Additive Model

In one-way ANOVA, the additive model is

$$ y_{ij} = \mu + \tau_i + \varepsilon_{ij}. $$

This says each observation is the sum of an overall level, a treatment effect, and a random error term. The model is called additive because treatment and error enter by addition rather than interaction.