Dear Statalist,
I have a clogit (or xtlogit, fe) model to describe data on school
district financial outcomes. The fitted model has three polynomial (X2)
terms. I am using "margins" to understand the impact of the X2 terms
over the range of variable values.
My first question is whether anyone has a better idea of how to
interpret an X2 term in a clogit model.
My second question relates to interpretation of the "margins" output:
If I run "margins, at (Var 1 = 100,200,300,400,500) predict (pu0)", I
get a list of average marginal effects associated with each value of the
variable I have requested. My plan at this point is to graph the
marginal effects output (value of variable on x axis, marginal change in
probability on the y-axis), because my understanding is that I cannot
get an absolute probability for specific variable values for clogit
inputs.
If the "margins" output demonstrates the following for "Variable 1":
At 100, ME = 0.292, SE = 0.066, Z = 4.410, P = 0.000
At 200, ME = 0.217, SE = 0.064, Z = 3.370, P = 0.001
At 300, ME = 0.159, SE = 0.064, Z = 2.490, P = 0.013
At 400, ME = 0.118, SE = 0.062, Z = 1.890, P = 0.059
At 500, ME = 0.088, SE = 0.059, Z = 1.490, P = 0.135
...does this then mean, "An increase in Variable 1 from 100 to 200 [a
1-unit change in the data] is associated with a 22% increase in
probability of success", and "An increase from 200 to 300 in Variable 1
is associated with a 16% increase in the probability of success", and so
on?
Any insight or suggestions will be appreciated.
Cora
Please access the attached hyperlink for an important electronic communications disclaimer: http://www.lse.ac.uk/collections/planningAndCorporatePolicy/legalandComplianceTeam/legal/disclaimer.htm
*
* 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/
