I would love to see a blog post on whether to teach sight words/high frequency words, and if there is any useful reason to track whether a student is learning them. My teachers are still teaching them in K and 1st, but more through reading and spelling them, decoding, and encoding them, in and out of text, and not by memorizing their shape. Yet, they are unsure of whether it is worth it to track which words they've learned and how much intervention to provide based on that data.
The short answer to both of your questions is, “yes,” but let’s make sure everybody gets this right.
What is sight vocabulary?
Sight vocabulary refers to all the words a reader can read or recognize immediately without hesitation or apparent sounding or mediation.
What are high frequency words?
High frequency words are those that appear relatively often in written or oral language. Frequency is determined by counting the words in texts.
Aren’t sight words and high frequency words the same thing?
No, it’s important to distinguish these terms. Their confusion may lead to some unfortunate misunderstandings. For example, it may suggest that only certain words can be sight words – that isn’t the case. Any word can become a sight word, no matter its source or frequency. Or it may suggest that readers only need a small collection of sight words, 100 or 220 words. The real goal is much more extensive. Or it could lead some to think that rote memorization is the main way to remember words. That’s not the case either.
What is the best way to learn sight words?
When I was a first-grade teacher I noticed that early in the year my students had trouble remembering new words. We’d review and review, and the next day, the kids often didn’t remember them. Later in the year, I’d introduce new words and that was all it took for many of my students. They seemed to remember those without any effort. What a change!
That means my students weren’t only learning words; they were learning how to learn words. Later, researchers (e.g., Ehri, 1998; Ehri, 2014; Share, 2004) provided more systematic proof of what I’d witnessed and more elaborate theoretical explanations (e.g., orthographic mapping, self-teaching). Basically, phonics instruction – along with phonemic awareness and morphology – helps students to form an internal cognitive memory system that allows them to efficiently remember words.
To help kids develop sight vocabularies in the tens of thousands (which is the real goal), we should provide systematic instruction focused on spelling patterns, relationships between letters and sounds and spellings and pronunciations and meanings. That is where most of our word teaching efforts should be focused.
Are there benefits from rote memorization of some words early on?
There is the obvious benefit of motivation. In my experience, learning a bunch of abstract sounds doesn’t enthrall the average 5- or 6-year-old, while knowledge of real words can be a source of pleasure and pride. Words give a greater sense of accomplishment.
Also, there’s no reason to perseverate on isolated words or word lists for a long period while the students try to master an extensive set of grapheme-phoneme relationships and spelling patterns. Including some high frequency words in a typical phonics curriculum enables students to begin to read texts almost from the beginning and that has both motivational and cognitive benefits (Solity & Vousden, 2009).
Most important, there is considerable evidence showing that students can generalize from memorized words to the decoding of not-yet-known words (Barr, 1972; Brunsdon, Coltheart, & Nickels, 2005; Fletcher-Flinn & Thompson, 2000; Kohnen, Nickels, & Coltheart, 2010; Kohnen, Levlin & von Mentser, 2020; Nickels, Coltheart, & Brunsdon, 2008; Thompson, Fletcher-Flinn, & Cottrell, 1999). As such, it might be best to think of this teaching as part of the phonics curriculum. No one knows exactly what gets stored in memory -- abstract patterns, rules, or even words themselves -- but including words in what we teach seems to help a bit.
Just as abstract spelling patterns and grapheme-phoneme correspondences can be generalized to new words, words themselves can serve as analogies facilitating the reading of new words and the application of decoding skills.
Is it harmful to have students memorize words?
There are folks who make that claim. They say that if you teach students to memorize any words you will disrupt their decoding or encourage guessing.
No research supports those cautions, and some of the sharpest eyes in the room argue against these groundless claims. For example, Castles, Rastle, and Nation (2018) wrote in response to this question:
“In our view, this concern is unwarranted, and the judicious selection of a small number of sight words for children to study in detail has its place in the classroom alongside phonics. As we have discussed, teaching phonics is crucial because it gives children the skills to translate orthography into phonology and thereby to access knowledge about meaning. However, when this is difficult because of spelling-to-sound complexities, there would seem to be a case for teaching children the pronunciations of a small number of such words directly, particularly those that they are likely to see very frequently in the texts they are reading (such as the, come, have, and said). In effect, this ensures that children can relate the visual symbols of writing to spoken language for as many words as possible and as early in their schooling as possible.” (p. 15)
Is there evidence showing that teaching students to memorize sight words improves reading achievement?
