Hello. My question is as much a Stata question as it is a general stats question. I have about 9 binary dep vars (y1, y2, ..., y9) that I would like to estimate with the same logit model, in which the indep vars are something like x1, x2, ..., x6. I have no doubt that the y variables are highly correlated. That leads me to believe it would be wrong to separately estimate...
y1 = a11x1 + a12x2 + ... + a14x6
y2 = a21x1 + a22x2 + ... + a24x6
..
..
..
y9 = a91x1 + a92x2 + ... + a94x6
I realize the exact way I rectify this has something to do with the nature of the correlations I want to assume. But, what method(s) should I be boning up on to make sure I adequately consider this issue and model the dep vars accordingly?
Thank you. Lloyd Dumont
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
