing guys are wrong. I believe that to be the case, but it changes nothing. I think in both cases the authors have had strong reasons for publishing what they did, and in both cases, they have had strong reasons for continuing to do so: money is not just a direct benefit, it is an indication that your work has wide appeal to educators and that it must be fulfilling some instructional need. As I’ve noted before, many things work in reading — they just don’t work equally well. Cognitive psychologists have explained how human beings fool themselves, looking at the positive evidence and rationalizing the data we don’t want to accept.
My point: Emily Hanford did the profession (and, most importantly, the students) a service by identifying how the most popular reading programs were out of alignment with the best knowledge that we have about teaching and learning reading. That’s really all that matters. That authors and publishers are allowed to publish what they want and to profit from that publication is a side issue that muddies rather than clarifies.
My bigger concern with Ms. Hanford’s most recent reporting (episode 1 of “Sold a Story”) has to do with the implied connection between the big problem (unnecessarily low national literacy rates) and her solution (add explicit decoding instruction to the agenda and eschew unproven approaches like 3-cueing).
The deep dive into the ugly NAEP scores was both informative (Hanford’s documentary-making skills were on fine display), but they also left me with the implication that we are only succeeding with 65% of fourth grade readers due to the ubiquity of 3-cueing and the dearth of phonics. That seems to be saying that if we addressed those curricular gaffs, all our kids would be successful readers.
That promises too much.
I know this same criticism can be aptly leveled at other reporters, politicians, and academics as well. Perhaps even me. We foreground our claims with pornographic NAEP literacy statistics without ever divulging that our nostrums are about improvement or mitigation only.
Do I believe it would be productive to have well-prepared teachers delivering explicit phonics lessons in grades PreK-2 using well-designed programs for about 30 minutes per day?
I believe it because of the many instructional research experiments that have been conducted over a long period of time that have shown such instruction to provide learning benefits to children.
I believe it because of the descriptive and correlational research evidence from neurological and cognitive psychological studies that suggest the potential benefits of instruction that guides students to connect letters and spelling patterns to phonology.
I believe it because of what has happened to fourth grade NAEP scores in the past when there have been increases (and decreases) in explicit phonics instruction. When we have had major emphases on phonics, scores have risen, and they have fallen or remained stagnant when attention to decoding lapsed.
Given that, what’s the problem?
The problem is that those changes are likely to only produce marginal improvement. Here I’m using the term “marginal gains” the way I think economists do: to refer to small incremental improvements that when added together with other similar improvements could result in significant improvement.
Why do I suspect the gains will be real, but relatively limited?
One reason the gains are likely to be marginal is due to those positive research findings I noted. The effect sizes in those studies average out to about 0.40 and when you control for other variables that attenuates to about 0.20.
Many elementary reading tests are calibrated to produce a 1-standard deviation difference between grade levels. That means that the average first-grade and average second grade reading scores often differ by 1-standard deviation. If each year’s phonics instruction managed to accomplish the amount of benefit suggested by those effect sizes (about 20% of a standard deviation or about 2 months added gain over a school year), our kids would be doing about 1 semester better in reading by 4th grade. That’s an amount of gain that I dearly desire, but an amount of gain that still would leave large numbers and percentages of kids far behind.
Of course, obtaining gains in a small study in which the researcher can carefully monitor the delivery of instruction is much easier than doing so in a large urban school district or a widely dispersed rural one. Usually, attempts to implement research-proven interventions on a large scale, witness diminished results. It is unlikely that those estimated gains would be accomplished statewide or nationwide year after year.
Another reason for my skepticism has to do with the findings reported by the National Institute of Child Health and Human Development. They funded an extensive body of research on reading development and instruction which was strongly supportive of phonics. But it also reported that more than 50% of those struggling readers whose decoding ability was boosted to average levels continued to struggle with reading because of other limitations. More recent studies (such as those done by Rick Wagner and his colleagues) have identified plenty of kids with adequate decoding abilities who, nevertheless, struggle with reading comprehension. No reason to believe phonics or more phonics would help those students.
Still another reason for concern is that past efforts that have significantly improved fourth grade reading achievement on the NAEP (most notably from 1991-2006), but they have done so only marginally. Large enough gains to be both statistically and educationally significant, but still with large numbers and percentages of kids who don’t read well enough.
An example might help. Much has been made of the recent reading gains in Mississippi and these have been ascribed to the wide implementation of a phonics curriculum. My own analysis of these scores is that phonics contributed incrementally or marginally to Mississippi’s surprisingly high (for their economic level) reading scores. Reading didn’t only improve in Mississippi after implementing its phonics reforms, however. Incremental gains had been building over a 17-year-period. Phonics was only one in a long series of incremental improvements that when added together made for noticeably significant results.
Setting aside that observation for a moment, let’s attribute the entire 16-point NAEP gain that Mississippi has experienced during this century to the universal implementation of high-quality phonics instruction. After those very real improvements, we see that 35% of Mississippi kids are still struggling with reading. They’ve managed, despite their high poverty levels, to reach the national averages. That’s wonderful. But even with those remarkable gains a very large percentage of Mississippi kids are struggling with reading.
