Assessing and Supporting Word-level Reading

Photo by Pixabay on Pexels.com

In this post, I’ll continue pulling together my notes on what I’m learning about reading. Thank you in advance for reading, sharing your thinking, and helping me to connect with a broader community committed to improving literacy instruction.

I want to first draw everything back to the Simple View of Reading as a friendly reminder that reading is big.

Word-level Recognition X Language Comprehension = Reading Comprehension

I’ve only been focused on the word-level piece, because that’s the part that was so new to my own understanding. But the language piece is HUGE!

Anyway, in this post, I’ll keep with the word-level recognition side of things and focus on what assessments and programs we might be able to use to tackle just that one side of things.

It’s one thing to have a clear theory and a model; it’s another thing to act upon it. This is where the real debates begin, because at some point, the rubber needs to hit the road:

  1. What will we use to screen and diagnose code-based and meaning-based literacy skills?
  2. What will we do in our core instruction to prevent reading difficulties?
  3. What will we do to intervene when core instruction is insufficient?

This means a school needs to have an RTI model of some kind, which is a level of sophistication, unfortunately, many schools struggle with. There’s a lot more that Seidenberg, Kilpatrick, and I have to say on this topic, but in this post, I’ll maintain a narrower focus. I’d like to dig further into the RTI piece of it in a future post (I have some criticism of the model, though I’m rethinking it in light of some of my new understandings).

Every School Needs a Universal Screener

One of my favorite things about the Advanced Literacy framework that both NY state and NYC have adopted is that it promotes the need to go far beyond the data provided by a state assessment. We need universal literacy screeners–a short test that can help immediately identify kids who are behind in either code or meaning-based ability, from which we can then drill into further with diagnostics.

Because the available tests are not always ideally suited for assessing the various components of reading, the best reading assessment tool is the evaluator’s knowledge of research on reading acquisition and reading difficulties.

David Kilpatrick in Essentials of Assessing, Preventing, and Overcoming Reading Difficulties

The problem is that there are no perfect assessments. As Kilpatrick notes, “Because the available tests are not always ideally suited for assessing the various components of reading, the best reading assessment tool is the evaluator’s knowledge of research on reading acquisition and reading difficulties.

And there are so many levels to reading assessment that it’s almost fractal in nature. You’d need a significant battery of subskill assessments to get a full and accurate picture of any individual child’s reading ability.

Another problem is that time is limited, and there are already a substantial number of tests that students are forced to take. Some are in-house, some are district mandated, some are used to evaluate teachers.

Ultimately, a school must make sense of them as best they can. This is where The Simple View of Reading really comes in handy. Different assessments provide you with different kinds of information.

There’s thankfully a lot of great resources in determining what screeners your will use. The Gaab Lab at Boston’s Children Hospital has an extensive compilation of screeners here.

Assessments of Phonological Awareness

Kilpatrick recommends using both the PAST and the C-TOPP2 as a further diagnostic after a universal screener. But I recommend using only the PAST because it’s free vs. $347 for the C-TOPP2 kit. Why is the PAST test free? Because Kilpatrick developed it and publishes it for free here: https://www.thepasttest.com/ He provides instructions on the site as well. Pretty darn cool.

I’ve begun piloting the use of the PAST at a few schools I support, and it’s been pretty eye-opening to see just how much need there is with phonological awareness in the students I’ve tested. I’ve administered it to an 8th grade self-contained class, and all of the students had phonological deficits — some at the most basic of levels. One student struggled to say the word “fantastic.” He couldn’t get that last syllable, even when I slowed it down and repeated it. 8th grade.

This has only gave me a greater sense of urgency in figuring this out.

The other thing I noticed is that the person administering the PAST really has to know their phonemes. It’s surprisingly hard to do well. In order to get an accurate gauge of student ability, you have to deliver the instructions swiftly and precisely. If you slow down or stumble when saying, “Say guide. Now say guide . . . but instead of . . . /g/ say /r/,” you can easily tax the student’s working memory, and they forget which word they are supposed to use while paying attention to the phonemes you’re saying.

