Confusing two or more factors so their effects cannot be separated; when two factors are set at the same levels throughout the experiment; that is, the columns are 100 percent correlated. See also confounding.