Hello all,
Piggybacking as I've appreciated this thread and particularly Clyde's points since I am in the process of doing more research on my data analysis technique. Are there citations I can use that support Clyde's posts in #2 and #4?
I am fitting a linear GEE model to my data (where my DV has 80% zeros and also takes on only positive values (except in some rare cases), about 20,000 observations) and I understand GEE has weaker distributional assumptions since it uses quasi-likelihood estimation as I understand. Is this correct about GEE? Also, if it does have weaker distributional assumptions, does it even matter or could I fit a linear model anyway.
Thanks
Piggybacking as I've appreciated this thread and particularly Clyde's points since I am in the process of doing more research on my data analysis technique. Are there citations I can use that support Clyde's posts in #2 and #4?
I am fitting a linear GEE model to my data (where my DV has 80% zeros and also takes on only positive values (except in some rare cases), about 20,000 observations) and I understand GEE has weaker distributional assumptions since it uses quasi-likelihood estimation as I understand. Is this correct about GEE? Also, if it does have weaker distributional assumptions, does it even matter or could I fit a linear model anyway.
Thanks
Comment