$H_0: \mu_{\text{new}} = \mu_{\text{old}}$ vs. $H_a: \mu_{\text{new}} > \mu_{\text{old}}$

$H_0: \mu_{\text{new}} > \mu_{\text{old}}$ vs. $H_a: \mu_{\text{new}} = \mu_{\text{old}}$

$H_0: \text{new method is better}$ vs. $H_a: \text{new method is not better}$

$H_0: \mu_{\text{new}} \neq \mu_{\text{old}}$ vs. $H_a: \mu_{\text{new}} = \mu_{\text{old}}$