PROC GENMOD question

[deleted_user]

I'm using a PROC GENMOD procedure to analyze differences in slopes between learning curves following Hartz et al 2001 (http://www.springerlink.com/content/u9bjxekyllj264g1).

Basically I'm analyzing differences in the rate of learning between two types of foragers from multiple colonies. The current model just analyzes group (or forager type) over trials to determine differences in slope or learning rates. I'd like to add colony to the model so that all of my analyses are done at once (I need to determine if there is also a colony effect).

The current main effects in the model are trial and group. I want to add another main effect: colony.

I'd like to know if anyone knows if I need to add the main effect for analysis as trial * colony or just as colony.

Thanks