Confounder

Such a variable "disturbs" the correlation between influencing variable and event investigated in a study, i. e. cause of disease or death. A disease investigated in correlation with ionising radiation is generally not only caused by radiation but also by other influencing parameters. Taking such confounders into consideration is therefore an essential point in the planning and carrying out of an epidemiological study. If this is not done, a correlation that does not exist in reality can be pretended or an existing correlation can be blurred.

If the influence of such additional factors that have not been taken into consideration pretends a correlation that does not exist in reality, this is referred to as "positive confounding", but if it overlaps the real correlation, this is referred to as "negative confounding". Frequently a narrow correlation between the disease frequency and the influencing variable is suggested which does not exist in reality since both are connected with a joint third variable.