The “science of reading” movement has shifted early reading practices across the country, with more than 40 states mandating that schools explicitly teach beginning readers how to decode words.
So why is it that so many older students still struggle with reading after they leave the early grades? And what can schools do about it?
These are the questions that have been driving Rebecca Kockler, the executive director of the Reading Reimagined program at the Advanced Education Research and Development Fund, a national nonprofit that conducts R&D on evidence-based approaches in K-12 education.
“We started down this journey because we saw so consistently this drop after 4th grade. We just were not able to see real change,” she said, referencing score declines in reading after that grade level. “We wanted to better understand what was happening there.”
The Reading Reimagined program funds teams of researchers to study how to support struggling readers in grades 3-8. In 2024, it published research with the testing organization ETS that confirmed the idea of a “decoding threshold”—a mastery of sounding of words that students needed to achieve in order to continue making progress in reading in upper elementary school and beyond.
But that threshold isn’t static—the decoding skills that students need grow and change as kids start to read more complex texts, Kockler and her colleagues argue in a new report from AERDF.
The report makes the case that all students could benefit from instruction that helps them decode multisyllabic words, the longer, more complicated words that often bear much of a text’s meaning in upper grades’ classrooms. This instruction is so central to reading comprehension, the report argues, that states should include advanced foundational literacy skills in their academic standards for grades 3-8.
Education Week spoke with Kockler about the report, what advanced decoding instruction would look like, and why she thinks it’s such a powerful lever to improve students’ reading abilities.
This conversation has been edited for clarity and brevity.
The report argues that schools stop teaching foundational skills too early, and that this underpins a lot of the literacy crisis for older readers. What does that mean, and why is that the case?
This is one of the important gaps, and one that has not been well-researched and well-talked about. There are a couple of other important things happening, related of course to vocabulary and background knowledge. We do not mean to suggest those things do not matter.
What we’re suggesting is that you learn to read and read to learn as early as pre-K and kindergarten, and both of those things continue at a minimum through middle school. Both pieces of that, all pieces of Scarborough’s Rope [a commonly used diagram to show how foundational skills and comprehension knit together into skilled reading] are important for a lot longer. We don’t intend to undervalue the components of reading comprehension, vocabulary, and background knowledge.
One of the things we found in digging into research that does exist, and working with researchers who have been thinking about this and doing some more of our own, is that there is this continued set of complex foundational reading skills that show up when words get more complicated.
We can feel like we taught all of the skills when kids are younger, but then when words get more complicated, around 4th, 5th, 6th grade—words get longer, word structures change—we stop teaching skills around how to break words apart and make sense of them. And the skills that worked when words were simpler aren’t the same skills that work when words are bigger and more complex. We have to keep teaching the new skills that show up.
How are the skills that go into decoding a short word like “cat” or “dog” different from the skills that would go into decoding a word like “photosynthesis”? And why is it that the skills that are taught early don’t necessarily transfer?
Students learn more basic decoding skills, which we put under the banner of the “science of reading,” when they’re younger, in kindergarten through 3rd grade: To put sounds to letters, how to blend sounds together, how to sound out words. Identifying that “c” makes the /c/ sound, and how to put those sounds together into “cat.”
Those same three letters show up in bigger words around 4th and 5th grade in the middle of a word, and those same three letters now make different sets of sounds because there’s other pressures in the English language on those sounds. So “cat” goes now into the middle of a word, and instead of making the sound “cat,” it makes the sound [that] shows up in the middle of “vacation” or “education.”
When we see kids who have more simple decoding skills, they’ll look at “education,” and they’ll pronounce it “ed-u-cat-ee-on,” saying cat in the middle.
But as words change, the English language is rooted in all of these different languages, and there’s different reasons why an “ion” sound changes how the word sounds, and that the “e” changes how the word sounds. So decoding starts to connect even more into vocabulary and morphology.
How do we differentiate, when students are struggling at the word level in these grades, whether it’s a decoding issue or whether it’s a vocabulary issue?
I’ll start with “education,” then I’ll get to “cathartic.” I cannot think of a middle schooler who doesn’t know the definition of education, and yet that’s a great example of a word where we see 4th and 5th graders struggle with fluency.
There’s quite a bit of reading where we see students struggling. If you really look at the words, they’re not such complex vocabulary that students don’t know what they mean. Cathartic is certainly a more complex word, and might signal that actually it is more [a problem of] vocabulary, background knowledge.
