[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: help with Heckman
I need some help with the FLM and two-step Heckman commands in Stata 9 (or
8). My data requires the use of cluster and individual weighting, which are
causing some problems with the command. If I choose to specify weights
myself, it only allows me to put the weight option at the end of the
substantive equation-do the weights apply to the selection equation as well?
How can I specify for the cluster variable? If I use the svy: Heckman
command, I can not use the twostep option. My questions are: How can I use
Heckman twostep and still weight the analysis appropriately?
As an alternative I estimated a probit model with weights to predict the
probability of the selection variable. I used the results to construct
an inverse mills ratio. I then added the IMR to the substantive OLS
model with weights, in essence controlling for the probability of the
selection variable. Is this an acceptable alternative to using the
I would appreciate any help I can get with this.
Teresa Casey, M.A.
Sociology, UC Davis
Office phone 530-754-9484
* For searches and help try: