ANOVA

[Nyougo Omae.]

Hello,

 

Does anyone know of a way for controlling for categorical variables (say demographics such as gender etc) when running factorial anova?

Thanks.

[Art Kendall]

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.

[SR Millis-3]

Reformulate the model as a generalized linear model, ie, in this case, as an OLS multiple regression model in which group and gender are covariates in the model.

Scott Millis



--- On Tue, 3/17/09, Nyougo Omae. <[hidden email]> wrote:

> From: Nyougo Omae. <[hidden email]>
> Subject: ANOVA
> To: [hidden email]
> Date: Tuesday, March 17, 2009, 10:22 AM
> Hello,
>
>
>
> Does anyone know of a way for controlling for categorical
> variables (say demographics such as gender etc) when running
> factorial anova?
>
> Thanks.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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

[statisticsdoc]



 

Enter the categorical variables as covariates, but be careful how you code them.  Binary variables,such as gender, do not present a problem, but categorical variables with more than two levels will need to be dummy coded.  For example, if you have data from three locations, and you want to enter location as a covariate, you would need to create two dummy codes representing the three levels of this covariate and enter the dummy codes (the same principle applies to ordinal categorical variables, such as age categories) .

 

Best,

 

Steve Brand 

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of Nyougo Omae.
Sent: Tuesday, March 17, 2009 10:22 AM
To: [hidden email]
Subject: ANOVA

Hello,

 

Does anyone know of a way for controlling for categorical variables (say demographics such as gender etc) when running factorial anova?

Thanks. 

For experienced and personalized consulting, visit www.StatisticsDoc.com 

 

[Art Kendall-2]

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


[hidden email] 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" [hidden email]
To: [hidden email]
Sent: Tuesday, March 17, 2009 12:17:16 PM GMT -05:00 US/Canada Eastern
Subject: Re: ANOVA

You're welcome.

Art

[hidden email] wrote:

Thank you very much. I have 700 cases so I believe this should work.

----- Original Message -----
From: "Art Kendall" [hidden email]
To: [hidden email]
Cc: [hidden email]
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.