A statistical model in which the main effects explain variance of the dependent variable such that the total portion of explained variance is the sum of the portions explained by each individual effect. This is in contrast to an interaction effect, which is based on a multiplicative model.
See Analyses of variance (ANOVA)