The dark red arrows on the outside run forward, and the last one closes the loop: a conclusion usually raises the next question. The teal arrows on the inside run backward, because you can return to an earlier station at any point. A result that makes no sense often means the setup was wrong, not the world.
Real investigation moves both ways. The circle is the order you learn it in.
The Six Stations
What happens at each one
- Question
- Something you noticed and want to explain. It has to be narrow enough that a test could answer it.
- Hypothesis
- A possible answer, stated plainly enough that a test could show it is wrong. A hypothesis is a statement, not a question — that difference is worth a point on the test by itself.
- Variables
- The one thing you change on purpose is the independent variable. The thing you measure is the dependent variable. Everything else you hold the same, and those are the constants. A control group gets no treatment, so that you have something to compare against.
- Test
- Do it — and then do it again. One trial tells you very little, because any single result can come from chance or from a mistake. When the test asks how an investigation could be improved, the answer is usually more trials, a larger group, a control group, or changing only one thing at a time.
- Data
- What you actually recorded, written down before you decide what it means.
- Conclusion
- What the data supports, and no more. "The hypothesis was wrong" is a real conclusion, not a failed experiment.
The Outer Ring
Can other people get the same result?
A conclusion is not yet knowledge. It becomes knowledge when it can be repeated by people who were not there the first time. That happens in three ways:
- The scientist repeats her own experiment.
- Another laboratory runs the same test with its own equipment.
- Another team asks the same question a different way and gets a comparable answer.
If nobody else can reproduce a result, it does not stand, however famous the person who announced it. In 1989 two chemists called a press conference to say they had produced nuclear fusion in a jar of water at room temperature. Laboratories around the world tried to repeat it. Almost none could, and within a year the claim had collapsed. Nothing was suppressed and no one was silenced. The ring did what it is there to do.
Run it once and you have a story. Run it again and you have a finding. Run it in someone else's hands and you have knowledge.
Two words the test will use. A measurement is reliable when you get the same reading each time you repeat it. It is valid when it measures the thing you say it measures. A bathroom scale that reads ten pounds heavy is perfectly reliable and not valid at all: the same wrong number, every time. One way to keep them straight — reliable is about repeating, valid is about truth.