Imagine that we're doing a comparison of the variances of two groups, but since the data is not normally distributed we want to check the results of our F test (since the normality assumption is violated) with a more sophisticated technique. We take our two original data sets, combine them, randomize the order of the values, generate two new data sets of the same size as the originals by dividing this new set in half, compute the variance ratio of these new data sets, and repeat this procedure 1,000 times. When we use the proportion of the new sets that have an F ratio higher than our original F value as our p value for the ratio, what technique have we employed? A two-way ANOVA Bootstrapping A Parameterization test A one-way ANOVA A Permutation test