turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- SAS Programming
- /
- SAS Procedures
- /
- An amateur question- Overdispersion error message

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

03-17-2009 11:43 AM

I thought that I was going to test my logit model for overdispersion, and tried to estimate the williams parameter. I got this message:

Overdispersion is not possible for data with a single trial for each binomial experiment.

The Williams method for modelling overdispersion cannot be performed.

I am indeed modeling Y = (0 or 1) so I have r = (0 or 1 ) and 1 observation per individual per event so I assume n = 1 giving me a model statement MODEL R/N = X;

So is it true, in such a case with a dichotomous dependent variable I don't have to worry about over dispersion?

( I tried to test for overdispersion just with MODEL Y = X, but it would not let me specify the SCALE =WILLIAMS statement in this form.

So am I testing for something that is not a potential problem?

Overdispersion is not possible for data with a single trial for each binomial experiment.

The Williams method for modelling overdispersion cannot be performed.

I am indeed modeling Y = (0 or 1) so I have r = (0 or 1 ) and 1 observation per individual per event so I assume n = 1 giving me a model statement MODEL R/N = X;

So is it true, in such a case with a dichotomous dependent variable I don't have to worry about over dispersion?

( I tried to test for overdispersion just with MODEL Y = X, but it would not let me specify the SCALE =WILLIAMS statement in this form.

So am I testing for something that is not a potential problem?