I have a question re bootstrap predicted probability in order logit. I have
an ordinal dependent variable with four categories with one category having
insufficient cases. I have a bootstrap routine to sample with replacement
and estimate teh model, but sometime, the sample does not contain that
category in the dependent variable. How to adjust for that? Should I just
ignore it when calculating the bootstrap average and the percential (or
normal) confidence intervals? Could you suggest some quick reference into
this problem? Really appreciate.
Jun Xu
Doctoral Student
Department of Sociology
Indiana University
_________________________________________________________________
Get 10MB of e-mail storage! Sign up for Hotmail Extra Storage.
http://join.msn.com/?PAGE=features/es
