I'm working on a model to predict urban bat activity based on park characteristics such as size, distance to water, etc. I did quarterly acoustic monitoring samples in 21 parks, except I missed the first sample period for 3 of the parks (didn't get permission in time.) So I want to be able to run genmod on response variables such as total bat activity and use sampling effort (number of quarters) to remove any bias from the total amounts. Does this make sense? I think I know how to do this in SPSS but not in SAS.

I'd prefer to use all 4 qtrs for the rest of the parks, and to not throw out the parks that only have 3 qtrs. I would appreciate any SAS-specific suggestions on how to do this.
Thanks so much!
Jennifer

Apparently you can do this in proc reg by making the sampling effort (3 or 4 quarters) the first variable and specifying "include=1" after the / in the model statement. But, it doesn't appear to work the same way in genmod.