A pharmaceutical researcher compares the variability in blood pressure reductions between two independent treatment groups. The sample variances are $s_A^2=6.4$ and $s_B^2=4.9$. She defines the $F$ test statistic as $F=\frac{s_A^2}{s_B^2}$.

True or False: An $F$ statistic calculated this way can ever be a negative number.

In the context of hypothesis testing using the $F$-distribution, which of the following is required to determine the critical value for an $F$-test?