An honest review of what the research really says about babies, toddlers and "educational" screens: where the label is marketing, where it is true, and what changes the answer.
For children under about two and a half, the honest answer is mostly no. Decades of research point to something called the video deficit effect: very young children learn dramatically less from a person on a screen than from the same person in the room. A baby watching a video labelled "educational" is not getting a lesson. She is watching coloured shapes move while the learning machinery in her brain, which is built for live human interaction, largely sits idle.
For children over three, the answer becomes more nuanced, and this page will be honest about that too. Well-designed, slow-paced content can teach specific things to preschoolers, especially when a parent watches alongside and talks about it. But the word "educational" on an app store listing or a streaming thumbnail is, in most cases, a marketing claim rather than a tested one. Knowing the difference is the whole game, and it is knowledge the people selling the content have little incentive to give you.
The video deficit effect is one of the most consistently replicated findings in early developmental science. In study after study, children under roughly 30 months who are shown new words, actions or problem-solving steps on video learn far less, and often nothing measurable, compared with children taught the same material by a live person. The information is identical. The delivery is what fails.
Patricia Kuhl's work at the University of Washington showed this vividly with language. American infants exposed to Mandarin through live tutors picked up the sounds of the language. Infants exposed to exactly the same tutors, on exactly the same schedule, via video or audio recordings learned essentially nothing. Sarah Roseberry's research with Kathy Hirsh-Pasek and Roberta Golinkoff added the crucial detail: when toddlers interacted with a person over live video chat, where the adult could respond to them in real time, learning came back. The missing ingredient was never the screen itself. It was social contingency: a real person noticing the child, responding to the child, and adjusting to the child, moment by moment.
Key citations
Kuhl, P. K., Tsao, F.-M. & Liu, H.-M. (2003). Foreign-language experience in infancy: effects of short-term exposure and social interaction on phonetic learning. PNAS · Roseberry, S., Hirsh-Pasek, K. & Golinkoff, R. M. (2014). Skype me! Socially contingent interactions help toddlers learn language. Child Development
This is why the question "do babies learn from TV" has such a clear answer. A pre-recorded programme cannot notice your baby. It cannot follow her gaze, repeat the word she almost said, or pause when she looks away. For a brain that learns through back-and-forth exchange, broadcast video is a conversation with someone who never listens.
Here is the uncomfortable part that parents are rarely told: the word "educational" in an app store or a programme description is largely unregulated. No body tests these claims before they appear. A developer can build a game of tapping balloons, add a letter of the alphabet to each balloon, and ship it under the education category.
When Kathy Hirsh-Pasek's team reviewed the educational app marketplace, they found that the overwhelming majority of apps marketed as educational failed basic criteria drawn from the science of how children actually learn. The researchers identified four pillars that genuine learning experiences share: the child is mentally active rather than passively tapping, engaged rather than distracted by bells and rewards, the content is meaningful and connects to her life, and there is social interaction around it. Most "educational" apps deliver none of the four. They deliver swiping.
Key citation
Hirsh-Pasek, K., Zosh, J. M., Golinkoff, R. M., Gray, J. H., Robb, M. B. & Kaufman, J. (2015). Putting education in "educational" apps: lessons from the science of learning. Psychological Science in the Public Interest
None of this means app developers are villains. It means the label is doing sales work, not science work, and a tired parent scanning a category page at 9pm has no way of telling the difference. That is not your failing. It is a gap in regulation that the attention economy is happy to leave open.
Honesty cuts both ways, so here is the part of the evidence that favours the screen. From around age three, children can genuinely learn specific, nameable things from well-designed television: vocabulary, letters, numbers, and prosocial ideas like sharing and kindness. The research on carefully constructed, curriculum-based children's programming, the slow-paced kind with a clear story, direct address to the child, and deliberate repetition, shows real and lasting gains for preschool viewers.
The conditions matter enormously, though. The gains appear when the content is designed around a learning goal rather than around retention metrics, when the pace is slow enough for a young child to follow cause and effect, and when an adult is nearby to extend what the programme started. A 4-year-old watching twenty minutes of thoughtful, story-led content with her dad, who then talks about it at tea, is having a different experience from the same child autoplaying short-form clips alone. Both will be called "screen time" in a survey. Only one resembles education.
If there is one practical lever in this whole literature, it is co-viewing. When a parent watches alongside a child and talks, the screen starts to behave a little more like a picture book: a shared object of attention that generates conversation. You name things, ask questions, connect the story to the child's own world. The social contingency that video lacks, you supply yourself.
