Ask, predict, test, look, decide — then go around again. Here is the simple cycle that turns plain curiosity into knowledge you can actually trust.
Take a lapScientists don't march up a tidy staircase to The Answer. They go around a circle — ask a question, make a prediction, test it, look at what actually happened, decide what it means — and then loop back to ask a sharper question.
Each lap doesn't have to be perfect. It just has to teach you enough to make the next lap better. That patient circling is how a wild hunch slowly turns into something you can actually believe. By the end of this page you'll have walked the loop yourself, run a couple of real little investigations, built a fair test, and spotted the myths people get wrong about how science works.
Press Next step (or tap any stage) to walk the glow around the ring. Watch a real worked example grow as you go — and keep an eye on the loop-back arrow at the end.
It starts with a real itch of curiosity. You notice something odd and wonder why — then you shape that wonder into one clear question you could actually answer. Scientists call this your research question.
Example: you always study with music on. So — does music actually help you focus, or is it secretly distracting you?
↻ The loop-back arrow is glowing — one lap rarely settles a question, so you head back to "Ask" with something sharper to try.
Stage 1 of 5 is glowing. Every investigation begins as a plain, honest question.
On a staircase you climb once and you're done — and if a step is rotten, you're stuck. A loop is different: every time around, you carry what you just learned into a better question. Press play and watch the difference.
The climber on the left reaches the top and stops — that's it. The dot on the right just keeps circling, and each lap is a fresh chance to learn.
Real scientists loop back all the time. A surprising result sends them back to re-test. A messy test sends them back to redesign it. A weird question sends them back to ask it more clearly. "Failing" is just information for the next lap — which is why no single trip around the circle is ever truly wasted. The loop is also why a single experiment never has the final word: it is one lap among many, and the picture only gets sharper as the laps add up.
You just walked them on the cycle. Here they are side by side, with the plain words on top and the words a real scientist (or your IB lab report) would use underneath. They mean exactly the same thing.
Turn a fuzzy "huh, that's weird" into one clear thing you could actually find out.
science word: research questionA careful guess of the form "if I do this, then that will happen" — written down before you test.
science word: hypothesisA fair test that changes one thing on purpose and keeps everything else the same.
science word: experimentWrite down what actually happened — the numbers and observations — exactly as they came out.
science word: data & evidenceSay whether the evidence backed your guess or knocked it down — then ask the next question.
science word: conclusionEvery conclusion hands you a sharper question. So you start again, a little less wrong than before.
science word: iterationDon't worry about memorising the fancy words. The shape is what matters: a question becomes a guess, a guess becomes a test, a test becomes evidence, evidence becomes a decision — and the decision becomes the next question.
The first two stages live or die together. A great hypothesis (say it: hy-POTH-uh-sis) is just a prediction you could check — a "if I do this, then that should happen." The secret rule sounds backwards: a good prediction has to be the kind of thing a test could prove wrong. If nothing could ever disprove it, then no experiment can ever support it either, so it isn't really science yet.
Compare two sentences. "Music is amazing" can't be tested — there's no measurement of "amazing." But "I'll remember more words after studying with calm music than in silence" can: you can count the words. The second one risks being wrong, and that risk is exactly what makes it useful.
Below are six statements. For each one, decide: is it a testable prediction (something a fair test could check) or just an opinion or wish (something no test could ever settle)?
Calm music will help me remember more words than studying in silence.
Read it, then choose.
Notice the pattern: testable predictions name something you could measure — words remembered, centimetres grown, minutes melted. Opinions and "nothing could ever prove me wrong" claims slip right through your fingers, because there's no number to check.
Back to our example: does music help you focus? To trust the answer, you change only the music and hold everything else steady. Tap a card below to "change" it too — and watch a fair test fall apart.
Fair test ✓ — only the music is different, so only the music can explain the result. You can believe the answer.
When just one thing changes, the result has only one thing to blame. Scientists call the thing they change on purpose the variable, and the things they freeze the controls. A test where two or three things changed at once is a muddle: even if the scores differ, you'd never know which change did it. Keeping everything else steady is the whole reason a fair test is worth trusting.
Let's run the whole loop, start to finish, on a question you could test in your own kitchen. Watch how each stage hands off to the next.
Your kitchen has two paper-towel brands — a cheap thin one and a pricey thick one. You wonder: does the expensive one really hold more water, or are you paying extra for nothing?
You write your guess down first: "I predict the thick, pricey towel will soak up more water than the thin one." It's bold and checkable — the test could easily prove it wrong.
