The Scientific Method Explained

How do we know what we know?

Why do doctors trust some treatments and reject others? How did we discover that the continents drift, that diseases are caused by germs, or that light bends around the Sun? The answer is a powerful way of thinking called the scientific method — humanity's best tool for separating what is real from what merely seems true. It is less a rigid recipe than a disciplined habit of mind, and once you understand it, you can apply it to almost any question.

The cycle of inquiry

The scientific method is usually drawn as a series of steps, but it works as a loop that feeds back on itself:

1. Observe something in the world.
2. Ask a question about it.
3. Form a hypothesis — a possible, testable explanation.
4. Make a prediction about what you should see if the hypothesis is right.
5. Test the prediction with an experiment or careful observation.
6. Analyse the data and compare with your prediction.
7. Conclude and refine — support, reject or adjust the hypothesis, then go round again.

Each loop sharpens our understanding. Let's unpack the parts that matter most.

A hypothesis must be able to be wrong

The single most important idea in science is falsifiability, championed by the philosopher Karl Popper. A scientific hypothesis must make a claim that could, in principle, be proven wrong by evidence.

Compare these two statements:

• "Plants grow taller with more sunlight." → Testable. You can grow plants in different light and measure them.
• "An invisible force makes plants happy, but it leaves no trace anywhere." → Not testable. No possible observation could ever disprove it, so it isn't science.

If no result could ever contradict an idea, that idea sits outside science. The willingness to be wrong is exactly what gives scientific claims their strength.

Controlled experiments and variables

To test a hypothesis fairly, you must isolate cause and effect. That means changing one variable at a time:

• The independent variable is the one thing you deliberately change (e.g. hours of light).
• The dependent variable is what you measure in response (e.g. plant height).
• Controlled variables are everything you keep the same (water, soil, temperature) so they can't muddy the result.

A control group is a comparison kept under normal conditions, with no change applied. If the treated group differs from the control and everything else was held constant, you have strong evidence the change caused the difference. This is the backbone of medical trials, where one group gets a real drug and another gets a placebo. To practise designing such tests, see What Is a Fair Test? Variables Explained.

Data, evidence and honest analysis

Once an experiment runs, scientists collect data and analyse it carefully. Good practice includes:

• Repeating the experiment so a one-off fluke doesn't fool you.
• Using large samples so chance has less influence.
• Looking for patterns and correlations — while remembering that correlation is not causation. Ice-cream sales and sunburn rise together, but ice cream doesn't cause sunburn; hot weather drives both.
• Reporting results honestly, including ones that contradict the hypothesis.

The hallmark of a good scientist is following the evidence even when it overturns a favourite idea.

Hypothesis, theory and law

Everyday language and scientific language differ sharply here, which causes endless confusion.

• A hypothesis is a testable proposed explanation, still under investigation.
• A theory is a broad, well-tested explanation supported by a huge body of evidence — like the theory of evolution by natural selection or the germ theory of disease. Calling something "just a theory" in science is like calling a champion "just a winner." Explore one example in Evolution and Natural Selection.
• A law describes a reliable pattern, often mathematically (such as the law of gravity), but doesn't necessarily explain why it happens.

A theory never "graduates" into a law — they are different kinds of statement.

Peer review: science checks itself

A discovery isn't accepted just because one researcher claims it. Before publication, the work goes through peer review: independent experts scrutinise the methods, data and conclusions, looking for flaws. Then other labs try to reproduce the result. Only findings that survive this gauntlet become trusted. This is why science is self-correcting — errors and even fraud are eventually exposed when results can't be repeated. The system isn't perfect, but no other approach builds knowledge so reliably.

A real example: handwashing

In the 1840s, doctor Ignaz Semmelweis observed that far more mothers died of fever in a ward staffed by doctors than in one staffed by midwives. He hypothesised that something on the doctors' hands — they often came straight from dissecting corpses — was causing it. He predicted that washing hands in a disinfecting solution would lower deaths, and tested it. Death rates plummeted. His evidence was strong, yet many colleagues rejected it because it lacked a known mechanism — germs hadn't been discovered. Decades later, germ theory vindicated him completely. The story shows both the power of evidence and how stubbornly people can resist it. Read more in Germs and Staying Healthy.

Hands-on investigation: the dissolving race

Here is a safe experiment that uses the full method — no hazardous materials.

1. Question: Does temperature affect how fast sugar dissolves in water?
2. Hypothesis: Sugar dissolves faster in hot water than in cold.
3. Set up the test. Fill three identical glasses with the same volume of water — one cold, one room temperature, one warm (warm, not boiling; an adult helps). Add the same amount of sugar to each at the same moment.
4. Control your variables: same water amount, same sugar amount, same stirring (or no stirring) for all three. Temperature is the only thing you change.
5. Collect data. Time how long until the sugar fully disappears in each glass. Repeat the whole experiment three times and average your results.
6. Analyse and conclude. Did warmer water dissolve sugar faster? Does the data support your hypothesis? What could you improve?

You have just run a controlled experiment exactly as a scientist would. For the wider mindset behind it, see How to Be a Scientist: Observe, Predict, Test.

Why this matters

The scientific method is one of the most important inventions in human history. It gave us vaccines, electricity, spaceflight and an understanding of the universe stretching from atoms to galaxies. More than any single fact, it teaches a way of thinking: be curious, demand evidence, test your ideas, listen to criticism, and be ready to change your mind. In a world full of confident claims, those habits are among the most valuable skills you can build.