This is also the kindest finding in the research, because it hands control back to the parent. You cannot redesign a programme, but you can sit down for ten minutes of it, point at the screen, and chat. Researchers consistently find that the same content produces better language and comprehension outcomes when it is co-viewed and discussed than when it is consumed alone. The screen does not teach. The conversation around the screen can.
There is a second axis besides content, and that is pace. Dimitri Christakis and colleagues have studied the relationship between early exposure to television and later attention problems, and a recurring theme in this line of research is pacing: rapid scene cuts, constant novelty, and stimulation arriving faster than real life ever delivers it. The concern is that very fast-paced media trains a young brain to expect a tempo of reward that sand pits, jigsaw puzzles and conversations cannot match, making ordinary, slower forms of attention feel effortful by comparison.
Key citation
Christakis, D. A., Zimmerman, F. J., DiGiuseppe, D. L. & McCarty, C. A. (2004). Early television exposure and subsequent attentional problems in children. Pediatrics
The evidence here is genuinely mixed in places, and serious researchers debate the size of the effect. But the practical guidance it points to is uncontroversial: if a young child is going to watch something, slow-paced beats fast-paced. A programme where a scene lasts long enough for a 3-year-old to understand who wants what, and why, is doing less to her attentional expectations than a cascade of two-second cuts.
Step back and admire, for a moment, the most effective marketing sentence ever aimed at tired parents: at least it's educational. It is brilliant because it does not sell a product. It sells absolution. Every parent needs twenty minutes to cook, answer a work message, or simply sit down, and every parent feels the low hum of guilt when the screen goes on. The word "educational" converts that guilt into virtue. The screen time you were going to allow anyway becomes, with one label, a lesson you have thoughtfully provided.
No individual show or app needs to be named or blamed here, and this site will not do it, because the issue is structural rather than personal. An entire industry has learned that parents do not buy children's content. They buy permission. Once you see the label that way, it loses its power. You can still use the screen when you need it. You simply stop letting a marketing department tell you what your child is learning from it.
The same science that deflates the "educational" label points clearly at what works: hands-on play, conversation, and materials that make the child do the thinking. A wooden shape sorter teaches geometry through the fingers. A simple board game teaches turn-taking, counting and losing gracefully. Building blocks teach physics by falling over. None of this needs to be branded educational, because the learning is built into the activity rather than printed on the box. This is the principle behind traditional educational toys → and it has not changed since long before screens existed.
The richest version of all is the one the video deficit research keeps pointing back to: another person. A parent rolling a ball, naming animals in a toy ark, or playing a slow game of skittles in the garden is providing exactly the socially contingent, responsive interaction that no recording can. Simple, open-ended wooden toys → are not magic. They are simply very good at creating the conditions, shared attention, conversation, repetition, in which young children demonstrably learn. Jaques has been making them since 1795, which is less a boast than a reminder that childhood ran on these materials for a very long time before anyone needed a study to defend them.
Realistically, no. The video deficit effect is strongest at this age: 1-year-olds learn very little from pre-recorded video, however it is labelled. The one meaningful exception is live video chat with a responsive person, such as a grandparent on a call, because the real-time back and forth restores the social ingredient that learning depends on. If you use video at this age for a breather, do it without guilt, but do it without expecting a lesson.
Not automatically, and often not at all. Most apps marketed as educational fail the basic criteria of learning science, and many add reward loops and effects that a calm television programme does not have. An app is only better if it makes the child genuinely think, stays free of distracting gimmicks, and ideally involves you. A frantic tapping game with letters on it is closer to a fruit machine than a lesson.
Yes, and comparatively it is much better than solo viewing. When you watch together and talk about what you see, you supply the responsive interaction the screen lacks, and research consistently shows children understand and retain more this way. If television is part of your family's day, sitting down for some of it and chatting about it is the single most useful change you can make.
Slow-paced content holds each scene long enough for a young child to follow it: few cuts, a clear story, characters who speak directly and pause for a response. It matters because rapid cuts and constant novelty deliver stimulation at a rate real life cannot match, and researchers including Dimitri Christakis have raised concerns that heavy early exposure to that tempo is linked to later attention difficulties. A useful test: if you can describe what happened in the last minute of the programme, it is probably slow enough.
The evidence says no. Patricia Kuhl's studies found infants learned the sounds of a new language from live tutors but not from identical video recordings, and later research has repeatedly confirmed that babies need a responsive human partner to learn words. Talking, singing and playing with your baby, in any language and at any pace, does more for her speech than any video designed for the purpose.
Britain's oldest toy maker. Every toy open-ended, wooden, and built for exactly the kind of play the research recommends.