Suppose we used a crossover design to test for differences in bruising from subcutaneous sodium heparin injections at three sites in a sample of 15 patients. Surface area of bruises (measured in mm2) and is measured 72 hours after injection, which are administered in random order at 8-hour intervals. Use the following to compute the F statistic to determine if there were significant bruising by site at the 0.05 interval.

Means:

Arm = 212.0 mm2

Leg = 99.0 mm2

Abdomen = 93.0 mm2

Sum of Squares:

SSsite = 17,993.00

SSerror = 48,349.00