My child approaches CAT differently now that he knows it. He looks at the word and recognizes it immediately. With a new word, he still needs to produce the sounds and blend them to work out what it says.
That observation prompted a question. If practicing a word makes it easier to read, how much of that improvement belongs to the particular word, and how much helps with the next one? I had expected the techniques used with familiar words to help with other words too. But that is something to investigate, not something my child’s experience can establish.
Repeating a word can help a child learn that word, which is a valuable part of reading. It doesn’t, by itself, demonstrate an improved ability to read other words. Introducing more different words doesn’t automatically solve that problem either. The useful distinction is between becoming familiar with particular spellings and developing knowledge that can be applied elsewhere.
What has the child learned?
When a child encounters unfamiliar print, they need a way to work out its pronunciation. Decoding means using the relationships between written letters and speech sounds to do that. These relationships can be used across words, which is why learning them is more powerful than learning a separate answer for every word a child might encounter.
Familiar words need not be approached with the same visible effort. A substantial 2018 review of reading development describes how decoding and repeated encounters with print can help children build knowledge of particular written words. This is called orthographic learning, meaning learning the word’s spelling so that it becomes easier to recognize on another occasion.
Successful decoding can do more than supply a pronunciation for that moment. It requires the child to attend to the letters and their order while connecting the spelling with a spoken word. That gives the child an opportunity to learn something specific about how the word is written. Further encounters can make that knowledge more precise and easier to use, so the child may no longer need to work out the pronunciation from the beginning each time. The important development is knowledge of the written word itself, which can support recognition beyond the particular exercise in which it was learned.1
That makes it misleading to dismiss immediate recognition as merely remembering rather than reading. A child who recognizes a familiar written word may be using knowledge that earlier reading helped establish. We don’t want every word to remain an exercise in laborious sounding out indefinitely.
At the same time, recognizing CAT tells us relatively little about how the child will approach an unfamiliar word. My observation fits the distinction between familiar and unfamiliar print, but it doesn’t tell me exactly how the knowledge was acquired. Nor does a fast answer alone tell us whether a child attended to the spelling or relied on another cue, such as a picture or a predictable position in an activity.
What should carry over to another word?
Suppose a child has practiced CAT and is then shown MAT. If the relevant relationships have been taught, the child can use them again, with a different sound at the beginning. That is a reasonable example of transfer, meaning applying learning to material that was not practiced.
But there are two different claims here. One is that knowledge of letters and sounds can be applied across words. The other is that repeatedly practicing CAT will improve the child’s performance on MAT. The first explains an important feature of decoding. The second is a question about what that particular practice achieved.
To answer it, a study needs to look beyond the practiced words. Otherwise, improvement might be entirely genuine but narrower than the claim we are making. The child could have become more fluent with CAT without becoming equally more fluent with other words using familiar relationships.
Would practicing more different words be better?
A Dutch experiment published in 2006 compared these possibilities directly. Researchers randomly assigned 74 Grade 2 children with reading difficulties, averaging 8.3 years old, to two computer practice conditions. One group practiced 20 words 20 times each. The other encountered 400 different words once each. Both completed 10 sessions with 40 target presentations per session.
The practice was unusual. Children made judgments about meaning, such as whether a word answered a question. Target words appeared for only 350 milliseconds before being covered. These were relatively accurate but slow readers, not children first learning to blend simple words. Before and after practice, they read target words and unpracticed words aloud so that accuracy and speed could be compared.
Repetition produced larger gains on the practiced targets. Neither condition showed clear transfer to unpracticed words, including words similar to the targets. There was no later test to establish how long the gains lasted.2
The comparison is interesting because both groups had the same number of target presentations, but those encounters were distributed very differently. Returning to 20 words strengthened performance on those particular words more than seeing each of 400 words once. The much broader exposure did not automatically yield broader transfer. More varied exposure and more transferable learning are therefore not interchangeable descriptions of the same result. That doesn’t make repetition superior reading instruction. It shows why we need evidence about what children can do afterward, rather than judging the practice by the size of its word list.
