Date:         Tue, 17 Mar 2009 12:44:22 -0400
Reply-To:     Art@DrKendall.org
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Art Kendall <Arthur.Kendall@VERIZON.NET>
Subject:      Re: ANOVA
Comments: To: nyougo@comcast.net
In-Reply-To:  <249908775.5760991237307900801.JavaMail.root@sz0115a.emeryville.ca.mail.comcast.net>
Content-type: multipart/alternative;

Include them as factors so that you can see both main effects and
interactions.

This is a concept of  accounting for all the dv variance you can.



Art


nyougo@comcast.net wrote:
>
> Art,
>
>
>
> Just a quick question: when I include them as factors should I still
> run interactions on them or limit the analysis so that they are main
> effects only?
>
>
>
> Thanks,
>
>
>
> Nyougo.
>
>
> ----- Original Message -----
> From: "Art Kendall" <Art@DrKendall.org>
> To: nyougo@comcast.net
> Sent: Tuesday, March 17, 2009 12:17:16 PM GMT -05:00 US/Canada Eastern
> Subject: Re: ANOVA
>
> You're welcome.
>
> Art
>
> nyougo@comcast.net wrote:
>
>     Thank you very much. I have 700 cases so I believe this should work.
>
>
>     ----- Original Message -----
>     From: "Art Kendall" <Art@DrKendall.org>
>     To: nyougo@comcast.net
>     Cc: SPSSX-L@LISTSERV.UGA.EDU
>     Sent: Tuesday, March 17, 2009 10:34:14 AM GMT -05:00 US/Canada Eastern
>     Subject: Re: ANOVA
>
>     As long as you have enough cases you just include them as factors
>     in the design.
>
>     Art Kendall
>     Social Research Consultants
>
>     Nyougo Omae. wrote:
>
>         Hello,
>
>
>
>         Does anyone know of a way for controlling for categorical
>         variables (say demographics such as gender etc) when running
>         factorial anova?
>
>         Thanks.
>

[text/html]