Consider the multiple linear regression model with an interaction term: $\hat{y} = 5.0 + 2.0x_1 + 3.0x_2 + 1.5(x_1 x_2)$. If $x_2$ is held constant at $x_2 = 4.0$, what is the effect of a 1-unit increase in $x_1$ on the predicted value $\hat{y}$?