Slope t statistic after doubling all data values

A simple regression $\hat y=\hat\beta_0+\hat\beta_1x$ gives a $t$ statistic $t_0$ for the two-sided test of $\beta_1=0$. You multiply both coordinates of every point by $2$, refit, and get $t_1$ for the same test on the new slope. What is $t_1^2/t_0^2$? Enter $-1$ if undetermined.

$1$. The slope is unchanged by scaling both axes equally. The residuals double, so the residual standard error doubles, while $\sqrt{S_{xx}}$ also doubles; the slope's standard error is therefore unchanged and so is $t$.