Analysis of Covariance

Analysis of covariance (ANCOVA) allows analyzing the relationship between two variables while controlling for a third variable. In a first step, its effect on the dependent variable is calculated, so that in a second step the unique effect of the independent variable on the dependent variable can be assessed. ANCOVA is useful in analyzing both experimental and observational data, but with different implications. Throughout this entry, the logic of ANCOVA and its assumptions are described. Examples of when and when not to apply ANCOVA are provided for observational as well as experimental data.