Clinical trials
Researchers will set up a clinical trial in order to test the effectiveness of a drug or other treatment for a particular disease or disorder. Essentially, a clinical trial is an experiment conducted within a hospital, lab or other controlled environment. For example, researchers might give a carefully screened group of subjects a new allergy medicine and then observe or test their reactions. Subjects need to be carefully screened so that unaccounted-for differences in people do not influence the results. For example, if the study subject group includes people from widely varying age groups or health statuses, some unknown variable from their background -- say, forgetfulness about taking pills or an unknown interaction with an over-the-counter medication or ethnic food -- might be mistaken for an effect of the drug. Clinical trials are good at establishing cause-and-effect relationships -- for instance, proving that diets low in folic acid can cause certain birth defects.

The down side of a clinical study is that its artificiality might mask an effect that would arise in people's real lives. For example, many years ago, an antidepressant drug passed several clinical trials with flying colors, only to have it discovered that it seriously interfered with specific kinds of intellectual performance. The scientists running the clinical trials had tested the drug for all kinds of common side effects -- headache, nausea, dehydration and others -- but had never thought of testing its effect on activities such as playing chess or solving mathematical problems. Because what they test is usually focused on a specific intervention for a specific problem, clinical trials may need to be repeated using different kinds of subjects before their results can be applied to the general public.