Put the stages of the scientific method in the order a researcher actually moves through them.

1. Notice a pattern or phenomenon worth explaining (observation)
2. Propose a testable, falsifiable hypothesis to explain it
3. Derive a specific prediction the hypothesis commits you to
4. Run a study and collect data to test that prediction
5. Draw a conclusion, then invite others to replicate it

Observation seeds a hypothesis, which yields a testable prediction, which a study checks, ending in a conclusion others can replicate. Deriving the prediction before committing to a hypothesis, or 'replicating' before concluding, would break the logic of the cycle.