When does a scientist test a hypothesis?

You test a hypothesis after you form it. But lets go over the scientific method anyway.

You start with a situation, then you come up with a question, then a hypothesis:

there are two kinds of hypothesis, null and alternate, null means that the results dictate that the treatment shows no Significance, where alternate shows that the groups have difference, which is significance.

You test the hypothesis in an experiment, and there are many different tests that you can apply to the DATA you collect.