I depend upon studies designed to determine if the use of a certain curriculum or instructional approach provides a learning advantage to students; particularly those that consider whether that teaching generalizes to overall reading achievement rather than just gains in the skill taught. That kind of gold standard evidence does not exist with sight word teaching.
However, there are several studies showing that including such sight word teaching in phonics curricula can be effective – and even that such inclusion improves performance (Kohnen, Nickels, Coltheart, & Brunsdon, 2008; McArthur, Castles, Kohnen, Larsen, Jones, Anandakumar, & Banales, 2015; Price-Mohr & Price, 2017; Shapiro & Solity, 2008; Solity & Shapiro, 2008; Torgesen, 1999; Vellutino & Scanlon, 1986; Wright & Ehri, 2007). Adding word teaching to the phonics regimen was beneficial both in the regular classroom and with dyslexic students. The inclusion of such word teaching can speed beginning reading progress along a bit (Browder & Lalli, 1991; Colenbrander, Wang, Arrow, & Castles, 2000).
How do you teach a child to memorize a word?
Teaching this kind of memorization is pretty straightforward, but there are some key steps. For instance, it’s important to isolate the word from other visual distractions. Students when trying to learn a word need to look at it word, not at pictures or other words in a sentence. Make sure students’ attention is on the written word, then say the word and have students repeat it.
It may help to put the word into meaningful context to convey meaning or usage (Miles, McFadden, & Ehri, 2019). I do this orally. When teaching the word “with”, the teacher may say something like, “I play with my brother.” Who do you play with? With.” Students need to link the visual image of the letters “w-i-t-h” to the phonological representation /w/-/i/-/th/, and we want this visual and phonological information linked to the word’s meaning and grammatical function.
Have students analyze the letters and sounds of the word. Don’t focus on a word’s shape, but on its sequence of letters. Have students spell the word. Perhaps ask if they know the sounds of any of those letters, even if the spelling is irregular, and certainly point out any of the letters that have their usual sound. This not only helps students to learn that particular word but to use this knowledge more generally (Murray, McIlwaom, Wang, Murray, & Finley, 2019). Encourage students to visualize the word (“take a picture with your eyes”) and get them to write/spell the word without looking. Some teachers use the “copy-cover-compare” approach successfully (Joseph, Konrad, Cates, Vajcner, Everleigh, & Fishley, 2011).
Finally, a bit of drill and practice is in order. Reading that word again and again over time helps with memorization. This is where things like flashcards, word rings, word ladders, and the like can come in handy.
What about text?
When first teaching a word, it is important to isolate it. But remember that one of the purposes for introducing words is to enable earlier text reading. Students should see these words in early instructional texts. They need to be able to read these words not just on flashcards, but in texts. The opportunity to confront these words in varied sentence contexts within controlled vocabulary readers or decodable texts should be part of the ongoing experience that follows their introduction.
Which words should we teach that way?
As has been recommended for 100 years, it is sensible to focus on high frequency words since those will be the most useful to the child’s reading. A high frequency word might be taught through memorization if its spelling is irregular or unusual or if the student has not yet learned the phonics skills that would support its decoding. Studies have shown that a relatively small number of words represent a large portion of the words that readers confront in text; for example, the first 100 words account for more than half the words we see in text (Fry, 1980). Knowing such words well should facilitate fluency and allow greater attention to the rest of the words.
Any other instructional advice?
Yes, those estimates of word frequency can be misleading. The first 100 words make up more than 50% of the words in texts, but only if you consider those words along with their common variants. The word “make” is included in those first hundred words, but it’s there because of make, making, and makes. That fact should suggest morphological work in which students transform sight words into various forms. What a great opportunity to explore the spellings, pronunciations, and meanings that inflectional and derivational morphemes present.
How many words should we have students memorize in this way?
Sight word teaching tends to be overdone. Some commercial programs go over the top including way too many words. The National Academy of Education released a report in which the experts recommended that kindergartners master about 18-20 such words (including their names). There is no research on this, but I think that is a reasonable (and smart) recommendation.
As for Grade 1, I’ve long encouraged teachers and parents to make sure students can read the 100 most frequent words in written English (Fry’s first 100 words). That sounds like a lot, but it includes those 18-20 words mastered in kindergarten. Likewise, more than 50 of those 100 words can be read directly and completely through the most common grapheme-phoneme relationships (e.g., it, he, but, not), and most of the others can be partially decoded with those skills. Of course, that means not all those words need to be taught through memorization. For second graders, emphasize the 300 most frequent words (which includes the 100 first-grade words – and, again, many of those can be learned and read through decoding).
How much time should be devoted to such word memorization?
I’d not put much time into it… only about 3-5 minutes per day. I’ve found reports of 30-45-minute sight word memorization lessons, which I think is nuts. I have also located studies of 1-3-minute instruction – which makes me wonder if my 5 minutes is overkill.
Many teachers push at least some of this work out of the classroom. It’s an easy thing for parents to help with. One of my favorite principals tested all her first graders on the first hundred words. She gave the results to the teachers and parents and told them the kids needed to know them all by the end of year. Most of the kids nailed it by Thanksgiving, mainly due to parental involvement. Word memorization should be a tiny part of the word knowledge instruction that students receive.
Should we monitor student progress with sight vocabulary?
I think so. That’s the only way to know which words the students know and how much progress they are making. But this can be done during the instruction. For instance, while students are practicing reading those words with partners, a teacher can easily and efficiently check several individual students’ progress with 10-20 words. No extra testing time needed.
Barr, R.C. (1972). The influence of instructional conditions on word recognition errors. Reading Research Quarterly, 7(3), 509-529.
Browder, D.M., & Lalli, J.S. (1991). Review of research on sight word instruction. Research in Developmental Disabilities, 12, 203-228
Brunsdon, R., Coltheart, M., & Nickels, L. (2005). Treatment of irregular word spelling in developmental surface dysgraphia. Cognitive Neuropsychology, 22, 213–251.
Castles, A., Rastle, K., & Nation, K. (2018). Ending the reading wars; Reading acquisition from novice to expert. Psychological Science in the Public Interest, 19(1), 5-51.
Colenbrander, D., Wang, H., Arrow, T., & Castles, A. (2020). Teaching irregular words: What we know, what we don’t know, and where we can go from here. Educational and Developmental Psychologist, 97-104.
Ehri, L.C. (1998). Grapheme-phoneme knowledge is essential for learning to read words in English. In J.L. Metsala & L.C. Ehri (Eds.), Word recognition in beginning literacy, (pp. 3–40). Mahwah, NJ: Erlbaum.
Ehri, L.C. (2014). Orthographic mapping in the acquisition of sight word reading, spelling memory, and vocabulary learning. Scientific Studies of Reading, 18(1), 5–21. https://doi.org/10.1080/10888438.2013.819356.
Fletcher-Flinn, C. M., & Thompson, G. B. (2000). Learning to read with underdeveloped phonemic awareness but lexicalized phonological recoding: A case study of a 3-year-old. Cognition, 74, 177–208.
Fry, E. (1980). The new instant word list. Reading Teacher, 34(3), 284-289.
Joseph, L.M., Cates, G., Vajcner, T., Eveleigh, E., & Fishley, K.M. (2011). A meta-analytic review of the cover-copy-compare and variations of this self-management procedure. Psychology in the Schools, 49(2), 122-136.
Kohnen, S., Nickels, L., & Coltheart, M. (2010). Skill generalisation in teaching spelling to children with learning difficulties. Australian Journal of Learning Difficulties, 15, 115–129.
Kohnen, S., Nickels, L., Coltheart, M., & Brunsdon, R. (2008). Predicting generalization in the training of irregular-word spelling: Treating lexical spelling deficits in a child. Cognitive Neuropsychology, 25, 343–375.
Levlin, M., von Mentser, C.N. (2020). An evaluation of systematized phonics on reading proficiency in Swedish second grade poor readers: Effects on pseudoword and sight word reading skills. Dyslexia, 26, 427-441.
