Other confounding influences must be controlled for so they don't distort the results, either by holding them constant in the experimental creation of evidence.
There are two research methods for exploring the cause-and-effect relationship between variables: Experiments are typically conducted in laboratories where many or all aspects of the experiment can be tightly controlled to avoid spurious results due to factors other than the hypothesized causative factor(s).
Alternatively, field experiments can be performed, as with medical studies in which subjects may have a great many attributes that cannot be controlled for but in which at least the key hypothesized causative variables can be varied and some of the extraneous attributes can at least be measured.
Field experiments also are sometimes used in economics, such as when two different groups of welfare recipients are given two alternative sets of incentives or opportunities to earn income and the resulting effect on their labor supply is investigated.
In areas such as economics, most empirical research is done on pre-existing data, often collected on a regular basis by a government.