This isn’t a reason to conclude that varied practice can’t help. It is a reason to question the assumption that a larger word list will necessarily produce broader improvement. Encountering 400 words once is also quite different from being taught how to apply one relationship across a carefully chosen set of examples.
Variation needs a learning purpose
The distinction becomes clearer when we ask what is being held constant and what is changing. With the English spelling EA, for example, MEAT and HEAT share the same ending, while MEAT and LEAF vary the surrounding consonants. Those examples help illustrate a question: does variation make it easier to identify the relevant spelling and sound relationship, rather than associate the answer with a particular combination of letters?
A 2019 experiment in Iowa examined this kind of practice structure in first graders. Data from 417 children were analyzed. Researchers varied the surrounding consonants and whether six vowel spellings were practiced together or in pairs, with the pairs either sharing a vowel letter or not. All groups encountered multiple words and invented words, and repeated items.
Greater variability had no overall advantage. Its effects depended on the grouping, with some analyses favoring varied consonants when overlapping vowel spellings were paired. Effects were small, and the comparison did not consistently establish a benefit. Assessments included items and tasks withheld from practice, with another check 1 to 2 weeks later. However, children mainly selected or constructed answers from spoken prompts rather than independently reading unfamiliar print aloud.3
That result leaves us with a narrower conclusion than “variety improves reading.” The choice of examples and the way they are organized may interact. It also matters whether a test measures choosing the right spelling after hearing a word or working out the pronunciation from print. Both can involve useful knowledge, but they do not demonstrate exactly the same ability.
The researchers worked with the company behind a commercial reading platform. This was a short experiment about practice structure, not a test of the complete product’s effectiveness.
What I would look for beyond the familiar word
The research doesn’t justify treating repetition as wasted practice. I had hoped for a clearer answer about how much practice with one word helps with the next, and these studies leave that question partly open. That doesn’t establish that repeated practice can’t help with new words. The studies are older, concern different reading stages and spelling systems, and don’t establish one ideal balance between familiar and unfamiliar material. These Dutch and English findings also can’t simply be assumed to apply unchanged to Swedish reading instruction or preschool children.
A reasonable practical approach is to revisit words a child is learning while also giving them opportunities to use taught relationships in different words. That is an interpretation of the evidence, not a combined home method these studies have tested and optimized.
The new word needs some thought, including whether the child knows it in speech. CAT and MAT can preserve a vowel relationship while changing the beginning. SUN changes the vowel, and ANT changes the arrangement of consonants and vowels. A word being short doesn’t mean that it asks the child to apply precisely the same knowledge. If the relevant relationships haven’t been taught, difficulty with the word tells us little about whether previous practice transferred.
I would also distinguish recognizing a practiced word now from recognizing it after a delay. Neither is the same question as reading an unpracticed word. These studies don’t establish that 4, 10 or 20 encounters mean the word, or the broader skill, has been mastered.
When my child recognizes CAT immediately, that can be meaningful progress. It doesn’t have to prove more than it shows. The next useful observation is how he approaches another appropriate word, without being given its pronunciation first. Repeated success with CAT and an attempt at a new word tell us different things. Keeping that distinction visible seems more useful than deciding that familiar words or varied words must always be the better choice.
Sources
1. 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 to 51. DOI: 10.1177/1529100618772271. Public full text.
2. Berends, I. E., & Reitsma, P. (2006). Remediation of fluency: Word specific or generalised training effects? Reading and Writing, 19, 221 to 234. DOI: 10.1007/s11145-005-5259-3. Public full text.
3. McMurray, B., Roembke, T. C., & Hazeltine, E. (2019). Field tests of learning principles to support pedagogy: Overlap and variability jointly affect sound/letter acquisition in first graders. Journal of Cognition and Development, 20(2), 222 to 252. DOI: 10.1080/15248372.2018.1526176. Public full text.