st: query pooling data on dependent variable logit mnl

[Christine Zozula <christine.zozula@gmail.com>]

Hi statalist,

Our dependent variable is dichotomous: whether or not someone received
a sanction for misconduct. But there are actually two kinds of
sanctions (we can call them A and B), and we’re interested in knowing
whether the same variables predict both sanction A and B in the same
way.  In other words, are we justified in combining the two sanction
types into a single (logistic) regression, or must we treat them
separately in something like a multinomial logit regression? There are
no a priori exclusion restrictions that suggest themselves, so what we
really want to do is test whether the same set of explanatory
variables influence getting sanction A (vs no sanction) in the same
way they influence getting sanction B (vs no sanction).  Is there a
way to test whether all the coefficients in one panel of a multinomial
regression are the same as all the coefficients in another panel?

Thanks for your time and attention
--
Christine Zozula
Graduate Assistant
Department of Sociology
University of Connecticut



--
Christine Zozula
Graduate Assistant
Department of Sociology
University of Connecticut

*
*   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/