Sign Test:
The sign test is a nonparametric test used with paired replicates to test for the difference between the 1st and the 2nd measurement in a group of "subjects". For each pair, you assign a "1" if the 1st measurement has the larger value, a "0" if the 2nd measurement has the larger value. The test statistic is the sum of these values, and the question of interest is the frequency with which a random assignment of "1" and "0" within each pair yields as extreme a value (whether high or low) as the observed data.