Screens, words and growing brains: what the research says

1. Screens on their own teach toddlers less than people do

The most consistent finding in this field is the video deficit: between roughly 12 and 30 months, children learn less from a screen than from the same thing happening in front of them. A 2022 review of screen exposure and cognitive development concludes the deficit holds for both interactive and non-interactive screens in that age range.

Large observational studies point the same way. Madigan and colleagues followed 2,441 Canadian children and found that more screen time at 24 and 36 months was associated with slightly lower developmental-screening scores a year or two later. Hutton and colleagues scanned 47 preschoolers and found that heavier screen use was associated with lower integrity in white-matter tracts that support language and literacy.

Two caveats matter. These effects are small (Madigan calls them "small" herself), and both studies are correlational: they cannot show that screens caused the difference. What they do suggest is that time spent passively watching tends to crowd out the play and conversation a young brain needs.

2. What makes a screen worth learning from

The research also shows when screens do work. Roseberry, Hirsh-Pasek and Golinkoff taught new verbs to 24- to 30-month-olds over live video chat, pre-recorded video, or in person. Children learned from the live person and from video chat, but not from the recording. The ingredient was responsiveness: something reacting to the child in real time.

Kirkorian, Choi and Pempek tested 116 children aged 24 to 36 months on touchscreens. The youngest learned a new word only when the screen asked them to touch the exact object being named. Touching anywhere did not help. The interaction has to point the child's attention at the thing that matters.

Hirsh-Pasek and colleagues distilled the science of learning into four pillars for judging an "educational" app: the child must be mentally active, engaged rather than distracted, learning something meaningful to their life, and ideally doing it with another person. A 2021 content analysis using those pillars found most apps marketed as educational score low, and Radesky's team has documented how common manipulative design is in children's apps: instant rewards, time pressure, lures that keep a child tapping.

3. Real objects are where words stick

By their second birthday most children can fast-map: hear a new word once, link it to an object, and remember it later. Spiegel and Halberda showed 2-year-olds doing this after a single brief exposure. The mechanism that makes it work is joint attention: child and adult looking at the same thing while the word is said.

The video-deficit literature adds a hopeful detail: letting toddlers handle the physical object while learning from a screen reduces the deficit. A camera-first app is unusual in that it only works when a real object is present. The screen is a label on the world, not a replacement for it.

Research on augmented-reality vocabulary tools with 5- to 6-year-olds finds they teach words at least as well as paper flashcards and children enjoy them more, but these are small classroom studies and none tested camera recognition of everyday objects. We treat them as encouraging, not conclusive.

4. Three languages: a feature, not a risk

Parents sometimes worry that hearing words in several languages will confuse a toddler. The vocabulary research says otherwise. Bilingual toddlers typically know fewer words in each language than monolingual peers but a comparable number across both, and each language's vocabulary tracks how much of it the child hears. A 2024 study of preterm toddlers, a more vulnerable group, found no delay from bilingual exposure.

The practical lesson: a child will learn the words in the language they hear most. An app that says "cat", "고양이" or "con mèo" on request gives families one more source of input in a minority language, which is exactly what the exposure research says helps.

5. What the guidelines say about age

The World Health Organization (2019) advises no sedentary screen time under age 2, and no more than one hour a day for ages 2 to 4, with less being better. The American Academy of Pediatrics advises avoiding screens other than video chat before 18 to 24 months, co-viewing high-quality content from 18 months, and about an hour a day of quality content for ages 2 to 5. Its 2026 policy statement shifts the emphasis from counting minutes to the "5 Cs": the child, the content, whether it keeps things calm, what it crowds out, and whether it opens communication.

Notice that both bodies treat sedentary, passive viewing as the thing to limit. An app a child uses standing up, moving around the room, looking for things to point at, with a parent naming them back, is closer to play than to television. It still counts as screen time, and we think it should be short.

6. Which ages fit KidWorld

Putting the evidence together, here is how we think about age. This is our reading of the research, not a clinical recommendation; talk to your paediatrician about your own child.

7. How we built KidWorld around this

Nothing in the app is meant to hold a child's attention; it is meant to hand attention back to the world. The camera only shows a word when it is confident, so it never teaches a wrong label. There are no rewards, timers, streaks or ads. The daily gift box offers one word a day, then waits. Settings sit behind a parent check. Recognition runs on the phone and no camera image leaves it.

What we do not have yet is evidence about KidWorld itself. We plan to observe real families using it and will publish what we learn here, including the parts that do not work.

For parents: five habits that make it work

1. Sit beside your child, say the word back in your own voice, and hand them the real thing.
2. Keep it to a few minutes. Stop while it is still fun.
3. Use it as a bridge: "We found the cup! What else is in the kitchen?"
4. Pick the language you want to strengthen and keep that one on for a while.
5. Skip it entirely before 18 months and whenever a screen would replace play, sleep or a meal.