Knowing only what is in your post, that you want to compare the 6 groups, I would say that you want equal sample sizes. If your dv is response rate, you are probably ok. However, you should use poisson estimates in comparing the rates. For other dv's you need to argue strongly that nonresponse is not related to what answer would be given if you could know the "true" answer.
Hope this helps. Art [EMAIL PROTECTED] Social Research Consultants University Park, MD USA Brad Branford wrote: >Hi, > >I'm in the Direct Mail business. Given an expected response rate of >0.50% how do you think the mail quantity (150,000) should be split >among 6 groups (5 test and 1 control) ? > >Thanks in advance. > >Brad > > . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
