Your child is a natural scientist, and the method has three beats: predict, test, observe. You can run all three at the kitchen sink. Ask what your child thinks will happen. Let them test it with their own hands. Then say out loud what actually happened. Toddlers already poke at things to work out what causes what, which Zero to Three calls exploration and discovery. Here’s our position: the words you add afterward matter as much as the activity.
Key Takeaways
- Three beats, one question each: Ask what your child thinks will happen, let them test it with their own hands, then name the result out loud so the discovery gets words.
- A wrong guess is the useful one: Toddler studies found children explored more selectively when an event looked unexplained, so a surprising result is what keeps the testing going.
- Curiosity supplies the motivation: Children invent their own problems to solve, from how high a tower can go to how a pop-up toy opens, and internal motivation drives that discovery.
Is my child doing science, or just making a mess?
Usually both, and the mess is part of how it works. Zero to Three describes science for babies and toddlers as exploration and discovery, inquiry that’s best understood as a “way of thinking and acting”. Nothing in that definition asks for a clean table or a right answer.
Your child sets the agenda, too. Leave a baby or toddler alone with ordinary objects and they’ll invent problems to solve. Zero to Three walks through two of them: a toddler who wants to know how high a block tower can go, and a baby who turns, pushes, bangs, and taps the door of a pop-up toy until it opens. Curiosity and internal motivation are what drive that work. So your raw material shows up free, usually right before dinner. Our guide to inquiry-based learning carries the same idea into the preschool years.
Telling exploration apart from chaos
Here’s the distinction we’d draw. Watch for repetition with variation. When your child pours water from the same cup again and again, each time from a slightly different height, that’s a test. Dumping everything at once and walking off usually means the question is finished, or was never there. Calling repetition with small changes the live signal is our own read of the evidence, and we’d pair it with the serve and return approach that Harvard’s Center on the Developing Child breaks into five simple steps for adults and young children to practise together.
Some spills are just spills, and dinner still has to happen. But before you reach for the towel, give it one beat and see whether a question is being asked with hands. Curiosity keeps its own schedule, so your real work is catching it while it’s happening.
What does predict, test, observe actually look like at home?
A few short sentences wrapped around something your child was already doing. A question about what might happen, a chance to try it, then some plain words naming what came of it. Here are three sequences that need nothing you don’t already own, drawn from the situations Zero to Three uses to describe toddler inquiry.
Three sequences worth trying this week
The block tower. Zero to Three names this one as a problem toddlers set for themselves: how high can this go? Ask before the next block lands. “What do you think happens if we put one more on top?” Then keep your hands down, even when the lean is obvious to you. It falls, and you describe it: the tower got tall, the bottom block was small, over it went. The describing is the part that’s easiest to skip.
The pop-up toy, and a bucket of water. A younger baby predicts with their body, so you narrate: “you’re pushing it, and it opened.” For a sink-or-float trial, Zero to Three suggests a bucket of water, carefully supervised, with seeds, leaves, and pinecones to sort by what sinks and what floats. Ask for a guess on each item first. Our post on sensory bins using backyard finds has more on collecting the materials.
Rainy afternoons suit this better than sunny ones, honestly. Everyone’s already indoors and slightly bored, and rainy-day play ideas tend to be exactly the low-stakes setup a prediction fits into. A guess, a trial, and one sentence naming the result fit inside almost any ordinary afternoon, with no kit required.
How do I ask for a prediction without turning it into a quiz?
Ask it as a real question, with a real chance you’re wrong too. Predictions curdle into quizzes when there’s one answer you’re fishing for, and children hear that in your voice long before they parse the words.
Scripts that stay open. “What do you think will happen?” “Which one do you think gets there first?” “Where should we put it so it goes faster?” Making predictions is one of the cross-cutting practices Zero to Three lists for young children, alongside asking questions, watching cause and effect, and noticing patterns. Its summary of early STEM practice gathers these into a short set of key practices and notes that, done well, they look a lot like play: children moving, doing, testing, discovering. If you want more phrasing, our list of open-ended questions to ask your toddler is built for this.
Wild guesses are welcome here. When your child says the pinecone will sink, they’ve committed to something, which gives the trial stakes and gives you something to name afterward. Leave the answer genuinely open, and your child spends that moment thinking instead of guessing what you want.
What do I say when the prediction is wrong?
Describe what happened, and sound interested while you do it. “You thought it would sink, and look, it’s floating.” A guess that misses is worth just as much as one that lands: either way there’s something real to name.
