Should We Teach with Decodable Text?

And, truth be told, this collection of studies is a bit of a mess; with little evident agreement as to what decodable text is, what it should be compared with, and what learning outcomes should be expected.
Over 20 years ago I was doing a masters level paper on reading and did a lit review on sustained silent reading and found the same the research was so poor you couldn't make any strong conclusions. My own theory was that where it was well implemented with a teacher who modelled and knew and loved children's lit it was great for 80% of the class and a waste of learning time for the 20% of children who low motivation to read and low ability - they usually go together.
Would love a Tim Shanahan on the topic

There was an unanswered question from the previous post regarding students with dyslexia that I would be very interested in. Any different benefits or detriments for this population regarding decodable text?
The majority of my career has been in first grade but recently I am working with K-5 students needing intensive intervention, many with a dyslexia diagnosis. Our district has adopted a multi sensory approach program which routinely includes decodables. I like them enough but question spending my limited time hunting and highlighting words with the phonics pattern before reading them. Anything in the research supporting this practice?

We are remediators,we see students that read the first syllable and guess the rest.
We see students from grade 1 to high school.
The students are anxious.
They are overwhelmed.
When we test them,they show difficulties with grapheme phoneme gaps and hardship blending sounds.

Without a structured literacy intervention and some decodable text they would remain illiterate.

I have read your stuff for decades.
You made me feel guilty :)
What we do now is we get them out of those decodables after we teach syllables because they need to know how to handle multisyllabic words and stop guessing.
That`s around 35 hours of remediation because naturally,the kids are vocabulary impoverished due to their difficulties, we need to hit heavy on other stuff to get them to where they need to be.

I see the problem with decodables that is now overkill.The kids are kept in them for too long,there`s no question!
'You gotta know when to fold em"
One of our interventionists showed me the pictures of books the school sent home for a grade 4 student. All CVC`S.
After only 3 months of intervention, we put him into real books.
Revolutions in systems are difficult.

It would make sense that the mix of decodables and non-decodables showed the highest results. Simple view of reading says decoding plus language is reading. Not decoding alone.
Decodables make good practice of skills. The application of those skills in non-controlled text is the goal.
I am curious how reading is measured in these studies. If the students haven't been taught the phonics patterns it is no surprise they cannot read them. Much like testing a student on division if they are only taught addition. This doesn't mean we don't teach addition. I would caution extrapolating this to mean that decodables are not effective.

It is very frustrating to me. Three years ago I literally had an administrator screaming at me that I had to use pure decodables only with no pictures for my primary students. I shut my door and used decodables and regular leveled texts with my students. I ended up making my own decodables because none were really provided for a time, and the ones that were later provided were terrible. Why is our educational system vulnerable to these wild mood swings where common sense leaves the building?

I can see that you may may not think that a picture is worth a thousand words!

Do you think that Artificial Intelligence will play a role in being able to adapt text that a child read to recognize the degree to which they benefit from decodable text and progressing to more complex text in their reading regimen?

Patricia,
When using early predictable (patterned) text, the teacher is focusing on left to right reading, return sweep, fluency, vocabulary development, and reading comprehension. If the book is Fruit Salad, each page will name a fruit. ‘In goes the…’ children learn the names of fruit, which fruit are peeled or not, and can talk about other fruit the author did not include. In writing about reading, they can sound out words to spell phonetically, practicing phonological awareness. They can consider how writers use the title to tell what they want to write about.
These are all valuable items and add to reading development. That is why predictables are used in combination with decodable text. Early decodables often contain pictures that help children build meaning and comprehend text. Unfortunately, many decodables are written so that language comprehension is more difficult. Beginning readers need text that support decoding and language comprehension and opportunities to write about reading (encoding comprehension.)

Bill--
I think AI already is being used to do such tasks. As AI gains sophistication and is trained up on certain language variables, I think we will see texts that allow us to expose kids not only to particular sound patterns, but to specified lexical, syntactic, and coherence patterns. That will get us more to what i've been arguing for during the past decade-- text centered comprehension instruction.

tim

Barbara--

On the studies of decodables, no one has yet shown any advantage to them in terms of reading achievement -- the kids with the non-decodables or less decodables repeatedly have done as well in learning to read as the kids who work with decodables.

tim

John--
You are right about the collection of studies, but even the best done of them always has concluded that there are no differences in reading achievement. The best it has done was that with 12 hours of work with decodables the kids were more likely to try to decode words (but even in that, no learning differences). Personally, I have no problem with decodables as part of the phonics instruction -- but that would be their sole use in my classrooms.

tim

This totally makes sense. I’ve noticed that kids who only get decodables in their hands are more rigid and don’t develop set for variability like the kiddos who have access to less controlled texts. Kids become good problem-solvers when they work a text that is high-quality and interesting to them - with motivation and support, they shouldn’t be limited!

