A seven-year-old plays four minutes of a word game on Tuesday and gets reluctant right every single time it appears. On Saturday the word turns up in a bedtime book — the dog was reluctant to go outside — and she stops, rereads the line, and asks what it means. Nothing has gone wrong. That is roughly what one round of practice buys you.
The distance between Tuesday and Saturday is the whole design problem. A vocabulary game either closes that gap or it produces a child who is very good at the game.
Knowing a word is a slider, not a switch
Word knowledge arrives in stages, and any single practice session can only nudge a word along by a stage or so. A rough version of the ladder:
- Never seen it.
- Seen it, no idea.
- Vague sense — something about being slow or unwilling.
- Understands it in a sentence that gives plenty of help.
- Understands it in a sentence that gives no help at all.
- Reaches for it in speech or writing, more or less correctly.
Most word games move words from step two to step four. That is genuinely useful, and it is not the same as the child owning the word. When a nine-year-old asks "does meticulous just mean tidy?" he is sitting at step three and doing fine. Partial knowledge is the normal state of most words in anyone's head, including yours.
This matters because it changes what you should expect from a screen. Four minutes a day will not produce step six. It can produce a steady trickle of words arriving at step four, which is exactly the state a word needs to be in before real reading finishes the job.
Matching a word to its definition is the weakest task in the box
It is also the most common, because it is the easiest thing to build and the easiest thing to mark. There are three problems with it.
First, a child can solve it without understanding anything. Given reluctant and four options, three of which are obviously silly, the answer falls out by elimination. Kids are excellent at this. They learn the shape of the puzzle, not the meaning of the word.
Second, definitions are written in words the child may not have either. "Meticulous: showing great attention to detail" hands a nine-year-old two abstractions to swap for one. Watch a child read a definition aloud and you can often see the moment where they decide to just memorise the string.
Third, and most importantly, matching builds a link between a word and another set of words. What a child actually needs is a link between a word and a situation. Those are stored differently and they behave differently on Saturday night with a book open.
Context does the teaching; the game just arranges the encounters
Compare two sentences a game might show:
- "Ben was reluctant."
- "The soup was still steaming, so Ben was reluctant to take a big spoonful."
The second one does something the first cannot: it constrains the meaning. A child who has never met the word can get within striking distance from the surrounding information alone. That guess — made by the child rather than handed over — is what makes the word stick better than a definition would.
The catch is that one context is not enough, and can actively mislead. A child who only ever sees reluctant next to hot soup may quietly file it as "cautious about food". Two or three different situations across a couple of weeks — hot soup, a dog at the door, a boy who does not want to apologise — are what tell the child which part of the meaning is the stable part.
So the honest description of a good vocabulary game is not that it teaches words. It arranges enough varied encounters, at the right intervals, for the child's own pattern-finding to do the teaching.
How many encounters a word actually needs
There is no clean number, and anyone who gives you one is rounding hard. The general pattern from vocabulary research is that a single meeting almost never sticks, and that most words need somewhere in the range of a handful to a dozen or more meaningful encounters before a child uses them reliably. The spread depends on the word:
- Concrete nouns with a picture attached — volcano, anchor, hedgehog — go fast. Sometimes two or three encounters.
- Everyday abstract words — reluctant, gradual, deliberate — sit in the middle and need repeated exposure across different situations.
- Words with a near-twin already in the child's head — furious when they already have angry — take the longest, because the child has to learn a boundary rather than a meaning.
Run the arithmetic on a real week. If a game puts a target word in front of a child three times in a session, and the child does three sessions in a fortnight, that is nine encounters spread over fourteen days. That is a reasonable dose. If the game shows a word once and never returns to it, it is not teaching — it is testing, and it will look impressive on a summary screen while producing very little.
Spacing beats volume
Ten encounters spread across three weeks do more than twenty encounters in one sitting. The reason is slightly counter-intuitive: the small amount of forgetting that happens between sessions is the useful part. Pulling a word back from the edge of forgetting strengthens it far more than seeing it again while it is still fresh.
A sensible schedule for a word a child met on a Monday looks roughly like this:
| When | What the game should do |
|---|---|
| Same session | Two or three encounters in different sentences |
| 1–2 days later | Bring it back once; expect some wobble |
| About a week later | Bring it back in an unfamiliar context |
| 2–4 weeks later | One check-in; if it holds, retire it from active rotation |
| Occasionally after that | Let it reappear as a distractor or in a reading passage |
You can test whether an app does this without reading any marketing copy. Note down three words your child met in week one, then watch for them in week three. If they never come back, the schedule does not exist.
What a well-built round looks like from the sofa
A few mechanics separate a real vocabulary task from a decorated quiz.
The child produces, rather than recognises. A gap in a sentence with the word missing asks more of the brain than four definitions with one right answer. Typing or saying the word asks more again.
The wrong answers are plausible. This is the single clearest quality signal a parent can check in ten seconds. Good distractors are the same part of speech, the same rough register, and all fit the sentence if you read it carelessly. If three of the four options are obviously wrong, the item teaches nothing.
Feedback restates the meaning in a situation. "Not quite — reluctant means you do not really want to do it, like Ben and the hot soup" is worth ten times a red cross and a retry.
The word count per week is small. Six to ten target words a week is plenty for a six- to nine-year-old. Thirty words a week produces thirty words at step two. Apps such as Kid Genius World that keep a weekly set deliberately narrow are making the right trade.
Audio is present. Plenty of children have heard anxious and never seen it written, or the reverse. Pairing sound and spelling closes that gap in a way a silent screen cannot.
The parts that are decoration
Coins, streaks, avatars and levels are attendance mechanics. They get a child to open the app, which is not nothing, but they teach no words and should not be mistaken for evidence of learning.
Countdown timers deserve more suspicion than they usually get. A word game on a five-second clock rewards fast guessing, which is the precise opposite of the slow retrieval that strengthens memory. If your child is racing a bar and tapping before finishing the sentence, the timer is working against the point of the exercise — and turning it off, where that is an option, usually improves what sticks by the weekend.