Research on toddlers does suggest how closely children track events they can’t yet explain. In one toddler study, toddlers watched a toy light up, sometimes on its own and sometimes after an adult acted. The researchers followed where the children looked, what they reached for when prompted, and what they went after with nobody prompting at all. All three pointed the same way. When the light seemed to come on by itself, these toddlers were more likely to treat a nearby button as the cause, and they kept poking at that puzzling version even when no obvious cause sat in front of them. The researchers read it as exploration steered by how well an event explains itself.
So resist tidying up the surprise. Try “huh, that’s not what we thought, want to try the heavy one?” We’d keep building emotional vocabulary in the same conversation as science talk, since a child who can name “surprised” has somewhere to put the feeling. A surprising result is worth lingering on, and a plain description of what just happened is one of the most useful things you can hand your child.
Which materials are worth putting on the table?
Fewer things than you’d expect, chosen on purpose. Zero to Three uses the word provocations for objects you’ve picked deliberately to deepen or extend children’s learning, put out to invite observation and exploration.
In practice, open-ended wins. Water and a few containers of different sizes. Cardboard and a stack of books for a ramp. Seeds, leaves, pinecones, and stones to sort, float, drop, and line up. Toys with one correct outcome get solved once and then abandoned. Give your child something they can vary and it keeps handing them new questions. When you want options that pull a child away from a screen, our roundup of screen-time alternatives leans the same direction.
Safety before the table is set
Water play earns its place, and it carries one condition worth repeating: close, careful supervision, every time, without exception. Yard finds need checking before they join the play. Not every leaf, berry, or seed belongs in a toddler’s hands, so clear anything you’re unsure about with Poison Control before it reaches the table, and our guide to toxic plants and poison control covers what to look for and who to call. While your child is still mouthing everything, follow your pediatrician’s guidance on choking hazards and on what’s safe to leave within reach.
One more practical note: put the materials at your child’s height. Set a bin on the floor and it gets used. Up on the counter, it turns into one more thing you have to remember to offer. A few materials chosen on purpose, with water supervised and yard finds checked, give your child’s thinking plenty to work with.
Do I need to explain why it happened?
Name it first, and naming is often enough on its own. Harvard’s Center on the Developing Child includes “give it a name” among its serve-and-return steps, noting that naming what children are focused on helps them understand the world and gives them words to use.
Zero to Three makes the division of labour plain: children take the lead in discovering, and adults supply language and scaffolding. So your sentence doesn’t need the word density in it. “The heavy one went straight down, and the light one stayed on top” is good science talk for a young child, plain words they can borrow and start using themselves. A What Works Clearinghouse practice guide distils early-childhood math research into a short set of recommendations, two of which are to talk with children about mathematical ideas using mathematical language, and to plan activities deliberately so they build vocabulary and language.
Put those three together and what emerges for a young child is, we’d argue, simpler than it may first appear: much of what the adult contributes is the naming. That’s the beat we’re most confident recommending, because every one of those sources lands on language as the adult’s contribution. What none of them settles is how long to keep at it, or which of your child’s questions is worth chasing. That part is yours, and it depends on the child in front of you. If you want somewhere to put the practice, ordinary play is it, and our piece on how to build language skills during playtime and the one on what early math looks like before age three both run on the same engine.
How young is too young to start?
Infancy, honestly. A baby working out the door on a pop-up toy is already testing, and Zero to Three describes that sequence of turning, pushing, banging, and tapping until something gives. Nobody taught that baby a method.
What shifts as children grow is your side of the exchange. A baby gets narration of the body: “you pushed, and it popped.” A toddler may offer a one-word guess, or simply a point, and it’s worth treating that as a prediction, even with no sentence attached. Ask a preschooler for a reason and you’ll get some gloriously strange ones. Our overview of development milestones from birth to five helps when you’re trying to pitch a question at the right level.
| Stage | What you ask for | What you say afterward |
|---|---|---|
| Baby | Little needs to be said here, simply notice what their hands reach for | Narrate the body: “you pushed, and it popped” |
| Toddler | A one-word guess, or a gesture toward something nearby | Name the result plainly |
| Preschooler | A guess, then a reason for it | Describe what happened, then ask what they’d try next |
Don’t wait for language to arrive before you start talking about causes, either. A child who’s been hearing cause-and-effect words all along reaches their own first explanations with the vocabulary already in the room. Testing starts in infancy with hands and mouths, and your prompts grow alongside the child doing the testing.
What does this look like in a preschool classroom?
Busier and more talkative than people expect. Children moving, doing, testing, and discovering is how Zero to Three characterises early STEM practice when it’s working, and a strong preschool room sounds like that. Cresthill Academy is an early-childhood education and child-care provider, and what follows is what we’d listen for in any program you’re considering.
