"...the way to proceed when investigating how the world works is first to carry out experiments and/or make observations of the natural world.

Then, develop hypotheses to explain these observations, and (crucially) use the hypotheses to make predictions about the outcome of future experiments and/or observations.

After comparing the results of those new observations with the predictions of the hypotheses, discard those hypotheses which make false predictions, and retain (at least for the time being) any hypothesis that makes accurate predictions, elevating it to the status of a theory"