Cut one sheet of each to the exact same size. Dunk each in a bowl with the same amount of water for the same ten seconds, then lift it out the same way. The only thing different is the brand — everything else is a control.
The soaked thick towel weighs 22 grams of water; the thin one weighs 14 grams. You write both numbers down — exactly as the scale reads them, no rounding to make your guess look good.
The thick towel really did soak up more water, so your hypothesis was supported. But it's three times the price for only about 1.5 times the water. That's a new question worth a new lap.
Same five moves, a completely different question. Notice that this time the surprise comes with a brand-new science word at the end.
The plant on the windowsill is leaning toward the glass. You wonder: do bean plants really grow toward light on purpose, or did this one just happen to lean that way?
"I predict that if a bean plant gets light from only one side, its stem will bend toward that side within a week." Clear, checkable, and easy to imagine being wrong.
Grow two identical seedlings. One sits by a window with light from one side; the other sits under a lamp directly overhead. Same pot, same soil, same water, same warmth — only the direction of the light changes.
The windowsill plant has clearly leaned toward the glass; the lamp plant grew straight up. You sketch both and note the lean — that drawing is real evidence too, not just numbers.
The evidence backs your hypothesis: the plant grew toward the light. There's even a name for it — phototropism ("photo" = light, "tropism" = turning). New question: how fast does it turn if you spin the pot around?
Here are the five steps of a new little experiment — about a melting ice cube — but they've been shuffled. Click them in the order they should actually happen, from first to last. When you've placed all five, they'll check themselves.
Click the steps in the order they should happen — first one first.
If you got tangled, remember the shape of the loop: you can't test a guess you haven't made yet, and you can't decide what it means until you've looked at the results. The order is always Ask → Predict → Test → Look → Decide.
These ideas sound right, which is exactly why they fool people. Here's the myth — and what's actually going on.
"Science is a straight staircase that ends when you reach The Answer."
It's a loop. Every answer opens a sharper question, so good scientists are never quite finished — they just know more than they did last lap. "Done" isn't really a place science visits.
"Once science proves something, it's certain forever."
Science doesn't deal in "certain forever." Each fair test adds confidence, but a surprising new result can always send everyone back around the loop. A scientific theory — like the theory of gravity — isn't a wild guess; it's an explanation that has survived thousands of laps and could still be sharpened tomorrow.
"A test that doesn't work is a failed, wasted experiment."
A result that knocks down your prediction is still information — it tells you which path is a dead end, so your next lap starts smarter. In science, "I was wrong" is one of the most useful sentences there is.
"Real scientists just know the answers — they don't guess."
Scientists guess and check exactly like you do. The difference is they write the guess down first, test it fairly, and let the evidence — not their pride — decide. Then other people run the same lap to make sure.
It isn't only for people in white coats. Once you know the shape, you'll start spotting it in places that don't look like science at all.
Before a new medicine is allowed, half the people get the real pill and half get a fake one (a placebo), with only the medicine different. That's a giant fair test.
A forecaster predicts tomorrow's weather, then checks what actually happens. Every miss is a lap that makes the next forecast a little better.
Designers quietly try two versions of a level on different players (an "A/B test"), watch which one people enjoy more, then keep the winner.
Change one thing in a recipe — a little less sugar — bake it, taste it, decide. That's the loop running in your own kitchen.
Before a real bridge is built, engineers test models and materials, predict how much weight they'll hold, then push them until they break to see if the prediction was right.
A coach tries a new training drill, measures whether sprint times drop, and keeps it only if the evidence says it actually helped.
Here's the part most people miss: the loop is bigger than one person. When scientists finish a lap, they share exactly what they did so others can run the same test — that's called reproducing a result. If lots of people all around the world go around the loop and land in the same place, everyone's confidence grows. If they don't, the surprise sends the whole field back around again. Checking each other's work like this is called peer review, and it's the reason science can be trusted even though no single person is ever completely certain.
Read each line from a real little paper-plane experiment and pick the stage it belongs to. Friendly hints if you miss one.
"I wonder if a paper plane with wider wings flies farther than a skinny one."
Pick the stage you think this line belongs to.
It's not a one-way staircase. It's a circle you walk again and again, getting a little less wrong each time.
Turn curiosity into a clear question, then a guess you could actually be wrong about.
Run a fair test that changes one thing — then write down what truly happened.
Say what it means, then go around again with a sharper question.