Look at the materials and where they sit. Open-ended things at child height, in quantities a group can share, get used without an adult handing them out. Then listen to the talk. A teacher who follows a child’s focus and names it, who asks for a guess before the demonstration, and who answers an unanswerable question with “let’s find out” is running the same three beats you run at home. That What Works Clearinghouse practice guide asks for conversations about mathematical ideas and deliberate planning for vocabulary, and you can hear both of those from the doorway.
What to watch for when you visit
Small groups and familiar faces are what we’d look at first. A child who knows the adults in the room has someone to bring a discovery back to, and a teacher who knows your child can pick up a question your child left hanging. Ask who’s with the children at each point in the day, how long the teachers have been there, and what happened with the last question a child asked that nobody could answer. Our guide to responsive caregiving on a tour lists more of these observable signals.
Children moving, doing, testing, and talking about what they found: that’s the sound of a preschool room where inquiry is actually happening.
What if my child asks something I cannot answer?
Say “I don’t know, let’s find out.” That sentence costs you nothing, and it shows your child that not knowing is where the work starts.
Then make good on it in a small way. Write the question down somewhere visible. Pick one thing you can do about it: a trial at the sink, a look at a library book, a walk to go see the thing in question. Harvard’s serve-and-return steps include naming what a child is focused on, and a question written on the fridge is a kind of naming that stays put. Our post on turning one question into a week of play is built around exactly this.
Outdoors tends to generate the best unanswerable questions, which is one reason we keep coming back to it, our post on outdoor time and executive function picks up that thread. Why does the puddle shrink? Where do the worms go? An unanswered question has real value, because “let’s find out” turns one question into a run of small tests.
How do I keep this going on an ordinary Tuesday?
Attach it to something you already do, so it survives a tired evening. Snack, bath, and the walk home each hold one prediction question with no extra setup, and there’s no need to stack them up.
At snack: which do you think floats in your milk, the cracker or the blueberry? In the bath: what happens if we fill the cup underwater? On the walk: do you think the sprinkler is on today? Then let them test it, and name the result in one sentence. That last beat is where the What Works Clearinghouse recommendation on deliberately planning for vocabulary and language quietly gets satisfied, with no plan on paper.
Some days you’ll do none of this, and that’s fine. Children keep experimenting whether or not you’re narrating, since curiosity and internal motivation drive the discovery, as Zero to Three has it. A child playing alone is running these cycles without you, which is part of why fostering independent play pays off and why play-based learning holds up so well under scrutiny. Attach one prediction question to something you already do every day, and the habit survives the evenings when you’ve got nothing left to give.
| Fact | Detail |
|---|---|
| Founded | 2010 |
| Locations | 8 New Jersey campuses across East Hanover, Harrison, Hoboken, Lyndhurst, Paramus, and Parsippany-Troy Hills |
| Curriculum | EsteamED Curriculum® (proprietary) |
| Age range served | 6 weeks – 5 years |
| Programs | Infants, Toddlers, Preschool |
| Hours | Monday-Friday, open 7:00 AM-7:30 AM, close 6:00 PM-7:00 PM (varies by campus) |
| Languages introduced | American Sign Language and Spanish at every campus; Mandarin immersion at the Hoboken campuses |
| Tagline | Exceptional Early Education™ |
Frequently Asked Questions
Is my child doing science, or just making a mess? Usually both, and the mess is part of how it works. Zero to Three describes science for babies and toddlers as exploration and discovery, inquiry that's best understood as a "way of thinking and acting". Nothing in that definition asks for a clean table or a right answer.
What does predict, test, observe actually look like at home? A few short sentences wrapped around something your child was already doing. A question about what might happen, a chance to try it, then some plain words naming what came of it. Here are three sequences that need nothing you don't already own, drawn from the situations Zero to Three uses to describe toddler inquiry.
How do I ask for a prediction without turning it into a quiz? Ask it as a real question, with a real chance you're wrong too. Predictions curdle into quizzes when there's one answer you're fishing for, and children hear that in your voice long before they parse the words.
What do I say when the prediction is wrong? Describe what happened, and sound interested while you do it. "You thought it would sink, and look, it's floating." A guess that misses is worth just as much as one that lands: either way there's something real to name.
Which materials are worth putting on the table? Fewer things than you'd expect, chosen on purpose. Zero to Three uses the word provocations for objects you've picked deliberately to deepen or extend children's learning, put out to invite observation and exploration.
