All science experiments begin with an observation. This observation specifically looks for cause and effect. For example, if you noticed that a plant grows better when given fertilizer than a plant without fertilizer, then you could design an experiment that would test if it is the fertilizer that makes a plant grow more quickly, and if so, what sort of fertilizer works best.
A hypothesis is a testable statement that marks what exactly your experiment is testing. A hypothesis emerges from what you've noticed in the observation stage. So, using the fertilizer example, a hypothesis would be, "A plant given regular fertilizer will grow faster than a plant given less fertilizer or no fertilizer." The experiment you design needs to challenge that statement to see if it is valid.
Your experiment must try to support the assertion you made in the hypothesis. Using the fertilizer plant experiment, you need to give plants different amounts of fertilizer in regular intervals (1 cup, 2 cups, etc.) and do so at the same time for each plant. A plant with no fertilizer would be your control, a sample without the variable of the fertilizer to see what happens with no outside influence at all. If you are testing to see if the amount of fertilizer matters, then you need plants all of the same species given differing amounts of the same brand of fertilizer. Your experiment will then show whether the plant given the most fertilizer grows the best, as the hypothesis states.
Take the results from your experiment and compare them to the hypothesis. What you're looking for is to see whether the evidence gathered in your experiment supports the statement of the hypothesis or disproves it. If your hypothesis is disproved, record that and try to determine where you went wrong. If your hypothesis is supported, then you can begin considering ways to expand the experiment and think of ways to test additional variables.