As part of a larger experiment, 32 male hamsters were assigned to four treatments in a completely randomized fashion, eight hamsters per treatment. The treatments were 0, 1, 10, and 100 nmole of melatonin daily, 1 hour prior to lights out for 12 weeks. The response was paired testes weight (in mg). Below are the means and standard deviations for each treatment group (data from Rollag 1982). What is the problem with these data and what needs to be done to fix it?
