Date:         Mon, 22 Feb 2010 08:27:24 +0100
Reply-To:     García-Granero <mgarciagranero@gmail.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         García-Granero <mgarciagranero@gmail.com>
Subject:      Re: GLM Univariate
In-Reply-To:  <874889.867.qm@web76109.mail.sg1.yahoo.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Eins Bernardo wrote:
> Why is the posthoc goes like that?
>

By default, SPSS uses the residual MS to compute Post-hoc tests. In a
nested design, since the F test is computed dividing MS(Group) by
MS(Set), you need that Post-hoc tests uses the same MS to compute the
SE(diff.) for the means comparisons.

HTH,
Marta
>
>
>
>     Be careful when running POST-HOC comparisons for Group (in case of
>     a significant result at Group level):
>
>     /POSTHOC = Group ( TUKEY ) vs Set(Group)
>     /DESIGN = Group Set(Group).
>
>     Regards,
>     Marta
>
>     -- For miscellaneous SPSS related statistical stuff, visit:
>     http://gjyp.nl/marta/
>

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD