The permutation test is based on the premise that by (1) shuffling the domain assignment (i.e., randomly assign the LOR to either PAE or COA), then (2) computing the test statistic based on the random assignment, and (3) repeating this process a very large number of times, the distribution of the test statistic under the null hypothesis (i.e., there is no difference in LOR between PAE and COA studies) is approximated. A frequency plot of the test statistics from all permutations is created, and the permutation p-value is the area under the “curve” to the right of the test statistic computed from the actual data, i.e., permutation p-value = (# of permuted test statistic values to the right of the actual statistic)/total # of permutations.
