Hey list,
So, I was chatting with my roommate last night, who was explaining to me a problem he's facing. I am interested in what the list thinks, partly to satiate my own curiosity and partly because I wasn't entirely satisfied with my roomate's idea.
On the surface, the situation is your basic MMR problem: does the relationship between an IV and a DV differ based on the status of some third categorical variable. His problem pertained to international development interventions, which I'm not familiar with, so I won't even bother getting into specifics.
If I recall the conversation correctly, within a given nation, he wants to ascertain whether the relationship between some some predictor variable (we'll say income) and a given outcome differs on the basis of gender. Straightforward enough. However, the nation comprises several provinces or districts between which there may be differences. While he is recognizes and addresses these provincial differences, as well, he wants to control for them to answer the question, controling for provincial differences, does the relationship between income and the DV differ on the basis of gender?
As mentioned, the problem does not seem that complex, but I'm curious nonetheless. Without having anything in front of me to reference, I would think that the GLM procedure could be used to model this. Am I on track? If it's more complicated than I'm assuming, any suggested references?
Best,
Matt
---------------------------------
Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now.
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
