In a linear model $Y = \beta_0 + \beta_1 X + \epsilon$, if we perform a transformation $X^* = aX + b$ and $Y^* = cY + d$, how does the new regression slope $\beta_1^*$ relate to the original slope $\beta_1$?