Suppose that a one-way analysis of variance involves four treatments but that a different number of observations (e.g., $n_{i}$ ) has been taken under each treatment. Assuming that $n_{1}=3, n_{2}=2, n_{3}=4$, and $n_{4}=3$, write down the $\mathbf{y}$ vector and $\mathbf{X}$ matrix for analyzing these data as a multiple regression model. Are any complications introduced by the unbalanced nature of these data?