There are better assessments now that can examine … do students need morphology work? Do they need vocabulary work? Almost always they’re going to need a combination of the two, because morphology almost never sits separate from vocabulary, and vocabulary almost never sits separate from complex decoding and applied spelling.
Can research show us how common it is that students’ struggle is with decoding rather than other parts of reading comprehension that aren’t decoding-related?
We haven’t had a lot of good data on that historically. We’re currently running some studies with Stanford and ROAR [Rapid Online Assessment of Reading, a K-12 reading assessment developed at Stanford University] to better understand this.
There’s not a simple answer, because, again, it’s hard to disaggregate complex decoding from vocabulary, because they reinforce each other, and they support each other. We need to put those things together in tier one instruction for kids, for almost all kids, explicitly and systematically. We shouldn’t try as hard to separate them.
At least 40% of middle schoolers—we think probably more, but at least 40%—seem to be showing real struggle with complex decoding. Probably more like 10% of students are really struggling with single-syllable decoding, like early-stage kindergarten to 2nd grade decoding work.
What the decoding threshold research shows is this really interesting correlation that we haven’t always understood, that the better you get at decoding, the better you get at comprehension. It’s not just this light switch like, ‘Oh, now you can decode, now you know you automatically are going to start moving into phases of comprehension instruction.’
What would instruction that’s integrated in that way look like in an upper elementary classroom or a middle school classroom?
We funded research on a couple of projects; their materials are open source on our website. Those are great examples of what we call “word work,” this work to bring complex decoding and vocabulary together in a more meaningful progression.
Another way to do it would be looking at the vocabulary of the books you’re reading and identifying the ways that morphology or other components—etymology [the origins of words], some fluency—are playing out in the books that you’re teaching. That’s really the richest and best way to do it, because kids have immediate application. We want to get kids out of isolated skill progression and very quickly, having to apply it in the context of reading books.
We are really pushing states to consider just a small number of additional tier one standards through middle school. If they’re seen as tier one curriculum, companies and others will be required then to build them in naturally to the curricular materials, so that they are built in. It doesn’t have to be a huge, long external progression for all kids. It can be integrated into the books you’re reading. We also have to assess it, so we get more clarity with where kids may still be struggling with some of these skills.
That’s really the ideal setting—these start to just get more naturally included into tier one instruction. But some of the tools we have, you could do once a week or twice a week, working students through that progression and still [see] lots of improvement from those.
Recently, there’s been some conversation in the broader science of reading movement about the potential that schools are overteaching foundational skills. Reading researcher Mark Seidenberg, for example, has argued that while all students need to be taught phonics, students without reading disabilities might not need to learn every single phonics pattern they could encounter. Once they have the basics, they can pick things up through statistical learning.
I’m wondering if you think that there is a danger in recommending advanced decoding skills for everyone—might we be overteaching in some way?
Foundational skills, whether they’re early stage foundational skills in kindergarten and 1st grade or complex foundational skills in older grades, are essential. The research is incredibly clear, and they need to be built systematically. They need to be explicitly taught and explicitly practiced.
But the thing we see, and here I totally agree with Mark Seidenberg on this point, the research will show some students need five practices, and some kids need 200 practices to master a skill. This is where I think technology can play a really important role in learning to read. Technology can start to sift out more quickly who needs five practices, and can move a student on, and who needs 200 practices, and can give them applied practice where it makes sense.
We don’t stop students once they get something wrong one or two times. We do integrated skills right away, even if they got it wrong in the individual skill building. And then we do applied skills in text very fast, starting in kindergarten.
Students are moving faster through our foundational skill progressions, and they’re showing more mastery—including the bottom 50% of students that we work with—when we can move them more quickly through that progression, and we can check on mastery after they’ve had chances to get into integrated skills and applied skills.
In that sense, I totally agree. It’s taking us too long to teach the foundational sequence, and we have a lot of evidence that we can do it more efficiently for students, figure out who needs more practice where, and move kids forward where they don’t need it. That’s also true for older students with the foundational sequence, but I do not think students don’t need some explicit teaching of those skills.
And again, students can, should, and must still be in grade-level comprehension work, vocabulary work, background knowledge work. That should be where the bulk of time is spent in the classroom.