Dear List,
so this is an interesting case: having the choice using either iweights or
pweights.
May I ask, what is the difference? How does it affect the estimation
results?
When doing regressions, I always use aweights. I'm a bit ignorant about
pweights but my guess is that they give the same results. However iweights
are very different! When you use iweights (or fweights) in your regression,
you sort of 'blow up' your dataset. That is, you realy *multiply* each
observation with the number in your weight-variable. So if you have 100
observations, and the average of your weight-variable is, say, 11.5, then
Stata will run a regression on 1150 observations. If, as in most cases, the
weights are used to correct for over/undersampling, this is incorrect. As a
result your standard errors will be far too small and significance test wil
bee invalid (unless you really have 1150 cases in your sample, but then you
don't need iweights).
