I use xtivreg with vce(boot, reps(1000)). I turns out that not only the SE errors slightly change if I bootstrap them, also the coefficient changes a bit which I find it a little puzzling. I guess it arises as a consequence of the two-stage procedure. Using bootstrap SE in xtreg does not affect the coefficient.
Any thoughts on that?
Related Posts with xtivreg bootstrap SE
Unable to start SEM Builder - Stata/MP 15.1- Weighting regressions
- How can I calculate how many of my panel persons switched their dummy status between observation periods?
- Confidence Intervals upper bound above 1 (100%) for survey data prevalence
- The independence of irrelevant alternatives assumpition test of multinominal logit has an error
Subscribe to:
Post Comments (Atom)
0 Response to xtivreg bootstrap SE
Post a Comment