I concur with your article. I have taught year1-3 chidleren literacy for 43 years in NZ. I have been teaching using Phonics + decodable books for 2 and a half years. I have found that the children transfer their learned phonics knowledge very easily, and that after the first few weeks of using decodables, read just as well using the NZ PM levelled readers, but they read with better phrasing and fluency. My children who struggle, with the longer decodable texts, find the PM's a relief, as they are not having to decode so many difficult words on a page, that is often written with contrived text and words the child has never used in their life, and consequently aremore difficult to understand, especially ESOL children. I use a variety of decodable books: Sunshine, Nelson PM, Beanstalk books- great photos. I use LLLL books after a few months as children are often confused with the cursive font. Thank you for a balanced view in your article. I will be sharing it.

Ha, about 5 years ago I told all my teachers friends in districts far and wide...”If you see folks getting ride of leveled readers in their classrooms, grab them and use them! I have just seen that baby getting thrown out with the water too many times....
I’ll wait to read that better research on decodable/non decodable readers but in the meantime my experience has taught me that children learning to read, need to read daily. And they need to read both decodable books and non decodable books that increase in difficulty. I taught children with dyslexia to read for 30 years and whether we were using Phonics for Reading, SIPPS or an Orton Gillingham program, we made time to read both everyday. The decodable books are implemented within the scope and sequence of the program you are using, allowing for more orthographic mapping of the newly learned sound-symbol element. Time is made everyday for students to read leveled choice books and easy readers within a structured routine, along with challenging texts to try and read alouds with the class or group. Providing this variety of reading activities allows teachers to develop language, content knowledge and opportunities to develop that all important skill of written expression, all part of literacy.

Hi Tim,
Very interesting post. Since the research points to the benefits of using non-decodable texts along with decodables, where do we find these texts?

Are they predictable readers? Are they levelled books? When I try to google such texts I cannot actually find them. Perhaps you don’t want to endorse particular publishers but I’m at a loss to find the texts you are referring to.

Also, did you know that schools in England will fail their educational inspections (OFSTED) if they start using non-decodables along with decodables!

As usual, the answer is not "either/ or," it's "both/ and."

I'm an SLDS, and after 25 years of teaching struggling readers, I've found myself using decodable readers less and less. For one thing, decodable texts sometimes become ridiculous in their efforts only to use a prescribed set of words. What's more, I've found that parents often don't really understand the purpose of decodable books -- they'll play them audio versions, or read them to them as bedtime stories!

When it comes to teaching pure decoding, I've found that using nonsense words is probably the most efficient method, because they teach students to think, rather than to shut down, when they encounter unfamiliar words. You just have to use a lot of extrinsic motivation, because they're hard! And it can be hard to stick with it, because very few programs integrate them into the curriculum.

When learning music, we practise with the simple over and over to build mastery and automaticity before moving to complex pieces. While doing this, we listen to beautiful music and watch others play, knowing this is our end goal. This makes logical sense to everyone - we can’t play Mozart when we are just learning to play Mary had a Little Lamb.

I see the alphabetic code the same way. We build our knowledge of the initial code, we build automaticity and mastery as we then move to more complex texts.

Children can not access regular texts with multisyllabic words when they still can’t read ‘cat’. Do we have these books in our classroom? YES. Do we let them borrow them from the library? YES. Do we give them simpler, decodable texts, to build automaticity and mastery? YES.

I hope someone studies this soon. I would hate for the debate to take away the important opportunity decodable books give for independent practise of skills. We know that many whole language advocates take any opportunity to dismiss the importance of learning to decode.

Patricia -
We also need to keep in mind the difference between the earliest leveled texts which are predictable as you describe, and other leveled texts which increase in complexity based on however the creators of the series choose to level them. I think leveled texts past the first few levels could be seen more as authentic texts - just shorter and more simple than something an adult would read to a child. Some children will learn to decode perfectly well using leveled texts and some deliberate instruction, whereas others continue to rely a lot on sentence structure and meaning cues, and it's those children who I think benefit most from decodables. I'm interested to look a bit deeper into the research studies and see if they distinguish between children with dyslexia or other reading difficulties and "good" readers.

