Date:         Fri, 4 Feb 2000 19:14:01 -0500
Reply-To:     Stephen Salbod <ssalbod@FSMAIL.PACE.EDU>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Stephen Salbod <ssalbod@FSMAIL.PACE.EDU>
Subject:      McNemar and Cochran Q
Content-Type: text/plain; charset="us-ascii"; format=flowed

I analyze two related binary variables (N =54) using the nonparametric
procedures: McNemar and Cochran Q (k=2). The obtained p values were .227
and .132  for the McNemar and  Cochran Q , respectively. I was wondering
whether this is because the McNemar is an Exact Sig. while the Cochran Q is
Asymp. Any explanation about this difference will be appreciated.

TIA, Steve