In epidemiology, a method of study that can support and test hypotheses about supposed causes. Data are the past characteristics and events that are analyzed.
In epidemiology and medical research, an investigative technique in which cases are selected for study on the basis of the dependent variable, i.e., the presence (study group) or absence (control group) of the condition or disease being investigated. Differences in the rates of the factor, trait, exposure, characteristic, or possible cause (independent variables) are then compared between the two groups. For example, a study might involve two groups of patients from the same population-one that has cancer (study group) and one that does not have cancer (control group). The smoking rates in these otherwise similar groups could then be compared to see if exposure to cigarettes differed between them. It is important to remember that case-control studies cannot prove causation but can only suggest associations between the variables. Because this type of study is retrospective, there is no way to control for bias in the study from differential reporting between the groups, nor to match the two groups as closely as would be necessary to exclude possible confounding factors.