| Date: | Fri, 3 Mar 2006 06:57:25 -0800 |
|---|
| Reply-To: | nevin.krishna@GMAIL.COM |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
|---|
| From: | nevin.krishna@GMAIL.COM |
|---|
| Organization: | http://groups.google.com |
|---|
| Subject: | Re: proc genmod-criteria for overdispersion |
|---|
|
| In-Reply-To: | <1141368524.851945.235550@e56g2000cwe.googlegroups.com> |
|---|
| Content-Type: | text/plain; charset="iso-8859-1" |
|---|
Hi Sammy,
Thanks for the response, that was actually another question i had (what
is the appropriate test to evaluate (pearson chi sq vs scaled pearson
X2)..assuming your correct and i should look at the pearson the value i
have is 1.27 is that considered close enough to 1 to assume there is no
overdispersion ?
--thanks, nevin
Sammie wrote:
> It depends on what scale you use. I normally use pearson scale.
> Overdispersion = Pearson Chi/DF
|
