Date:         Wed, 30 Dec 2009 13:03:36 +0000
Reply-To:     goladin@gmail.com
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Murphy Choy <goladin@GMAIL.COM>
Subject:      Re: simple question
Comments: To: Adelina Gschwandtner <adelina.gschwandtner@UNIVIE.AC.AT>
Content-Type: text/plain

Hi,

Not sure if this is the right approach but I think that you can do a proportion test in sas using proc ttest.

The approach is to create binary level for each class and then doing a t test on them. I am not entirely sure whether this approach is appropriate but you might want to drop a mail to Art, Peter Flom, Dale Mclarren or David Cassell for more advices on this matter.

If you go to the ucla sas site, they suggested using proc frequency to test the differences. That version uses a chi square approach. In fact, they even have a site that points you to the right test.

www.ats.ucla.edu/stat/mult_pkg/whatstat/choosestat.html

Hope this is helpful!

------Original Message------
From: Adelina Gschwandtner
Sender: SAS(r) Discussion
To: SAS-L@LISTSERV.UGA.EDU
ReplyTo: Adelina Gschwandtner
Subject: simple question
Sent: Dec 30, 2009 8:34 PM

Hello hard working sas colleagues!

I have a very simple question. I have tow samples of different sizes
that contain each a character variable with 4 different parameters (m1,
m2, m3, m4). Is there any possibility to check if the percentage of m1
in sample one and sample two are significantly different? If yes how?

I know how to perform a ttest for two samples of different sizes on the
mean on some specific variable. But I have no clue how to deal when what
I want to check is not the mean but the percentage and when this
variable is not numeric but a character variable.

Please help!!!

Many thanks and all the best for the remaining days of the year,
Adelina.


Sent from my BlackBerry Wireless Handheld

--
Regards,
Murphy Choy

Certified Advanced Programmer for SAS V9
Certified Basic Programmer for SAS V9