When training teachers to administer the PAST, I first have to ensure they can pronounce the phonemes accurately, and then deliver the tasks with swift pacing. This takes practice!

So my advice is to practice delivering the PAST with someone else, multiple times, before you administer to a student.

Check out my new Resources page for a couple of trackers you can use once you’ve administered the PAST.

Assessments of Phonics Skills

Kilpatrick recommends using the TOWRE-2 Phonemic Decoding Efficiency subtest and KTEA-3 Nonsense Word Decoding subtests. The problem is that these are subkits of a larger kit, and the kits for each are expensive. If you’ve got a school psychologist in your building who uses these and can lend you a hand, that’s great.

Protip: “Nonsense word tasks appear to be the best way to evaluate a student’s phonics skills. In essence, all unfamiliar words a student encounters are functionally ‘nonsense’ words until they are correctly identified. . . . Timed nonsense word reading, such as in the TOWRE-2 and the KTEA-3, is arguably a better assessment of a student’s cipher skills than the traditional, untimed nonsense word reading tasks.  . . . It is recommended that any timed nonsense word reading task be administered after an untimed task, and not before.”

Though Kilpatrick recommends these normed assessments, he does acknowledge that they “do not provide much information about the specifics of what elements of phonics skills are weak or missing. By contrast, there are many criterion-based assessments of very specific elements of phonic knowledge. Some are commercially available assessments and others are free online. These criterion-referenced assessments will index the particular letter-sound combinations that the student knows, such as the various letters, blends, digraphs, and diphthongs, which can aid instructional planning.”

So my (admittedly amateur) advice? Normed assessments are great if you can afford them. But you can use something like the CORE Phonics Survey, the DIBELS Nonsense Word Fluency tasks, or Ruth Miskin Nonsense Word Test (all available for free). I also just got an OG (Orton-Gillingham) phonic screen from a colleague, and it was really short. Please let me know what else you might recommend.

I’ll stop here.

There’s much more to talk about with assessments for word-level reading, but I’ll stop here. Even out of these two, phonemic awareness and phonics, I’ve elected to only focus on one — phonemic awareness. Why? Because if Kilpatrick and Seidenberg are right, this is the core area of deficit that causes word-level reading gaps. And because I’m just trying this out and seeing what kind of practices and systems I can support a school in developing that are sustainable and scalable, and you have to start somewhere.

Even just administering the PAST is a much bigger endeavor than it seems at first glance. You need to train and practice with it. Then you need to test each student individually, in a space where you have enough quiet to be heard.

And then you need to figure out how to provide effective intervention in a consistent and effective way. From the first set of data I just collected last week, I can see this will be more complicated than I thought. Each student is at different levels of phonemic awareness, so how can we group them strategically while still addressing each student’s need?

Help! If you’ve used Kilpatrick’s Equipped for Reading Success program, especially with older students, any advice is much appreciated.

Applying What I’m Learning About How Kids Learn to Read

It was pretty cool to see my last post catch 🔥 and link me in to a vibrant and smart community of educators committed to the science of reading.

To review, in that post I laid out what I’d begun learning after realizing I knew absolutely nothing about learning to read:

Summary of critical points on word-level reading

The Simple View of Reading provides us with a clear and research-based model of reading comprehension

  • This doesn’t mean it’s completely definitive–no model is. But it does give us a useful map for aligning and targeting our assessments and instruction

Anyone who hears and speaks can be taught to decode words in print

  • IQ is not the basis for the ability to decode
  • Nor is it ever too late to address decoding issues

Units of sound (phonemes -> phonology) are the basis of written language (graphemes -> orthography)

  • Most word-level reading challenges are related to issues with hearing and speaking the sounds of the letters in words

