Date:         Thu, 14 Dec 2006 08:52:11 +0100
Reply-To:     Marco Venus <dutch.linguistics@gmail.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Marco Venus <dutch.linguistics@gmail.com>
Subject:      Tukey for unequal levels
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Is it possible/legitimate to carry out pairwise comparisons for a factor
within a MANOVA design if the levels of this factor are unequally
distributed? The levels are distributed as follows: level 1 (36), level 2
(36), level 3 (18) and finally level 4 (54).

Thanks,

MV