Vanessa, I love your comparison...."we listen to beautiful music and watch others play, knowing this is our end goal. “

Mat--
The trick to beginning reading books is that they have to be easy enough for kids to read -- and that means they usually have to have some intentional alteration of the language to support the kids (but not so much support that they get the wrong idea). Hyper decodability is one of those tools, so using decodables to give kids practice with the skills they are learning makes sense. However, if you overdo it with those texts (as is often recommended these days), then kids learn the alteration you have made which may cause them problems down the road. Another simplification is to use texts with high frequency words (words the students are very likely to have in their oral language -- so no barriers to understanding, but with high amounts of repetition, using those individual words over and over but in different contexts (not the repetitive sentence patterns of predictable books which discourage kids from looking at the words). There are a number of such easy readers (the beginning levels of the Pete the Cat books are an example) and for examples of texts that have both properties -- decodability and controlled vocabualry, I would suggest that you go to the Text Project website. Freddie Hiebert provides such texts for free (and at least some of hers have been used in her research). Those kinds of texts are also an alteration that can become a crutch for some kids but by using both kinds of texts you decrease the chances of kids learning the adjustment rather than how to read.

tim

Hi Tim,
Thank you for this blog post. I am interested in your thoughts on terminology. In previous posts you have discussed decodable, predictable and controlled readers. (https://www.shanahanonliteracy.com/blog/which-texts-for-teaching-reading-decodable-predictable-or-controlled-vocabulary).
You now use the terms decodable and non-decodable. Is there a reason for this change in language? Im wondering if I have missed something.
I actually like the change in terminology and it makes more sense.
Thanks,
Jacinta

Miriam P. Trehearne

With regard to decodable text, evidence suggests that decodability is one important factor in texts for beginning readers. The key words here are “one factor” and “beginning readers”. There is also evidence that suggests other factors in the text are important: “The diversity of genres represented, natural language, and the degree the text is engaging the kids” (Duke, 2019).
To be engaged two important factors are interest and motivation. If students are reading text they are interested in they will more than likely persevere even if the text is challenging! And the more time students spend reading, the better readers they will become. Research shows a significant correlation between high levels of engagement and improved attendance and achievement. “The number one factor that impacts student learning is engaging teaching. Schools must look at changing the culture of education so that teachers are excited about teaching, and students are excited and engaged in learning (Hattie, 2023)”. Highly motivated students who see reading as a desirable activity will initiate and sustain their engagement in reading and thus become better readers.

So let me share a firsthand example of engagement and reading perseverance. Yesterday I was reading with a six-year-old, beginning reader. The teacher had assigned a decodable text to be read and reread daily at home. The student also had a non decodable text that she brought to our session together. She read the decodable text with no interest and little perseverance. She then asked if we could read the non decodable text in which she demonstrated engagement and much more perseverance. Right from Day One we should not ignore the importance of reading engagement and the importance of hooking students on reading.

Notwithstanding, decodable texts, more focus should be on the degree of effective literacy teaching and learning with a strong emphasis on writing. Many students come to reading in part, at least, through writing, which unfortunately is often a curriculum casualty.

This recent study, 2023, is worth a look:
Efficacy of decodable texts and non-decodable texts:
Pugh, Kearns, & Hiebert (2023)

Study used effect size data from three recently published meta-analyses of the effects of reading interventions on reading achievement of students with reading difficulty in kindergarten through third grade.

Effect sizes for interventions with:
• Decodable texts: .50
• Non-decodable texts: .49
• Decodable & non-decodable texts: .66

Thanks again Tim for your ongoing support of literacy educators.

You said "I think it’s okay to use decodable texts as part of phonics instruction, but such practice should be limited, and even beginning readers should be reading (not just listening to) more than decodable texts." I couldn't agree more. I view decodables as one of several possible scaffolds that can be used early on- that means they are used for effect (encouraging & practicing the application of decoding skills) and then removed. By definition that's what you do with a scaffold- think gradual release. I recall a while back you said something to the effect folks claiming we need to use decodalbes for years and years were just making that stuff up. Very happy that you are highlighting the research that gives teachers some guidelines on how to make the most effective use of decodables.