We acquire new words as we read via a process called orthographic mapping

  • It is the phonological part of our brain that anchors the written word in our memory, not our visual memory
  • We learn the vast majority of words (after we have decoded them) by rapidly and unconsciously recognizing the sequence of the sounds of the letters in a word — even when they are irregular

The root cause of most struggles in word-level reading is a lack of proficiency with advanced phonemic skills

  • Students require fluency with deleting, substituting, and reversing phonemes to acquire a large stock of sight vocabulary

Since Then

Since writing that post, it’s felt like a whirlwind of learning. In the NYCDOE, I learned that there are K-2 supports in many elementary schools called Universal Literacy coaches, and they are trained in the science of reading. I spoke with a few and saw how they are attempting to bridge the various programs and curricula schools use to the science. I read Robert Pondiscio’s superb book on Success Academy, How the Other Half Learns, and struggled to square how SA consistently achieves the highest reading proficiency rates in NY state, while applying some reading approaches not fully aligned to the science. (More on that in another post; there’s a lot to dig into from that book, and I’d like to do it justice.)

I then went to a training on Equipped for Reading Success with David Kilpatrick, and got to ask him directly about the distinction between statistical learning and orthographic mapping. He views them as different processes — orthographic mapping refers specifically to the mapping of individual phonemes, and it’s far more quickly acquired (1-4 exposures), as compared to statistical learning, which is a more global pattern recognition process that requires far more exposures. He had a nifty little chart he pulled up to explain the distinctions. Either way, however, I found Marnie Ginsberg’s explanation in a comment on my last post to be a pretty good way to think of it, though with the key addition being that while proficient readers can rapidly do all of this on their own, we need to explicitly train and teach the skills required for orthographic mapping (a chart that outlines those skills below).

A graphic from Equipped for Reading Success that should be widely known in every school.

It can be hard to gain clarity on anything in the world of education, but most especially when it comes to reading. So even as I take one step forward, I often take two steps back further steeped in doubt. Yet I’ve decided to commit to Kilpatrick’s manual as my North Star for the next quarter.

The Knowledge

I’m still moving through the Equipped manual a little each day on my commute, marking it up and imbibing what I’ve taken to calling “the Knowledge” in my annotations, an allusion to the famed test for London cab drivers. The Knowledge, in this case, being terms like digraphs, blends, diphthongs, onset, and rime.

Terms like these, much like grammatical terminology, can seem unnecessarily technical and unessential to good teaching. Yet imagine a world in which it was required for teachers to learn and be assessed on the knowledge behind the terms of word-level reading! I never understood– nor was exposed to–what “onset-rime” means until I read Kilpatrick’s manual. Yet once I grasped it, it served as a threshold concept for understanding phonological awareness.

Here’s the passage from Equipped for Reading Success that expanded my mind and made me aware of a key distinction between the syllable level and onset-rime level of phonological awareness:

“The onset-rime level of phonological awareness goes beyond the syllable level because the child has to break apart the syllable. . . . Onsets and rimes can only be understood within the syllable. Not every syllable has an onset, but every syllable has a rime. This is because every syllable has a vowel.”

–David Kilpatrick, “Equipped for Reading Success” pgs. 20-21

Remember how in my last post I had the big realization that phonemes are an abstraction from our everyday experience of spoken language as a stream of sound? The onset-rime level of sound awareness is one further abstraction from hearing syllable level sounds. There are gradations of abstraction on the road to distinguishing those individual phonemes, and that progression moves from syllable level (“baseball” = 2 claps), to onset-rime level (“baseball” = 4 claps (“b” is onset, “ase” is rime, “b” is next onset, “all” is final rime), to phoneme level (“baseball” is 6 claps (/b/, /A/, /s/, /b/, /a/, /l/).

I’ve begun playing some of the “word games” in Kilpatrick’s manual with my two and a half year old son to cultivate phonemic awareness, and I’ve noticed he can’t yet isolate the second part of a two syllable word. He can identify the first part, however. Which is of absolutely no concern to me, given his age, but I found it revealing of an even more fundamental progression in terms of working memory and the awareness that we can break up multisyllabic words into smaller parts.

