When dealing with experimental methods you systematically vary one or more factors in order to see what effects the changes have on behaviour. An experiment is often performed in a laboratory because it’s easier to control the factors which one systematically wishes to vary. The main advantage is that you can draw conclusions. You can’t do that in the other two methods. One disadvantage is the ethical problem.
You can only do experiments that don’t affect the subjects in a negative way. Consider the belief that alcohol is a sexual stimulant, and that a few drinks can lower sexual inhibitions and increase sexual attraction towards someone else. Experimenters made a few decades ago an experiment about how alcohol affect ones sexuality. The researchers created two drinks that people could not separate apart from taste and the bartenders then fooled the costumers.
There were those who thought they were drinking vodka and really drank it, but there were also those who thought they were drinking vodka, but drank tonic. Regarding the non-alcoholic drink there were those who thought they were drinking tonic and did it, but there were also those who thought they were drinking tonic but in fact they were drinking vodka. This might seem a bit confusing. To make it clearer the table below represents the different possibilities.
The researchers now wanted to study how the sexual arousal had changed. When the subjects had been drinking for a while the researchers observed all of them and the collected data are showed in the table below. A pattern observed between two different variables is called a correlation. The correlational approach is commonly used in non-experimental studies, since finding patterns is one of the important ways of increasing our understanding