Jacinta--

Sorry for the confusion. Decodable and undecodable are both misnomers. Decodability is not a discrete variable (yes/no, black/white, up/down), but a continuous one. Texts are manipulated for young readers to increase the degree of decodability. Supposedly, undecodable text may actually be quite decodable. The same can be said about vocabulary controls -- how much repetition is there? how many words only appear once or twice in a text? and so on. When studies claim a text is decodable they are either talking about the percentage of words that are spelled/pronounced in the most frequent way or about whether an instructional program had taught the skills needed to pronounce words with those spellings. Undecodable either means that no one manipulated the text to increase its decodability or it really means less decodable (we manipulated these passages to be 70% decodable while these were only 30% decodable by our definition). In any event, by any definition or measure, there is some evidence showing that students will be more likely to sound out words if they are asked to read decodables for a very short period of time -- lessons over 12 days in first grade, and no research showing an advantage when it comes to learning to read.

tim

This is an even more complex issue in an English language learning context, especially late arrivals to secondary education, often with interrupted learning and non-alphabet codes. They do not necessarily have access to a wide range of vocabulary nor often Western contexts to rely on to decode. Often also, their language filters are a mix of helpful and hindrance - they struggle to both hear and produce the sounds orally or written. Certainly a blend of bespoke approaches is critical to learn subject matter at the same time as language acquisition AND literacy. Your thoughts???

Victoria--

You are absolutely right, that can happen, and those students may need basic phonics instruction along with some practice with decodable texts, perhaps. This is especially true of "newcomers." However, with most English Learners language (vocabulary and syntax particularly) are more likely the culprits. You can't know what the students really need without some kind of assessment. If they are falling short on basic decoding (or on more advanced decoding -- with multi-syllable words) then that should be provided. However, in those cases where that is not the case, a heavy emphasis on building English should be the priority.

tim

To begin with, let's be more precise and call these type of texts "phonically controlled' - a much more accurate description. When you think about it, all books are decodable if you know the code they contain!
As a writer of over a hundred phonically controlled texts, (Little Learners Love Literacy and Pocket Rockets) I feel I have something to offer to this discussion. It really depends on the quality of the decodables as to whether they will be of benefit to the learners.
There seems to be a misconception with many people that phonically controlled texts only contain CVC words. CVC worded texts are just the beginning. In a quality series, the stories are not stilted and build up to contain more and more words with more complex spelling patterns, producing naturally flowing phrases and sentences with entertaining plots and information, so much so that you wouldn't even know the text has been phonically controlled!
The odd unlearnt word here and there is not an issue, but to ask children to read both phonically controlled books along with those containing many words with unknown/unlearnt grapheme/phoneme correspondences makes no sense. You wouldn't give someone a piece of music containing chords they hadn't yet learnt.

Berys—
That’s a nice opinion but we determine quality by the measurable effects the books have on the children’s learning. There are no such data. The claim that one set of decodables is more effective than another is a claim with no evidence.

Tim

I think it is worth while reviewing what was done in traditional phonics teaching, which was practised in NZ from 1930-50. Children were not expected to be able to read meaningful stories until about half way through the first year (age 5-6yrs). By then students had learned all the sounds of the consonants, the short vowels, twelve of the simple double consonants, the digraphs, the long vowels, oo, ou and ow with which you have a pretty reasonable vocabulary and can then compose real stories . Up to this half way through stage in the first year the reading consisted of mostly simple sentences with some picture reading (drawings of objects in sentences instead of the words) and incorporating a couple of hundred memorized sight words. The rest of the phonemes and affixes are covered in the following 18 months with more sight words but no sight words in the second year for 6-7 year olds.

Somehow this seems more honest to me than attempting to make up contrived stories from only limited phonemes. Also anyone can create these sentences which saves costs in buying books. eg Ann runs on the wet sand, I see a web on a twig. There were several sentences on a page which can be fairly unconnected to begin with but develop into a single simple page story of 70 words, with or without a picture.

England is fond of Jolly Phonics , which was a programme based on an Austrian system and did not initially take account of dyslexics and other strugglers. But English isn't Austrian and we have a whole heap of irregular spellings and basic Anglo-Saxon non- phonic words which unfortunately are also the most frequently used . Dyslexics have trouble with sight (non phonic) words as well as regular phonics according to researcher Anne Castles of Australia . I find the phonic fanatics have annoyingly, turned sight words into bogey-men.

