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
- /
- Analytics
- /
- Stat Procs
- /
- deviance value for Proc Genmod

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

06-13-2011 10:49 AM

I am running two different genmod models: One is for a binomial distribution and the other is for a normal distribution. My SAS code is very similiar for the two models. However, the output for the binomial distribution does not give me all of the goodness of fit information that I need. Specifically, the binomial distribution output does not give me the criterian of deviance, scaled deviance, pearson chi-square, and scaled pearson X2. Yet, the output for the normal distribution does give me values for these 4 criteria. ??? I need to find out how to get the criterian listed in the output for the binomial model.

You will see that the SAS code for the two different model runs is very similiar.

Here is my SAS code for the binomial distribution:

proc genmod data=analysis desc;

model success = / dist = bin link = logit type3;

run;

Here is my SAS code for the normal distribution:

proc genmod data=analysis desc;

model cpue = / dist = normal link = identity type3;

run;

Any help on how to get SAS to report the 4 criteria for the binomial distribution would be greatly appreciated.

You will see that the SAS code for the two different model runs is very similiar.

Here is my SAS code for the binomial distribution:

proc genmod data=analysis desc;

model success = / dist = bin link = logit type3;

run;

Here is my SAS code for the normal distribution:

proc genmod data=analysis desc;

model cpue = / dist = normal link = identity type3;

run;

Any help on how to get SAS to report the 4 criteria for the binomial distribution would be greatly appreciated.

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

06-14-2011 12:19 PM

Presumably success in your example is a 0/1 variable (not a count out of n). The binomial is really the Bernoulli in this case. As can be shown (see book by Collett on binary data), the deviance (and many related statistics) "can tell us nothing about the agreement between the observations and their corresponding fitted probabilities" for binary data (page 65 in Modelling Binary Data by Collett, 1st edition [1991]). This is because the deviance in this special case depends only on the fitted values and not the observations. Presumably, this is why GENMOD does not display the deviance and related statistics.

Now, if your binomial data was number of successes (s) out of n individuals per sampling unit (n), you could use:

model s/n = / dist=bin link=logit type3;

and you would get the deviance (etc.).

Now, if your binomial data was number of successes (s) out of n individuals per sampling unit (n), you could use:

model s/n = / dist=bin link=logit type3;

and you would get the deviance (etc.).

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

06-30-2011 05:16 PM

In cases where you have predictors in the model, specify the AGGREGATE= option to obtain these statistics. See this usage note: