Date:         Fri, 5 Aug 2005 18:56:10 -0400
Reply-To:     Andrea <andreawknospam@NOSPAMCOX.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Andrea <andreawknospam@NOSPAMCOX.NET>
Organization: Cox Communications
Subject:      Re: Correct syntax for GLM statement
Comments: To: sas-l@uga.edu

I think that I need results for the group as well as looking at how well
individuals did. What procedure did you have in mind for that?

-Andrea

*Please remove NOSPAM to contact me by e-mail.*

"BILL WEST" <wm_a_west@YAHOO.COM> wrote in message
news:20050805203648.91441.qmail@web34310.mail.mud.yahoo.com...
> I'm not so sure this is best approach as the post test scores could be
> higher as a group but inconsistent. I'd suggest some other approach seeing
> how well each person improved, predicting post test by pretest score for
> same person.
>
>
> Xiangyang Ye <xiangyang.ye@GMAIL.COM> wrote:
> Andrea,
>
> You have 2 variables, one for pre-test score, and the other for
> post-test score. In proc anova or proc glm, you have to put the 2
> variables into one variable, and create a dummy variable to show the
> score is pre- or post-test.
> For example, suppose you have the following real scores
>
> pre-test post-test
> 87 76
> 55 87
> 67 88
>
> Your SAS dataset should like this(I use 1, 2 as group dummy variable.
> '1' for pre-test score and '2' for post-test score):
> group score
> 1 87
> 1 55
> 1 67
> 2 76
> 2 87
> 2 88
>
> Based on the above dataset, you can use proc glm or proc anova(since
> your dataset is balanced) to get the expected result.
>
> Hope it helps.
>
> Sean
>
>
> Regards,
>
> Bill West
> Home # 978 635 1754
> Cell # 978 394 1746
>
> ---------------------------------
> Start your day with Yahoo! - make it your home page