In science, an experiment is a carefully controlled study of a phenomenon through the manipulation of conditions variables. For example, in a clinical trial of a new drug, individuals may be randomly assigned to groups that receive either the new drug or a placebo; the responses of the people under the two treatments are then analyzed to determine whether the drug had an effect significantly different from (or better than) that provided by the placebo.
While the scientific method is the universal application for how experiments are to be performed there are common indicators that can indicate an experiment was performed in a manner that eliminates bias.
Experiments are sometimes contrasted with observational studies in which observations are made of pre-existing populations without direct manipulation by the researcher. For example, one may study patients who undergo open-heart surgery to determine which factors, such as age, gender, race or socio-economic status, influence post-operative health. Often observational studies are undertaken rather than true experiments because of ethical, legal or practical considerations involved with the kinds of treatments which would have to be imposed in an experimental setting.