Another concern about the NAEP evidence: Even during those eras when phonics instruction and 4th grade reading performance rose together, they have not managed to have a big influence on NAEP 8th grade or 11th grade scores. One would think that 4 or 5 years after accomplishing those phonics gains, better readers would continue to display their early learning gains in middle and high school. That has not been the case.
The take aways?
By all means, please address those educational defects that Emily Hanford has Paul Revered for us. Primary grade kids should have high quality phonics instruction and that should provide precious gains in early reading achievement. More kids will succeed in learning to read, and the level of average performance should go up as well.
However, if what you seek is the solution to the low literacy attainment problem that “Sold a Story” started with, then you had better be prepared to do a better job with those other needs that research has also identified.
Our kids need high quality instruction in phonemic awareness, phonics, text reading fluency, spelling, reading comprehension (both in terms of comprehension strategies and written language skills – vocabulary, syntax, cohesion, text/discourse structure), and writing. Indeed, our kids need to learn to read challenging literature and informational texts from the different disciplines in sophisticated ways, and they need to get used to using text for building extensive stores of knowledge about their social and natural worlds.
That prescription is for a PreK-12 response, not a primary grade one. Our goal shouldn’t be better fourth grade readers, but more literate 12th grade readers. Having more 4th graders reaching proficiency levels only matters if we’re willing to build quality on quality to make sure they maintain and advance those early successes.
It is okay for our reach to exceed our grasp. But we’ll do best if everyone fully understands both what it is that we are reaching for and what it will really take to accomplish it.
(And to those of Hanford’s critics who are now chortling, “See, that’s what we were saying,” I would say— “No, what you were doing was fanning the flames of a reading war instead of embracing every initiative that has a reasonable chance of contributing to the accomplishment of universal literacy.”)
Comment from Sheila
I wish that everyone that heard Emily Hanford's podcast would also read this. Thank you, Tim. Hats off and peace!
Perhaps part of the problem has been that while teaching phonics was "right" from the start, that did not mean the way we've been teaching phonics was (or is) optimal. There is now a phonics program available that promises to more than triple the effect sizes that you cite for beginning readers- not marginal gains at all.
The experiment is yet to be run that if most (upwards of 90%) students become master decoders by third grade by using only optimal programs and methods instead of "business as usual" phonics instruction, does that translate into significant gains in the 4th and 8th grade NAEP scores? I'd like to run that experiment. (Not claiming that better linguistic comprehension instruction is not needed — it is!)
Reply from Tim Shanahan
Such experiments would certainly show us the limitations of what we promote. However, we are working with real kids and the best thing would be comprehensive instructional programs that address all the key elements that kids must master. Phonics is certainly one of them, but there are several others.
Comment from Bill
Beside quality and phonics-based curriculum, we need to consider the elements of pedagogy and assessment. Only 1/2 of the colleges of education instruct research-based instruction resulting in woefully unprepared teachers. District PD triage will never make up for this huge gap. Additionally, school districts demonstrate weak ability in planning and almost 0 ability to implement well and monitor and hold accountability. Inevitably school districts will throw effective research-based program on the ash heap of failed educational fads and initiatives. Sometimes not because of their quality but because of their failure to implement, monitor, and hold accountable.
Reply from Tim Shanahan
Thanks. I don't disagree that quality of instruction, but it tends to matter a bit less than the other legs of our 3-legged stool — the most important ones to get right first are teaching a sufficient amount of those things that kids need to learn to be successful readers. Then making sure the teaching is good too, improves effects. We can teach well and fail to teach the right things. However, if we are teaching the right things poorly, some kids can still figure out what they need to from the information.
Comment from Karen
I'm just wondering about history — has there ever been a time when reading "scores" were good? Are we working to bring readers to a level we had in the past? Or are we working towards something we have not seen before in history? Do other countries' readers do better?
Reply from Tim Shanahan
What would be good scores is related to how much literacy is needed in a society. Given that globalization, digitalizing, and increases in legal protections have over 50 years increased the amount of literacy that people need. What you would like to see is clear growth in literacy. We saw reading scores climb in the 1970s and early 1980s, but then schools backed off on certain teaching practices (like explicit teaching) and scores went down lower than they had been in the 60s and 70s. Then states and federal governments focused heavily on following research, teaching explicitly, monitoring learning, etc. and in the 1990s and early 2000s scores climbed. Then governments backed off on those policies and scores languished for a decade and then started to fall again in the 2018 and more recently. One could argue scores were the highest ever in 2006, but we're no longer at those levels. Other countries that we used to lead now match our performance, and some even are doing better. We definitely have room to grow and there is reason to believe we can do better (even if it means finding new ways of accomplishing that).
Comment from Wendy
Can the education system ever out-teach literacy disadvantage and academic struggle? We spend a lot of time measuring our impact only to see that we make little difference; put only a little dint in the problem, despite everything we try. I must admit, I feel a bit defeated after reading your article, though it is an excellent analysis of the state of play, and its clarity and rationality superb. This teaching game is not for the feint-hearted!
Reply from Tim Shanahan
As Mississippi has shown, it is possible to overcome many of the effects of disadvantage when it comes to literacy. It is hard work and it is worth doing. Unfortunately, not many policymakers are interested in providing the continued support and discipline needed for the continued incremental grind needed for real success. But it is possible.