Give a six-year-old twenty pieces of dry spaghetti, ten marshmallows and one instruction, build the tallest tower you can, and the tower falls over. It usually falls near the bottom, where a single strand was carrying everything above it. What happens in the two minutes after that is the part that matters, and it is the part most home learning skips entirely: does she build it again, and does she change the bottom or the top?
That is engineering, it costs about the price of a packet of pasta, and it is the letter of STEM that almost never happens at home.
Science is a question habit, not a topic list
For a four- to six-year-old, science is not facts about volcanoes. It is the habit of saying what you think will happen before it happens, then finding out. The prediction is the exercise. Fill a bowl, gather five things off the counter — a grape, a cork, a metal spoon, a plastic bottle top, a candle — and ask which will float. Most children file the grape with the fruit and expect it to bob. Grapes sink. The surprise is what teaches, and the useful follow-up is not "why?" but "what made you think it would float?"
Write the predictions down, even as ticks and crosses on a scrap of paper, because a written prediction cannot be quietly revised after the fact. Being wrong on paper twice a week is how a child learns that wrong is a step in the process rather than a verdict on them.
Somewhere around seven to nine, add one rule: change one thing at a time. Fold four paper planes the same way, alter a single fold on one of them, throw all four from the same spot. Whether that plane flies further is beside the point; the idea is that if you change everything, you learn nothing.
Technology means tools, not tablets
The T is the letter most often mistranslated. For a five-year-old, technology does not mean a device with a screen. It means human-made things that solve a problem. A zip. A tin opener. A bicycle brake. An umbrella. A door handle.
Two activities carry this letter almost by themselves. The first is taking something apart: an old wired keyboard, a dead torch, a wind-up clock, a retired mouse. With a small screwdriver, an adult nearby and no requirement that it ever works again, a six-year-old will spend forty minutes on this and come away knowing that objects have insides. Keep batteries and anything that plugs into a wall out of it, and keep the screws in a jar.
The second is asking, of anything on the table, what people did before it existed. The tin opener is a good place to start, because the honest answer involves a hammer and a chisel and is genuinely awkward. That question is design history, and it is the door through into engineering.
Screen technology has a real place here — typing, block-based coding, drawing tools — but it is one branch rather than the trunk.
Engineering is the letter that gets skipped
Ask parents what they did at home this month and you will hear about reading and sums. Engineering comes up rarely, for practical rather than philosophical reasons. It makes mess. It needs floor space. It takes longer than twenty minutes. Above all it produces failure deliberately, and failure at seven in the evening, when everyone is tired, is unpleasant to sit through.
But the failure is the content. Science ends in an answer and maths ends in a right answer; engineering ends in "it fell over," which is the only one of the three that forces the child to decide what happens next. The loop is: build, watch it fail, work out where it failed, change one thing, build again. A child who has been round that loop fifty times has something that does not arrive from anywhere else.
Constraints are what make it work. An open-ended "make something" produces a five-minute cardboard blob. Specific limits produce engineering:
- A bridge between two chairs 40 cm apart, newspaper and tape only, that holds a tin of beans.
- The tallest free-standing tower from twenty pieces of dry spaghetti and ten marshmallows.
- A boat from one square of kitchen foil that carries twenty coins without sinking.
Language matters more here than anywhere else in STEM. "Good try" closes the loop. "Which part gave way first?" opens it. Point at the joint, not at the child, and let them do the diagnosing; "the bottom bit bent" is a perfectly respectable engineering statement from a six-year-old.
Budget for three attempts. If there is only time for one, the lesson a child takes away is that building things ends in collapse and an adult saying never mind. One twenty-five-minute session with rebuilds in it is worth more than three separate ten-minute builds.
A rough guide by age. Four- and five-year-olds need big parts, cardboard boxes and cushions and kitchen roll tubes, because their hands cannot manage precision yet and the frustration sits in the fingers rather than the thinking. Six to eight can handle tape, string and a stated constraint. Nine to eleven can take a measurable target: spans 50 cm, holds 500 g, rolls three metres.
Maths is measurement before it is worksheets
The maths worth doing at home is mostly the maths with a reason attached to it.
In the kitchen, halve a recipe. Fractions stop being shaded circles and become the real problem of what half of three-quarters of a cup is when you own one measuring cup.
At the shop, from about seven, ask for an estimate of the total before you reach the till. Not the exact figure, a guess to the nearest note. From about nine, two sizes of the same product and the question of which is better value covers division, decimals and a decision that actually gets made.
For four- to six-year-olds, measure with the wrong units on purpose. How many shoes long is the sofa? How many cups fill the saucepan? Non-standard units teach that measuring is repeated comparison, which is the idea a ruler hides.
The habit worth protecting through all of it is the question "how did you work that out?" — asked when the answer is right as well as when it is wrong, so it is never heard as an accusation. Two children in one house will often reach 8 + 7 by different routes, and both routes are real mathematics.
STEM, STEAM, and what your school actually calls it
If you read advice written in another country, the letters may not match the ones your child's school uses. Many educators and school systems add an A for Arts and talk about STEAM, on the argument that design, drawing and making are not decoration but part of how the other subjects get done. Others keep four letters, on the argument that adding more blurs the point. In England, Design and Technology is a named school subject in its own right, so part of the T and the E arrive through the timetable rather than through anything labelled STEM. Elsewhere, "technology" at school can mean workshop and craft skills, or can mean computing, and those two produce very different homework.
None of this deserves parental anxiety, but one practical move falls out of it: ask the school what they actually do, then spend your home time on the letter that is thin. If the school week is heavy on written maths and demonstrated science — a common pattern in many places, because those are the parts that examine cleanly — then engineering is your gap, and twenty-five messy minutes on a Saturday is the correction.
One more thing for families who have moved between countries. The age at which formal instruction begins varies considerably, and so does the order topics arrive in. A child who has not met a topic yet is on a different sequence, not behind.
A week of this, with nothing bought
| Day | Time | From the house | Letter |
|---|---|---|---|
| Monday | 10 min | Bowl of water, five objects, predictions written first | S |
| Tuesday | 15 min | Old torch or keyboard, small screwdriver, jar for screws | T |
| Wednesday | 25 min | Twenty spaghetti, ten marshmallows, a ruler | E |
| Thursday | 10 min | A recipe halved, one measuring cup | M |
| Friday | 20 min | Wednesday's tower rebuilt with one deliberate change | E and S |
That is eighty minutes across a week and it needs a bowl, a screwdriver and a packet of pasta. Keep it that small; coverage is not the point.
You will know it is working from two small behaviours rather than from right answers: the child starts sentences with "what if we" without being prompted, and goes back to a failed build the next day without being asked. Both are worth more than any list of facts.