Date:         Fri, 23 Jul 1999 17:30:14 -0500
Reply-To:     "Nichols, David" <nichols@SPSS.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         "Nichols, David" <nichols@SPSS.COM>
Subject:      Re: NESTED ANALYSES in SPSS

You need to add the main effect of ipgt_grp to the DESIGN subcommand.

David Nichols
Principal Support Statistician and
Manager of Statistical Support
SPSS Inc.

> -----Original Message-----
> From: David Soliz [mailto:kckds@EMAIL.MSN.COM]
> Sent: Friday, July 23, 1999 4:21 PM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: NESTED ANALYSES in SPSS
>
>
> I am doing some analyses for a paper and have been struggling
> with something
> I was wondering if anyone has suggestions.  The analyses are
> for a treatment
> outcome paper examining the impact of three conditions -
> group treatment
> without BPD members, groups treatment with BPD members, and
> WL on various
> symptoms.  I have no problem doing a GLM repeated measures
> analysis but am
> stuck on how to nest groups within the treatment condition.
> Each condition
> has 4 groups within it.  I would ultimately like to know the
> main effect for
> time, the main effect for treatment condition, the main
> effect for group
> nested within treatment condition, and the interaction between
> timeXtreatment condition.  However, I am not getting this no
> matter what I
> try.  The syntax file follows.  What this gives me is the
> main effect for
> time and the interaction between time by group within
> treatment condition.
> If anyone has suggestions for what I should do I would
> greatly appreciate
> it.
>
> bsi01 - is a symptom measure at time 1
> bsitot_p is a symptom measure at time 2
> ipgt_gp is treatment condition
> txgp is for individual groups 4 of which are in each
> treatment condition
>
> GLM
>  bsi01 bsitot_p BY ipgt_grp txgp
>  /WSFACTOR = bsitime 2 REPEATED
>  /METHOD = SSTYPE(3)
>  /INTERCEPT=INCLUDE
>  /CRITERIA = ALPHA(.05)
>  /PRINT = DESCRIPTIVE ETASQ OPOWER PARAMETER HOMOGENEITY
>  /DESIGN txgp WITHIN ipgt_grp .
>
> Any suggestions would be greatly appreciated.  Thanks.  Desperate
>