McArthur, G., Castles, A., Kohnen, S., Larsen, L., Jones, K., Anandakumar, T., & Banales, E. (2015). Sight word and phonics training in children with dyslexia. Journal of Learning Disabilities, 48(4), 391-407.
Miles, K.P., McFadden, K.E., & Ehri, L.C. (2019). Associations between language and literacy skills and sight word learning for native and nonnative English-speaking kindergartners. Reading and Writing, 32, 1681-1704.
Murray, B.A., McIlwaom, M.J., Wang, C., Murray, G., & Finley, S. (2019). How do beginners learn to read irregular words as sight words? Journal of Research in Reading, 42(1), 123-136.
Price-Mohr, R.M., & Price, C.B. (2017). Synthetic phonics and decodable instructional reading texts: How far do these support poor readers? Dyslexia, https://doi.org/10.1002/dys.1581
Shapiro, L. R., & Solity, J. (2008). Delivering phonological and phonics training within whole class teaching. British Journal of Educational Psychology, 78, 597–620.
Share, D. (2004). Orthographic learning at a glance: On the time course and developmental onset of self-teaching. Journal of Experimental Child Psychology, 87, 267–298.
Solity, J., & Shapiro, L. R. (2008). Developing the practice of educational psychologists through theory and research. Educational and Child Psychology, 25, 119–145.
Thompson, G., Fletcher-Flinn, C., & Cottrell, D. (1999). Learning correspondences between letters and phonemes without explicit instruction. Applied Psycholinguistics, 20, 21–50.
Vellutino, F. R., & Scanlon, D. M. (1986). Experimental evidence for the effects of instructional bias on word identification. Exceptional Children, 53, 145–155.
Wright, D., & Ehri, L.C. (2007). Beginners remember orthography when they learn to read words: The case of doubled letters. Applied Psycholinguistics, 28, 115-133.
I'm surprised by the number of sight words you suggest to be memorized per grade level. This is the amount my former district recommended in its curricula, but it fell short supporting the benchmark that students were expected to read at various points during the year. On the other hand, my district also used a poor resource as its phonics curricula (Fountas and Pinnell), so maybe that explains the disconnect.
Dear Dr. Shanahan,
Thank you so much for this week's blog post! It is very timely and helpful!
I'm still trying to wrap my head around Mark Seidenberg's latest blog posts about how the "Science of Reading Movement is moving without much science." You state above "We should provide systematic instruction focused on spelling patterns, relationships between letters and sounds and spellings and pronunciations and meanings" and "Master an extensive set of grapheme-phoneme relationships and spelling patterns." While Seidenberg agrees with that, "Let there be no confusion: Beginning readers need instruction to gain foundational skills," he also states that there is an "Overemphasis on phonics rules, sight words, spelling rules," and that "Learning rules is slow. It is another form of rote learning. Learning a rule has a benefit: generalization. But, costs are very high and rules are not the only basis for generalization. Neural networks generalize without learning rules."
Can you please help explain where that leaves a first grade teacher? How do we sift through all of this information to make informed decisions on what is best to teach our students?
My advice is to stay to the "science of reading instruction" rather than the "science of reading." When people are telling you how and what to teach, you should ask for direct evidence that teaching those things or teaching in those ways has been proven to benefit children's learning. If someone says teachers "are teaching too much _______ (fill in the blank)" ask how they know that... how many schools/teachers have they observed in or what survey data are they dealing with? If they say that you should teach a particular thing, ask not about the basic cognitive or psychological research that makes them think that may be true, but about the classroom trials that have been run to test the claim?
I don't believe that practitioners in any field should be determining their practice on the basis of basic science -- the same standard that medical practitioners work under.
I gave not just those numbers but an explanation of them and also an argument for why most word time should be spent on decoding (and why even teaching those words is at least partially about decoding). How much phonics were you teaching (the typical study that has shown phonics to benefit kids has been for about 30 mins per day)?
Yes, you did explain those numbers and argue why most word time should be on decoding. The explicit phonics lesson was 10 to 20 minutes with further word practice at work stations or in small group if needed.
I didn't say decoding for 10-20 minutes. In past blog entries I've argued for 30 minutes per day based on the research studies that have been done. Up to 5 minutes on sight words and 30 minutes on PA and phonics.
