It's true that experiments are great for the control they allow over variables that might influence your dependent variable. Well-controlled experiments have good 'internal validity', meaning that if you've designed a good experiment, you can be quite certain that any changes in your dependent variable are due to changes you've made in the independent variable (i.e., to your manipulation).
However, with this benefit comes a loss of 'external validity', that is, a loss of generalizability. Because the experimental situation is so tightly controlled and done in a very limited situation, you are not certain that your results would be seen beyond the lab setting. You cannot know that without doing your study in a more natural setting, but if you do that, you lose the high internal validity that comes with an experimental design. This is a constant trade-off in science, and is not ...
The control in experimental studies are determined.