When it comes to foundational reading skill knowledge like this, it’s always been something I’ve wished I’d known, but didn’t consider it essential, because the expectation was that I focus on grade-level texts and content. And yet I had students reading far below grade-level. One would think that this would have compelled me to learn it at that point–and I did try, I went through some of the files from my first years of teaching, and I found a whole set of phonics related stuff I’d amassed–but the reality is that it was something else on top of many other things I needed to know and do, and I put my primary focus on grade-level texts and skills. Not a bad focus, of course, but I look back on my many students who were struggling with decoding words, and I feel like I have failed them. I have failed them.

Teaching is a hard job. But so is nursing, and I’m watching my wife as she goes through a nursing program and struggles to acquire a vast body of knowledge that must be applied on a daily basis in a clinical setting. Nurses have to acquire this knowledge and be able to apply it, their jobs demand it. People’s lives are literally on the line. And yet, when it comes to teachers, our society seems to be perfectly fine to let them off the hook.

In How the Other Half Learns, Pondiscio has an especially wry zinger (in a book full of them) in Chapter 1 when he states, “Teaching is the easiest job in the world to do badly. . . But it’s the hardest job to do well.”

We are graduating too many students who are functionally illiterate. We all need to step up our game.

My Theory of Action

My working hypothesis, based on Kilpatrick: many of the struggling readers in the schools I support are struggling with a core phonological deficit. Therefore, if I administer the PAST and identify where a student’s phonemic awareness level is (and train teachers to do so), and support targeted daily instruction in phonemic awareness until proficiency is attained, then those students’ reading levels will improve.

I’ve brought the PAST, a short phonemic awareness assessment from Equipped for Reading Success, to a few of the middle schools I work with, and have begun pilots with self-contained classrooms and students. I just administered the PAST to my 1st student last Wednesday. We selected him because we knew he was struggling with reading. But it still shocked me with just how basic his phonemic awareness level was. He was at nearly the lowest level, the syllable level, a pre – mid kindergarten level.

Let me frame the wider context of what we’re up against: in that school, roughly 40-50% of students across the 6-8th grades are identified as struggling with decoding, according to an iReady diagnostic. Of that ~50%, how many are struggling with a phonological deficit? I’d like to find out. And help to do something about it.

Finding a way to tackle something that massive, while continuing to ensure that core instruction demands grade-level expectations, is a tough challenge. Because let it be known that I am in no way suggesting that kids struggling with word-level reading should no longer be exposed to grade-level texts and content. What I am suggesting is that it is incumbent on teachers at any level (and schools) to be knowledgeable enough of foundational skills and grade-level content and skills to scale their instruction accordingly. And yes, this is a heavy lift indeed. There’s never enough time in the day.

Yet I’ve found Kilpatrick’s materials promising in this regard, because some of the phonemic awareness activities are “1 minute” practice sessions. Every single minute we have with a student is precious time, all too easily squandered.

I recognize there’s many other aspects to this, such as administering a phonics screen or oral fluency task and pairing students with different programs depending on the need. But I’ve got to start somewhere. I’m going to start small to see if my hypothesis is verified and if I can help to enact instruction that will target those needs. This is where the rubber hits the road.

I may fail. This whole thing is, ironically enough, a pet project of mine. It is no official aspect of my duties and role in the schools I support. And I take on too many side projects as it is. I’ve got a book I’m supposed to be writing, by the way, but can no longer find the time for, let alone post on this blog. But I have a hard time thinking of anything more important than getting this right. So I’m saying this publicly so the network I’ve begun connecting to can help support me, so I can better help support the students and teachers I touch each day.

If you are on a similar journey, please connect with me here or on Twitter @mandercorn and let’s work through this together. There’s a wealth of knowledge out there, we just have to each individually connect the dots.

Thank you in advance, and thank you for reading. In solidarity.