A new cream was developed to reduce the irritation caused by poison ivy. To test the effectiveness, researchers placed an ad online asking for volunteers to participate in the study, and 100 subjects replied. They are informed that one group of 50 will receive a new cream and the remaining group of 50 will receive a cream with no active ingredient. The subjects are numbered 1–100 and these numbers are put into a random number generator. The first 50 unique numbers will represent the subjects that will receive the new cream and the remaining 50 subjects will receive the inactive cream. All subjects are exposed to poison ivy and given their cream. They are asked to return in three days and report their level of irritation. The average scores for each cream are compared.

Does this procedure describe a completely randomized design for this experiment?

Yes, the subjects will be exposed to the poison ivy.
Yes, the subjects are randomly assigned to each treatment.
No, the subjects are asked to return in three days. It is possible that some of them may not return, which will introduce bias into the results.
No, the subjects are randomly assigned to the two treatment groups, but the groups are not guaranteed to have 50 subjects in each.