1. Is the F test of two population variances extremely sensitive to violations of the assumption that the populations are normally distributed? What is the key implication of this?
2. Explain why we cannot merely place a minus sign on the upper-tail critical value and obtain the lower-tail critical value when using the F distribution.