A word that was solid on Tuesday and gone by Friday is one of the ordinary experiences of this job, and it usually means the practice needs changing rather than increasing.
Four adjustments are worth trying first: make the set smaller, connect the word to its letters and sounds rather than showing it whole, correct errors the instant they happen, and practice briefly on more days rather than at length on fewer. Each has a published routine behind it, and the rest of this sets out what that is.
Which of the usual advice is actually evidenced?
Less of it than the confident tone online suggests, and the practice guides say so in a way worth borrowing.
The Institute of Education Sciences publishes practice guides through its What Works Clearinghouse, and each recommendation is assigned an evidence level by an expert panel against a formal standard. The guide defines those levels plainly: strong evidence means positive findings are demonstrated in multiple well-designed, well-executed studies, leaving little or no doubt that the positive effects are caused by the practice; moderate evidence means well-designed studies show positive impacts but questions remain about generalizing them; minimal evidence means there is not definitive evidence that the practice is effective.
In the foundational reading guide, four recommendations carry four different ratings:
- Developing awareness of the segments of sounds in speech and how they link to letters: strong evidence.
- Teaching students to decode words, analyze word parts, and write and recognize words: strong evidence.
- Ensuring each student reads connected text every day: moderate evidence.
- Teaching academic language skills including inferential and narrative language and vocabulary: minimal evidence.
Notice what is at the top. Not a flashcard technique. Not a multisensory trick. Sounds, letters, and decoding, which is the part of the work your child’s school is doing systematically and which most home advice treats as background.
That does not make flashcards useless, and this site uses them constantly. It does mean that when a word will not stick, the highest-value adjustments are the ones that pull practice toward sounds and letters and away from whole-word repetition.
Why do words disappear between sessions?
Because a partially stored word looks exactly like a learned word until it is tested.
Linnea Ehri’s account of how words get filed is useful here. In her 2014 paper, she describes an early phase in which a child forms connections using only the more salient letters of a word, often the first and last, because their knowledge of sounds and letters is still incomplete, particularly for vowels. She notes two consequences directly: similarly spelled words get confused, and unfamiliar words get predicted from initial letters and context rather than read.
A child in that state can name twenty cards correctly in a session, because in a session the context is doing half the work, and then fail on the same words in a book on Friday. Nothing has been forgotten, because nothing complete was stored.
Ehri is also explicit that the storage process is enabled by phonemic awareness and grapheme-phoneme knowledge. That is why the answer to a slipping word is rarely more repetitions of the same whole word. Orthographic mapping covers the mechanism in more detail.
What should you change first?
Shrink the set. It is the cheapest change and usually the one that moves.
The practice guide’s Star Words activity uses three to five high-frequency words on a ring. Johns and Wilke, writing in the Texas Journal of Literacy Education for teachers whose students have difficulty responding automatically to high-frequency words, choose three to five for a first session and add the next three to five only once the first are automatic. The guide explains the reason where it discusses words learned as wholes: it recommends limiting the number introduced at a time, because learning them holistically places considerable demands on a child’s memory.
If five is not working, try three. A short set read many times beats a long set read once, and it removes the daily experience of not knowing most of the pile.
What should you change second?
Stop showing the word whole. Take it apart.
The University of Florida Literacy Institute’s guidance on irregular words distinguishes between words that are permanently irregular, where the letter sounds are unique to that word or a very small number of words, and words that are only temporarily irregular, meaning the child has not yet been taught those letter-sound correspondences. Its approach is to sound out the decodable parts and identify the part that has to be learned by heart, and it notes that flashcards become an effective practice tool once that introduction has happened.
So the order matters. Teach the word by working through its parts, then use the cards to build speed. A card shown cold, over and over, is asking the child to store something they have not yet been shown the structure of. The heart words page has the routine step by step.
One quiet implication of the temporarily irregular category: some words are not sticking because the child has not been taught that spelling pattern yet. Those words are not a problem to solve. They are a word to come back to in a few weeks.
What should you change third?
Correct immediately, and keep sessions short and frequent rather than long and occasional.
The Iowa Reading Research Center at the University of Iowa recommends sight word practice that is brief, under ten minutes per session, but frequent, with immediate and specific feedback. Immediate matters mechanically: a word read wrong and allowed to stand is a wrong version being practiced.
Specific matters too. “Not quite, look at the middle: that ai is the odd part, the rest is just s and d” gives the child something to do. “Try again” does not.
Johns and Wilke’s envelope routine is a good container for all of this. Learned words live in the envelope, words still being worked on are clipped to the outside, every session starts with the current few and then runs the known pile for speed, and anything that stalls goes back to the clip. Old words keep getting revisited at almost no cost, which is exactly what a word that slipped once needs.
What about a child who has stopped wanting to?
Cut the length before you cut the frequency, and put the reading somewhere other than a table.
Two minutes of cards on the way to school, done most days, is worth more than a twenty minute session that ends badly, and it will not become the thing anyone dreads. Games that need no printer covers formats that produce repetitions without looking like practice.
It is also worth keeping the end of each session on something readable. The practice guide rates daily connected text at moderate evidence, and Johns and Wilke finish their routine by putting words into phrases and sentences to promote transfer to context. Finishing on a real sentence gives the evening a point that a pile of cards does not.
When is it time to ask someone?
When you have made those changes, given them a little time, and the same words are still going.
That conversation is with your child’s teacher. They see your child reading in a setting you do not, and they know which phonics patterns have been taught and which have not, which is often the whole explanation. It is a two minute conversation at the door or a short email, and it is an ordinary next step rather than a last resort. Nothing on this site, and no app, is a substitute for it or a way of working out the answer yourself.
Ask sooner than that if the difficulty is broader than a few words, if it is upsetting your child, or if the school has already mentioned it. There is nothing to be gained by working through a list of adjustments first.
Two things are worth holding onto while you wait for that conversation. The first is that a word not sticking is a normal event in learning to read, not a verdict on anything. The second is that the sight words pillar and the practice guides all place this work where it belongs: a small, useful piece of practice beside a much larger job that the school is doing over years.
The narrow help a phone offers here is accuracy. Sight Words Vocabulary Builder holds 94 high-frequency words in four stages, each inside a short sentence, with the words that cannot yet be sounded out gathered in their own deck, and a tap reads the word and the sentence aloud. It is free and has no ads. When the adjustment you are making is “correct it immediately and correctly”, having the right pronunciation available at seven in the evening is worth something.
A smaller set, taken apart before it is practiced, corrected on the spot, a few minutes on most days, and a sentence at the end. That is the whole adjustment.
Sources
- Foundational Skills to Support Reading for Understanding in Kindergarten Through 3rd Grade, NCEE 2016-4008, What Works Clearinghouse, Institute of Education Sciences, 2016
- Orthographic Mapping in the Acquisition of Sight Word Reading, Spelling Memory, and Vocabulary Learning, Ehri LC, Scientific Studies of Reading volume 18 number 1, 2014
- High-Frequency Words: Some Ways to Teach and Help Students Practice and Learn Them, Johns JL and Wilke KH, Texas Journal of Literacy Education volume 6 number 1, 2018
- Irregular and High Frequency Words, University of Florida Literacy Institute
- Teaching Sight Words as a Part of Comprehensive Reading Instruction, Iowa Reading Research Center, University of Iowa, 2018