Bioequivalence and Bioavailability Forum

Dear MGR!

❝ In this type of study (means conducting in two groups) Is it compulsory to generate a randomization schedule seperately for each group? or can we generate the randomization schedule at one time containing all the 30 subjects?

No, it isn't. But as I said, I would suggest to stratify the randomization for the two groups, i.e., randomize 3 blocks of 6 sequences (18 subjects) in the first group, and 2 blocks of 6 sequences (12 subjects) in the second.
By this procedure you start with a balanced design not only in the entire sample (30 subjects), but also within both groups.
Statistical inbalance may complicate the evaluation. If you randomize the entire data set of 30 subjects and split them afterwards, it's quite likely that you end up with inbalance to some degree in at least one group.

In real life inbalance occurs frequently due to drop-outs anyway. I think it's no good idea, to start a study already in an inbalanced fashion.

See also this thread.