The scientific method
Open the Choose the next step in an investigation room
This article opens a matching guided lab. Follow the on-screen prompts, make a prediction, and connect the model result back to this topic.
01 // What it is
The scientific method is a process for testing an idea in a way other people can check. It is not a single fixed checklist — different textbooks list four, five, or seven steps — but the core idea is always the same: ask a question you can test, make a prediction, run a fair test, look at what actually happened, and change your thinking if the evidence disagrees with you.
Real scientists rarely move through the steps once in a straight line. A weird result usually sends them back to an earlier step: to run the test again, to control a variable they missed, or to ask a narrower question. Treat the steps below as a loop, not a one-way staircase.
02 // The six steps
1. Ask a question
Notice something and turn it into a question you can actually test, like "does more light make bean plants grow taller?"
2. Do background research
Find out what is already known, so you are not testing something already answered and your hypothesis is grounded in real information.
3. Form a hypothesis
Write a testable if-then-because prediction. See how to write a hypothesis.
4. Test it
Design a fair test: change one independent variable, measure one dependent variable, and hold everything else constant.
5. Analyze the data
Organize your measurements in a table, graph them, and look for a pattern between the variable you changed and the one you measured.
6. Conclude and report
State whether the data supports or does not support your hypothesis, and share your method so someone else could repeat it.
03 // Common mix-ups
- A hypothesis that turns out wrong is not a failed experiment. The test still worked if it was fair — you just learned the answer was no.
- A scientific theory is not a guess. In science, a theory is a well-tested explanation backed by a large body of evidence, like the theory of evolution or plate tectonics.
- A scientific law describes what happens (for example, Newton's laws of motion). A theory explains why it happens. Neither one is "more proven" than the other — they answer different questions.
04 // Treat the method as a loop, not a recipe
The familiar steps are a planning aid, not a rule that every scientist follows in one fixed order. Researchers often review earlier studies before choosing a question, revise a method after a pilot run, or collect new evidence after an unexpected result. The steps are useful because they make the reasoning visible: what was asked, what was changed, what was measured, and why the conclusion follows from the evidence.
Consider a class that notices ice melts at different rates on two surfaces. A testable question identifies the comparison, such as how surface material affects melting time for equal ice cubes. Students predict an outcome, keep cube size and room conditions similar, collect timed observations, graph the pattern, and decide whether the results support the prediction. If the cubes differ greatly in size, they revise the procedure rather than hiding the mismatch.
05 // Choose a measurement before you begin
A question becomes easier to test when its important words have measurable definitions. 'Melt faster' could mean time until no solid ice remains, mass of ice lost in five minutes, or change in size over a fixed period. Those measurements can lead to different procedures. Write the operational definition in the plan, including units, starting conditions, and when a measurement counts as complete.
Plan how many trials are realistic and what will be held constant. Repeated trials help show whether a pattern is reliable or whether one observation was unusual. Set up a table before testing so each result has a clear place. If a measurement depends on judgment, such as deciding whether a tiny ice fragment remains, agree on the same rule before comparing samples.
06 // Separate evidence, interpretation, and conclusion
Evidence is the recorded observation or measurement. Interpretation describes the pattern in those records. A conclusion explains what the pattern means for the original question and names uncertainty or limitations. These parts are connected but not interchangeable. 'The sample took 6 minutes' is a measurement; 'it melted faster than the other sample' is a comparison; 'the tested metal surface was associated with a shorter melt time under these conditions' is a conclusion scoped to the experiment.
Do not treat a hypothesis as a promise that the result must support. If results do not fit the prediction, report them accurately and look for a method issue, an unexpected variable, or a better explanation. Changing the claim to fit the evidence is part of learning. Changing the data to protect a first guess is not. Keep original records, document revisions, and distinguish a corrected procedure from a selective deletion of inconvenient trials.
07 // When a classroom test cannot prove cause
A controlled experiment is not always possible or ethical. Scientists studying storms, earthquakes, or human behavior often observe naturally occurring conditions and compare groups or patterns. They can find evidence consistent with an explanation, but they must consider alternative causes and avoid claiming a single variable caused the outcome when the study did not control it.
After the investigation, share enough detail for another person to understand and repeat the work: materials, measurement rules, number of trials, unusual events, and the reasoning behind the conclusion. A strong next question follows from a remaining uncertainty. The scientific method therefore ends with communication and often begins again, because new evidence can improve a model or suggest a more precise test.
08 // Questions people ask
Is the scientific method always followed in the same order?
No. It is normally taught as a straight list, but real investigations loop back constantly — an odd result sends a scientist back to check a controlled variable, refine the question, or run the test again before drawing any conclusion.
How many steps does the scientific method have?
It depends on the textbook — you will see it broken into four, five, six, or seven steps. This page uses six: question, background research, hypothesis, test, analyze, conclude. All versions cover the same ground; they just split or combine steps differently.
What is the difference between a scientific theory and a scientific law?
A law describes a pattern, usually as a rule or equation (Newton's second law: F = ma). A theory explains why the pattern happens, backed by a large body of evidence gathered over time. A theory does not "graduate" into a law — they are different kinds of statement.