A method, not a recipe
The scientific method is a way of finding out about the world by making careful guesses and testing them against evidence. School textbooks often present it as a fixed list of steps. Real scientists do use a pattern like it, but they go back and forth, change their questions halfway and often learn most from results that surprised them. Think of the steps as a rough map, not a rail track.
The best way to see the pattern is to use it on a small problem. This article follows one that most people can imagine: a table lamp that will not light.
Step 1: Observe and ask a question
The starting point is an observation. You switch on the lamp and nothing happens. Observations lead to questions, and a useful question is specific enough to investigate.
“Why is my house cursed?” cannot be tested. “Why does this lamp not light when switched on?” can.
At this stage, gather what you already know. Does the lamp usually work? Did anything change recently? Do other devices on the same socket work? Careful observation avoids wasted effort. Suppose you find that the lamp worked yesterday and that a phone charger in the same socket still works.
Step 2: Form a hypothesis
A hypothesis is a proposed explanation that you can test. It is more than a vague hunch because it makes a claim that could be shown wrong.
Based on your observations, you might propose: “The bulb has burned out.” This fits the facts, since the socket has power and the lamp worked before.
Other hypotheses are possible: the lamp is not plugged in properly, the switch is broken, or the cable is damaged. Good practice is to list several and check the simplest first.
A hypothesis is not the same as a theory in the scientific sense. A scientific theory is a well-tested explanation that draws on a large body of evidence. A hypothesis is a specific proposal that is still waiting to be tested.
Step 3: Make a prediction
A prediction says what you should see if the hypothesis is true. It takes the form “if X, then Y”.
If the bulb has burned out, then putting a different bulb of the same type into the lamp should make it light. And a different working bulb is the only change we make. Equally, if the original bulb is placed in another lamp that works, it should not light.
The prediction matters because it makes the test concrete and shows what would count as evidence against the idea. A hypothesis that could not possibly be shown wrong, whatever you observed, is not useful for testing.
Step 4: Test it fairly
Now you run the test. Changing only the bulb is the key to a fair result, a principle explained in designing a fair experiment. If you replaced the bulb and also moved the lamp to another socket, a success would not tell you which change mattered.
Record what happens. Say you fit a working bulb and the lamp lights. The prediction holds.
Suppose instead nothing happened. The prediction failed, so the hypothesis is not supported. That is still progress: you have eliminated one explanation and can test the next, perhaps the plug or the switch. Experienced investigators treat a failed prediction as information, not as a defeat.
Step 5: Draw a conclusion, carefully
If the new bulb works, you conclude that the burned-out bulb was very likely the problem. Note the careful wording. A single successful test supports a hypothesis; it rarely proves it beyond all doubt. A loose connection might have been jiggled when you handled the lamp, for instance. If you want more confidence, you could test the old bulb in another lamp, where you would expect it not to light.
This is how evidence typically works. Confidence builds as different tests point the same way, and it falls when they do not. Wording like “this supports” or “this is consistent with” is more honest than “this proves”.
Step 6: Share and repeat
Scientists publish their methods so others can repeat the work. Repeating matters because one result might be a fluke, a mistake or a quirk of the particular equipment. In the lamp case, sharing means writing a note for the next person in the house. In wider science, it means reports that others can examine, which is why reading papers is a useful skill even for non-specialists.
What the lamp example leaves out
A household test is simple, and real research is harder in several ways.
- Many things vary at once. Bodies, ecosystems and societies are far messier than a lamp. Researchers use careful design, comparison groups and statistics to separate effects.
- Measurements are imperfect. Readings vary, so results come with a margin, as in measurement uncertainty.
- People are biased. We tend to notice evidence that supports what we already believe. Familiar cognitive biases such as confirmation bias are the reason scientists use blind tests and independent checks. Common logical fallacies can also creep into how results are described.
- Questions evolve. A surprising result often reshapes the question instead of closing it.
Using the method in daily life
You can apply the same habits without a laboratory:
Define the question precisely, list more than one explanation, ask what you would expect to see if each were true, change one thing at a time and be willing to change your mind.
These habits also support information literacy: when you read a claim online, you can ask what the evidence is, how it was tested and whether others have checked it. Projects such as citizen science show that interested non-specialists can take part in real investigations too.
Being willing to be wrong is the core of the approach. A method that lets evidence overrule your preferred answer is what makes the results worth trusting, though even careful studies can reach conclusions that are later revised.