If your school won't pay for the basic readers you want, then make up your own or have senior students in the school do it. The idea that elementary reading books have to be written by famous literary people or companies is silly. There are many free phonic books on line. Copy the words but have the children draw their own pictures in a cheap exercise book or take the pictures off the internet. Don't be afraid of including many sight words to make your little 50-100 word story flow well. Learn Fry sight words well with flash cards : word one side picture on the other. Make your own. These can be used for spelling as well, with the picture displayed first. (Shops will usually provide you with free reject cardboard from their products if you say you teach (handicapped )children to read ). Send the personalized cards home for parents or extended family to revise. Research has shown flash cards are one of the best ones of recalling and memorizing new knowledge ( See your Hechinger Report on teaching times tables 2023).

I wish NZ academics would read your site Timothy . They are stuck on only decodables and predictive texts ! I wouldn't have any dyslexic or struggler read a predictive aka levelled , reader until they have reached a reading age of eight.

I have always thought of decodable or phonetically controlled texts as the equivalent of the onramp to a highway. It's an important but short passage that leads to increasingly independent and conventional reading. Early childhood educators must always remember that part of our goal is to build that early skill foundation. Another, equally important part of our goal is to build that disposition toward why we read in the first place-- because it's a rich,rewarding and valuable experience on its own. A strict diet of decodable texts can't serve both these ends at once. Love this blog. Learn from it every time.

"And, truth be told, this collection of studies is a bit of a mess; with little evident agreement as to what decodable text is, what it should be compared with, and what learning outcomes should be expected."

Sounds like there isn't what Steve Dykstra calls "bullseye" research in his webinar "An Adaptive, Scientific Approach to Uncertainty." He says “We are not required to be right; we are required to have ‘right’ reasons that we can explain to other professionals . . . Know where the science ends and your best judgment begins.”

I've explained to chatGPT why I choose a variety of texts for my students, including multi-criteria text that have decodability as one criterion--and I was relieved to get the thumbs up. Does this count?

Long time reader, first time commenter; I've greatly appreciated your clarity, knowledge, writing style, and sense of humor over the years.

I have a question about decodables that relates to some of your previous writing on effective grouping strategies, uses of instructional time, and the lack of evidence for sustained silent reading. The literacy curriculum chosen by my district includes independent reading time, even for our youngest learners. The idea is that as they read, they will be practicing the most important skills they need to learn next, guided by teacher assessment and check-ins. A common problem of practice for our teachers is that when asked to read independently, children who can't yet decode the text may just flip through books without much direction - or find other, less productive things to do.

Our curriculum includes pattern texts for emergent readers in which they are to develop concepts of print and letter-sound automaticity. Once students know their letters and have developed some concepts of print, repeating patterns and using picture clues doesn't seem to provide much practice of the foundational skills students need to master to move toward reading fluency. For these children, who more or less know their letters, are getting instruction in phonics, and are writing with invented spelling, the question is how independent reading time can be used productively. Decodable text seems to provide a better answer than pattern text. There's also value in children choosing their own texts, but the books students would like to read can sometimes be far in advance of their decoding ability.

The questions this brings up for me are the following: 1. How valuable is independent reading with teacher assessment and occasional brief check-ins for students at the stage of development described above? and 2. If we think it is of value, how should we guide students to an appropriate balance of decodable text and other text types?

I realize that the research probably can't answer these questions definitively, and I'm curious about how you might make decisions about how to support teachers and students with these questions. Thanks!

Thanks for pulling together the research on decodables! Could you specify which of these references is the meta-anallysis that looked at the 90 studies? Even though I create decodable books for struggling readers, I completely agree with your assessment that they are only part of the solution. There are many rich and engaging types of text available for young readers such as graphic novels, novels in verse, and narrative non-fiction picture books for older students that provide rich resources for previewing topics covered in the curriculum. Many of these text types have features that support developing readers, yet many people in the field of learning disabilities are unaware of these books because of a myopic focus on the nuts and bolts of phonics. If we want to encourage children to fall in love with reading, we need to show them where all of their hard work is heading. Make friends with the school and city librarians. They are a wealth of information... and spend time exploring the children's section of the library.

Carolee--

The Pugh, Kearns, & Hiebert, 2023 article is the one you want.

tim

I am just curious if you have ever reviewed the study over decodable text done by Wiley Blevins in his fresh look at phonics book. I would like to know your thoughts.

Ashley--

Wiley Blevins isn't a researcher. He doesn't do studies. I assume he must have referred to somebody else's research. I'd stay with what I had to say in this blog.

tim