The 10 to 20 minutes of instruction time I mentioned was my answer to your question, “ How much phonics were you teaching…”
Our district uses sight word assessments as one of our major pieces of data. I find that there are kinders that come to me able to read over 100 word on the list but cannot read them in texts or come close to writing them. Any thought on this? I
Tim, the considerations are clear! Thank you. I work with student teachers using a strategic hand metaphor. The thumb symbolizes development of everything, the index finger symbolizes interests in significant objects, the middle finger symbolizes aspects of activity in the subject according to the curriculum, the ring finger symbolizes productive exercise, and the little finger symbolizes automaticity. Whether we treat texts or words, it always happens dialectically WITH the memory of all and not for the memory, as already Plato differentiated.
That isn't surprising. The benefits of rote memorization of lists of words are quite limited. Studies show that it is not enough to memorize a list of words but students need to see those words in text. In fact, that is one of the main reasons for teaching any words through memorization -- to enable kids to read texts that their decoding skills are not yet sufficient for. You teach the words and then have students doing the rest of what is counseled above -- have students read texts with those words in it, have students work with those words to recognize their different inflectional forms, etc.
Teach the words but then teach kids to use the words. You might consider testing students two ways on those words -- once in isolation and once in text.
Ah, I didn't understand that. No, that is too small an investment. Try 5 minutes on words and 25-30 on decoding and I think you'll see a difference. Good luck.
What is your response to neuroscientist Dr. Stanislas Dehaene's statement, “Whole word reading is a myth…the brain does not use the whole word shape”?
In the videos below, it appears that Dehaene is saying that skilled readers just become faster at accessing individual letters and their sounds rather than reading whole words. Neurologically, each sound is stored separately, so we actually have to activate and connect them to read words.
Dehaene is correct (as far as I know), there is no such thing as reading words by shape. We've known for decades that readers, when reading, are looking at pretty much every letter as they proceed through a text. However, Dehaene is looking at what the brain appears to be doing during reading, but he ignores what the research on teaching has to say. For example, studies have shown that having words stored in memory help students to develop the process that Dehaene describes. Neuroscience has revealed some fascinating things about how people read. Unfortunately, it has -- so far -- told us very little about how people learn to read. There is nothing wrong with trying to drawn inferences from that very fallible brain research, but those inferences need to be consistent with what we know about learning.
I’m not understanding why you would teach a child to memorize “with” when that can be decoded. Isn’t it better to save any memorization to those high frequency words that can’t be decoded?
The following quote indicates that Castle, Castle and Nation (2018) may not understand how high frequency words are taught in the typical American classroom. In my district, teachers and parents are specifically told not to allow the students to study the words in detail because they want an automatic response to the mere sight of the word. The people who believe teaching kids to memorized high frequency words from memory is harmful would not find it objectionable to teach the kids to study the high frequency words in detail. Studying a high frequency word in detail would be a form of "segmenting" not "whole word memorization by mere sight."
No research supports those cautions, and some of the sharpest eyes in the room argue against these groundless claims. For example, Castles, Rastle, and Nation (2018) wrote in response to this question:
“In our view, this concern is unwarranted, and the judicious selection of a small number of sight words for children to study in detail has its place in the classroom alongside phonics.
Can you give me any direction regarding sight words with striving 9th grade readers? They are reading three years below grade-level, so I'm wondering if we should spend some time on sight words.
No, I would not focus on sight word development with students three years behind (in the 9th grade). I would work on oral reading fluency, vocabulary, and comprehension (and possibly a small amount of work on decoding multi-syllable words depending on the students' skills). Sight vocabulary is the goal of decoding instruction -- and for students reading K-2 level some memorization of sight words is in order, too, but not beyond those early reading levels.
It makes sense to memorize sight words when (1) the words are unusual in their spelling patterns and/or (2) when a student does not yet know the decoding skills needed to take on that word successfully. Words like "with" can be very useful in a decodable text but the /w/ sound is probably not going to be taught especially early in the phonics sequence. There is no good reason to avoid that word or the texts that might include it.
Thank you so much for your perspective on Dehaene's statement. I appreciate your distinction between how the brain reads vs. how the brain learns to read. I agree!
