Paired comparison designs are a special case of randomized block designs with block size of 2. Pairing is effective if blocks of size 2 occur naturally such as in animal and medical studies with...

Paired comparison designs are a special case of randomized block designs with

block size of 2. Pairing is effective if blocks of size 2 occur naturally such as in

animal and medical studies with twins, pairs of eyes, and so on. Paired designs

are better than unpaired designs when pairing is effective. Use the paired t test

statistic in (3.1) for testing the difference between two treatments. Incorrect

use of the unpaired t test (also called the two-sample t test) will result in a

severe loss of efficiency. The square of the paired t test is equivalent to the F

test in the ANOVA for randomized block designs.