Dr. Roberta Golinkoff, the language scientist who put app-store claims under the microscope, on what children actually need in order to learn: and why ordinary play has had it all along.
Dr. Roberta Golinkoff is Professor of Education, Psychology and Linguistics at the University of Delaware, and one of the most influential researchers in the world on how young children learn language. With her long-time collaborator Kathy Hirsh-Pasek, she co-authored the landmark review of "educational" apps: the paper that asked, with the full weight of the learning sciences behind it, whether the tens of thousands of apps sold to parents as educational actually teach anything.
That pairing matters. Golinkoff is not a campaigner who decided screens were bad and went looking for reasons. She is a scientist who has spent her career working out, in fine detail, how babies and toddlers crack the code of language. When someone with that background examines the app store, the question is not "are screens evil?" but something far more useful: what does learning actually require, and does this product provide it?
In 2015, Golinkoff, Hirsh-Pasek and colleagues published a major review in Psychological Science in the Public Interest, one of psychology's most rigorous outlets. Its title gives the game away: Putting Education in "Educational" Apps. Their starting observation was blunt. The "educational" label in app stores is a marketing category, not a scientific one. Anyone can apply it. No body checks it. A developer's claim that an app teaches letters or numbers requires no more evidence than a cereal box's claim to be part of a balanced breakfast.
Rather than simply complain about this, the review did something more durable. It distilled decades of research from the science of learning into four pillars that real learning requires, whatever the medium. Learning happens best when a child is active: minds-on, thinking and doing, not passively receiving. When she is engaged: focused on the task itself, not distracted by bells, rewards and pop-ups that hijack attention. When the material is meaningful: connected to her own life, so new information has something to stick to. And when learning is socially interactive: happening with another person, in conversation, with shared attention and response.
The source
Hirsh-Pasek, Golinkoff et al. (2015), Putting Education in "Educational" Apps, Psychological Science in the Public Interest · Read the original
Held against those four pillars, most apps marketed to young children fail most of them. Many demand only tapping and swiping: physically active, mentally passive. Many are engineered around the very distractions, sound effects, animations and reward loops, that the engagement pillar warns against. Few connect to anything in a particular child's actual life, and almost none involve another human being. Meanwhile, the review's quiet counterpoint is the one parents most need to hear: ordinary physical play passes all four pillars by default. A child building a tower with her dad is active, engaged, working on something meaningful to her, and locked in social interaction, all at once, without anyone having designed it that way.
The practical gift of this research is that it replaces a vague worry with a usable test. Parents are routinely told to choose "high-quality" apps and "educational" content, advice that is almost useless because nobody says what quality means. The four pillars say what it means. Stand in front of any app, any toy, any activity, and ask four questions. Is my child's mind doing the work, or is the thing doing it for her? Is she focused on the task, or on the rewards and noises around it? Does this connect to anything in her real life? And is anyone playing with her?
Be honest about what the framework does and does not claim. It does not say a four-year-old can learn nothing from a screen; the review itself argues that apps built on the four pillars could teach, and a video call with grandma is genuinely social. Nor does it say every minute of a childhood must be optimised for learning. What it says is narrower and more useful: the word "educational" on a download page tells you nothing, and the burden of proof sits with the product, not with you. A parent who ignores the label and applies the pillars instead will, on the evidence, choose better nine times out of ten.
The most cheering implication is the one hiding in plain sight. The gold standard the pillars describe, minds-on, focused, meaningful, social, is not an expensive app or a tutoring programme. It is the floor of your living room. The science of learning, assembled with all the rigour a flagship journal demands, ends up describing what children playing with their parents have been doing all along.
Watch the four pillars run, not in a lab but at a kitchen table. A three-year-old is threading wooden beads onto a lace while her mother sits beside her. Pillar one, active: her fingers are doing delicate, deliberate work, steadying the bead, aiming the lace, and her mind is doing more, planning the next colour, holding the pattern she has decided on. Nothing threads itself. Pillar two, engaged: there are no pop-ups in a bead. The smooth weight of the wood, the small click as it slides down to join the others, the slowly lengthening string: the activity holds her attention because the activity is all there is. Pillar three, meaningful: the necklace is for grandma, who is coming on Saturday, and the colours are grandma's favourites. Pillar four, social: her mother is naming colours, counting beads aloud, asking which comes next, and the child is answering. That back-and-forth is precisely the kind of rich, responsive talk Golinkoff's language research has shown builds vocabulary.
No one designed that scene to satisfy a framework from Psychological Science in the Public Interest. It satisfies it anyway, which is rather the point. The same is true of a shape sorter solved in a lap, dominoes clicked end to end while someone counts the spots, a wobbling block tower built with a grandparent who gasps at exactly the right moment. At Jaques London, Britain's oldest toy maker, we have been making this kind of toy since 1795, and we would never claim the four pillars were our idea. They are simply what traditional toys have always quietly delivered. If you want to see what passes the test, the simple, sturdy things in our educational toys collection → are a fair place to look: not because a wooden toy is magic, but because a wooden toy leaves all four pillars to the child and the adult beside her, which is where the learning was always going to happen.
The full review, Putting Education in "Educational" Apps, is long but unusually readable for an academic paper, and worth the time if you have ever hesitated over a download button. For the wider case, start with why real play beats screens, see how the pillars echo through the rest of the research in the science, or browse more researchers like Golinkoff in the Expert Library.
No, and the review does not say so. Its argument is that the label is unregulated, so it cannot be trusted on its own. An app genuinely built on the four pillars, one that makes a child think, holds her attention on the task itself, connects to her life and ideally involves another person, could support learning. The trouble is that very few are built that way, and the label will not tell you which ones. Apply the pillars yourself rather than trusting the word "educational".
Not necessarily in the sense the research means. The engagement pillar is about being focused on the thing to be learned, not merely captivated. Many apps capture attention with sounds, animations and reward loops that actually pull focus away from any learning content. A child can be glued to a screen and mentally idle at the same time. The better question is not "is she absorbed?" but "what exactly is her mind working on?"
Yes, and it is honest to say so. A toy that performs by itself, flashing, singing and narrating at the press of a button, leaves the child passive and distracted in exactly the way a poor app does. The pillars are a test for any plaything, not a verdict in favour of toys as a category. Simple toys tend to pass because they leave the work, and therefore the learning, to the child and whoever is playing with her.
Solo play is valuable and no parent should feel guilty about it; independent, minds-on play still exercises the active, engaged and meaningful pillars. The social pillar is the multiplier, especially for language, so the realistic aim is some genuinely interactive play each day, not constant accompaniment. Ten minutes of real back-and-forth over a threading set or a domino line does more than an hour of supervised silence.
Britain's oldest toy maker. Every toy open-ended, wooden, and built for exactly the kind of play the research recommends.