Suppose that you are asked to design an experiment to study the effect of a hormone injection on the weight gain of pregnant rats during gestation. You have decided to inject 6 of the 12 rats available for the experiment and retain the other 6 as controls.
(a) Briefly explain why it is important to randomly divide the rats into the two groups. What might be wrong with the experimental results if you choose to give the hormone treatment to 6 rats that are easy to grab from their cages?
(b) Suppose that the 12 rats are tagged with serial numbers from 1 through 12 and 12 marbles identical in appearance are also numbered from 1 through 12. How can you use these marbles to randomly select the rats in the treatment and control groups?