A researcher wants to compare the variances of two independent normal populations $N(\mu_1, \sigma_1^2)$ and $N(\mu_2, \sigma_2^2)$. Which test is appropriate for $H_0: \sigma_1^2 = \sigma_2^2$?