What I see in classrooms and on worksheets teachers buy and send home concerns me very much. Such attention is being applied to "sight words" from the Fry list that many students intitially begin to believe that reading is memorizing. This belief becomes a hard habit to break. It seems crucial that our initial focus (and our major focus) be on helping kids relate sound to symbol. Neuroscientist Stanislas Dehaene makes this point in his book Reading in the Brain when he says this needs to be our entire focus in the beginning. If they can articulate the sounds n, t, p, and a and can write those letters correctly and begin to read and spell words like an, at, pat, tan, pan, etc. ..and they get SOOO good at connecting these sounds and symbols that these words become "mapped" and in the brain's visual word form area (where they'll never have to be sounded out again!! Woot!). Then we can add more sound/symbols. Similarly, even irregularly spelled words, like said, can become "mapped." To expect a child to "read" the word 'for' before the /f/ and /r/ are even taught is to send a mixed message. Why not teach decoding sound-symbol without question for a few months to make sure they truly understand what reading is; then gradually teach a few irregular words (as needed), but drawing attention to the parts that are regular and providing the repetitions so the word becomes orthographically mapped and into the visual word form area forevermore.
Thank you for your response! This really helps with curriculum plans for next year.
Lists can be very abstract. I’ve switched first to Fry word phrases/sentences. It’s working out super well. Ehri referenced using phrases and Tim Shanahan mentioned it too.
Hi Dr. Shanahan,
I have a specific question about assessing high frequency words. I agree with what you mentioned about assessing students in the moment as they practice their words. However, our district uses more formal one-to-one assessments, particularly for grading and reporting purposes. My question, therefore, has to do with the format of the assessment. Do you feel that having students read the individual high frequency words in isolation gives sufficient information on their level of mastery, or should students also read all of these words (all 100 in 1st grade, for example) within the context of a sentence on an assessment to show mastery? Obviously, reading the words in context is the ultimate goal, but this takes much longer during a formal assessment, as you can imagine. If a student can confidently read them one by one, would you consider that to be evidence that the child has a firm grasp on reading the words? Or would you not feel confident that the child has learned them until you heard them read all 100 words in the context of a sentence or text?
There are studies that show students may know a word in isolation but still have trouble with it in context (and vice versa). I find the same for myself in learning French -- words I will know on flashcards, sometimes stymie me in print. So doing some testing both ways is a good idea. I would suggest that you consider not testing students on all of the words to save some time that could then be used to check some words in context-- consider testing students with a random sample of words and sentences -- perhaps 30 words in isolation and 10 in context. The key to random testing in that way is to not select the words until you are ready for the testing (that way you won't find yourself giving some words more attention just because you know they will be evaluated). For the words in sentences, try to compose sentences with words that students are likely to be able to read (and only score them on that one word). If a student balks at a sentence, it is okay to help with any of the other words -- or if the student struggles, you may say something like: "Do you see any words there that you can read?"
Tim- I really do post reminders from you blog about how to best teach reading on my office wall. Starting to run out of room. This one's going up today "“In our view, this concern is unwarranted, and the judicious selection of a small number of sight words for children to study in detail has its place in the classroom alongside phonics. As we have discussed, teaching phonics is crucial because it gives children the skills to translate orthography into phonology and thereby to access knowledge about meaning. However, when this is difficult because of spelling-to-sound complexities, there would seem to be a case for teaching children the pronunciations of a small number of such words directly, particularly those that they are likely to see very frequently in the texts they are reading (such as the, come, have, and said). In effect, this ensures that children can relate the visual symbols of writing to spoken language for as many words as possible and as early in their schooling as possible.” (p. 15)" Also sending it to a tchr who was doing ALL high frequency words by rote. Thanks for your insights. I don't always agree with everything you say but on this one you are totally on the mark. Sam from St. Louis
I am an experienced primary school teacher in Melbourne, Australia. I have started using your steps to memorise sight words with my students and found it very effective so far. Thank you for sharing these strategies. Very helpful.
Leave me a comment and I would like to have a discussion with you!
Copyright © 2023 Shanahan on Literacy. All rights reserved. Web Development by Dog and Rooster, Inc.
See what others have to say about this topic.