Dispersion Statistic in proc logistic.

[GemmaR]

Hi All,

I'm trying to get the dispersion statistic out of proc logistic. It works when I use the aggregate option but not when I don't.

Should the pearson and deviance statistic only be used when subgrouping the observations?

I know that genmod produces the deviance and pearson statistic but in SAS 9.2 you seem to have to aggregate it again.

Am going around in circles with this so any advice would be much appreciated!

[StatDave]

Dispersion cannot be computed for individual data. You must define subpopulations and this is what the AGGREGATE option allows you to do. See this usage note on overdispersion: http://support.sas.com/kb/22/630.html and the link within it to a note on defining subpopulations.

The answers to many questions can be found in the Samples and SAS Notes in our searchable knowledgebase, http://support.sas.com/kb. You can use the search engine there to find the answers you need.