There are many different ways in for a researcher to test a hypothesis. The two large methods are non-experimental and experimental (Myers & Hansen, 2012).
In a non-experimental method, there isn’t any manipulation of the antecedent and the researchers are looking for a correlation between A and B.
In experiments, the researchers manipulate an independent variable and look for what happens to the dependent variable and researchers are looking for how A affects B.
Experiments do a better job of assessing causality because there is not a third variable. In an experiment, there is more concrete data and gives insight on what will/should happen.