A well-designed experiment will test that prediction. For an experiment to be designed well, it must have good verifiability (ability to show that your idea is true when it is true) as well as good falsifiability (ability to show that your idea is false when it is false).
You do not need to know specifically how to design experiments to pass this course, but you should be aware of the need for falsifiability and verifiability. Some experiments involve detailed lab procedures, precise instruments and planned control of variables. Other experiments are simply quantitative observations and might be better described as tests. Astronomy is a science in which many experiments are essentially what I have just described as tests.
6. Conclusion. This means that you reject or you do not reject the hypothesis. Regardless of the conclusion reached, additional investigations may be performed to gather more evidence regarding the hypothesis. I suppose it's great if the experimental observations (the experimental data) are so conclusive that all rational people will know whether to reject or not reject the hypothesis, just by looking at the data. But rejecting or not rejecting a hypothesis is often based on a statistical analysis of experimental data, and with statistical analysis, you'll have some known uncertainty about your decision. The uncertainty is frequently called error, but it does not imply a mistake or deliberate flaw. Error, in this use, expresses how certain you are in the truth (or correctness) of your decision. A good example of known uncertainty (error) is found with polling results. Those results are often presented this way:.
40% favor the proposition;.
60% are against it;.
with an uncertainty of plus or minus 3%.
That 6% range of uncertainty (error) is an example of known uncertainty, and its size depends on many things, including the size of the sample. All else being equal, a large sample is usually better than a small sample.